1 _:r:~:~.' l _‘ ,.- ...v..:.rx . I....:¢....;..... . 5.2:... .. 9. H4... 2.... 5. ._..1 “I... _ "goaaxl J“ _ .2 9r?” . . . . . . :3... .......r........». 3...... . .z .a. f)...u..t~....f—. . . a (5 9r .v. .I... .,|_..ma... xmm5... a}... . f... a ......m . a... 1.. .21..." ffwm: mm. in: 19.49.”... hummucm.‘ ’- m 130m". Runniflfl» .1. . . . 7 1.21m! , , if: N W ..vv11.v.: . .1...fr v... . $1... LPRO SAM!) min i s: : In .. AL :GNS ‘ SITUA for .2: n. H.123 .. hm .nrllvUrWnn JIM/IL..- 33 H. CTYONj . C E; STRU IN INT ma W ,2 . fl . .. . _ . .. .. v , _ . u .a . u . fl . w . u a r r . . . . . v 4.. . , . . a . , ,. .n Fir-0.. .. 1-5:... -A LIBRARY Michigan $9” UniverSiW This is to certify that the thesis entitled . THE INSTRUCTIONAL PROCESS AS A FUNCTION OF INTERACTIONS AMONG INSTRUCTIONAL SITUATION VARIABLES presented by Elizabeth Joan Salmi has been accepted towards fulfillment of the requirements for Ph.D. Education degree in BMW hhfinpnmuuu 0-7639 ABSTRACT THE INSTRUCTIONAL PROCESS AS A FUNCTION OF INTERACTIONS AMONG INSTRUCTIONAL SITUATION VARIABLES By Elizabeth Joan Salmi The purpose of this study was to examine the components of the instructional situation of a large undergraduate college course involv- ing a large lecture-small discussion section format. Through an interaction detection analysis an attempt was made to: 1) measure learning over time; 2) quantify learner characteristics contributing to performance in the instructional process; 3) measure instructor characteristics relevant to the instructional process; 4) and to relate the interactions of instructor and student character- istics to student learning. On this basis, three conceptualizations were examined: 1) a conception of the instructional process; 2) a conception of a set of instructor characteristics possibly important in problem solving activity among students; 3) and a conception of a strategy for research in the initial stage in the established classroom setting. The first of these conceptions was that of the instructional process defined by the variables in the instructional situation. Speci- fically, the study was concerned with how certain sets of predictor factors (sets of student and instructor characteristics) were related to the instructional process outcome variables. The second conception pertained to the role of the instructor in the instructional process. Elizabeth Joan Salmi The third conception concerned the nature, role and feasibility of research in the ongoing instructional situation. The structure of the research design approximated the instructional process over time by extracting data at different points in time as the instructional process progressed. An attempt was made to maximize explanation by looking at changes in structure as well as changes over time. Eleven analyses of nine course outcomes were made on the basis of three sets of predictor factors. The analysis indicated that coalescence occurred around cer- tain classes or levels of given predictor variables in their description of an outcome, coalescence differing by instructional process conditions surrounding outcomes, contextually and over time. Interactions emerged on three dimensions. The analysis provided examination of the following: 1) interactions of levels of student characteristics to course outcomes, and interactions of levels of instructor characteristics with student characteristic levels to course outcomes; 2) interactions of instruc- tional effects; 3) and interactions with and independent of multicollin- earity. Certain variables Operated differentially as facilitators or barriers to student performance in meeting outcomes. Factors Operating consistently across all levels were those tied to a conceptual approach to learning. There were differing barriers or facilitators for ranges of student performance. An estimate of congruence was provided between the goals and assumptions of this instructional situation and the actual instructional process in terms of the extent to which 1) pro- vision for individual differences among students was realized; 2) the extent to which the instructional process prepared students to use Elizabeth Joan Salmi diSparate information in a problem-solving approach to learning, 3) and the extent to which models of the behavior Suggested as goals for students were provided by the instructors. Elizabeth Joan Salmi disparate information in a problem-solving approach to learning, 3) and the extent to which models of the behavior suggested as goals for students were provided by the instructors. THE INSTRUCTIONAL PROCESS AS A FUNCTION OF INTERACTIONS AMONG INSTRUCTIONAL SITUATION VARIABLES By Elizabeth Joan Salmi A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Educational Psychology 1971 ACKNOWLEDGMENTS This study was part of a project under the authorization of the School of Teach- er Education, College of Education, Michigan State University, and was supported by a grant from the Educational DeveIOpment Project under the direction of the Office of the Provost. The help and concern of Leland Dean, Associate Dean and Director of of the School of Teacher Education, were greatly appreciated, ii TABLE OF CONTENTS Page LIST OF TABLES v LIST OF FIGURES vii Chapter I. THE PROBLEM 1 Need for the Study 1 Purpose of the Study 8 Expectations and Implications 11 Theoretical Considerations 12 Overview 28 II. REVIEW'OF THE LITERATURE AND BACKGROUND OF THE STUDY 30 Review of the Literature 30 Background of the Study Setting 51 Summary 61 III. RATIONALE FOR CHOICE OF FACTORS; INSTRUMENTATION; PROCEDURES 68 Rationale for Choice of Factors 68 Instrumentation 87 Outcome Variables 87 Predictor Variables 91 Student Variables 91 Instructor Variables 101 Procedures 122 Setting and Subjects 123 Design for the Study 123 Strategy for Data Analysis 127 Expectations and Implications 136 Summary 145 IV. DATA DISPLAY AND DISCUSSION OF "IMPORTANCE" 154 Data Display Form 154 Final Course Grade Based on Entering-Course Characteristics Discussed as an Illustration Representative of All Outcome Variables 163 Characteristic Data Patterns 163 Discussion of the Data of Final Course Grade on Entering-Course Characteristics 165 iii Chapter Page Summary Presentation of Data Analysis of Outcome variables 175 Final Course Grade on Mid-Course Variables 175 Final Course Grade on End-of-Course Variables 178 Pretest on Entering-Course Variables 180 Midterm Score on Mid-Course Variables 184 Midterm Grade on Mid-Course Variables 187 Final Exam-1 on End-of-Course Variables 189 Final Exam-2 on End-of-Course Variables 191 Final Exam-Total on End-of-Course Variables 197 Final Exam-Grade on End-of-Course Variables 199 Instructor Grade on End-of—Course Variables 201 Summary 207 V. DISCUSSION OF THE FINDINGS 217 Discussion of the Findings 217 Summary 269 VI. SUMMARY.AND CONCLUSIONS 273 Summary 273 Discussion 282 BIBLIOGRAPHY 302 APPENDIX A Course Criteria, Summary of Predictors, Data 310 Coding For Predictors APPENDIX B Summary of Variables By Number, Mean, Standard Deviation, Skewness, and Kurtosis 314 .APPENDIX C Intercorrelation Matrix of Study variables 315 .APPENDIX D Figures Dl-Dll: Formation of Student Subgroups on Sets of Predictor Variables over Each Outcome 319 .APPENDIX E Tables El-Ell: Proportion of Variation in Each Outcome by Subgroup within Branch Explainable for Each Variable by Predictor Set 341 iv Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table 10 11 12 13 14 15 LIST OF TABLES Summary of Course Predictors Summary of Student Predictor Factors Model of Course Structure and Time Sequence Predictor Factors in Stage I Analysis Predictor Factors in Stage II Analysis Predictor Factors in Stage III Analysis PrOportion of Variation in Final Course Grade by Group within Branch Explainable for Each Entering-Course Variables Proportion of Variation "Explained" in C0urse Outcome Variables by Entering-Course, Mid- Course, and End-of-Course Variables Mean Expected Final Course Grade, by Entering- Course ASpiration-Expectation for Final Course Grade PrOportion of Variation "Explained" in Course Outcome Variables by Course Performance Variables Proportion of Variation "Explained" in Course Outcome Variables by Student Personal Variables Proportion of Variation "Explained" in Course Outcome Variables by Student Personality Variables Proportion of Variation "Explained" in Course Outcome Variables by Student Attitude Variables PrOportion of Variation "Explained" in Course Outcome variables by Instructor Variables MBan.Expected Instructor Grade by Level of Instructor Course Load Across Student GPA/ Midterm Score Grouping Page 74 124 128 129 130 131 158 166 174 219 221 222 223 225 234 Table Table Table Table Table Table 16 17 18 19 20 21 MEan Expected Final Course Grade Using End-of- Course Variables, by Entering-Course ASpiration- Expectation for Final Course Grade Mean Expected Course Grade by Hihg and Low to Normal Student Test Anxiety According to Enter- ing GPA Level Mean Expected Course Grade by High and Low- Average Student Extraversion According to Various Grouping Mean.Expected Course Grade by High and Low Instructor Authoritarianism across Student GPA Performance Groups MEan.Expected Instructor Grade by Low and High Level of Instructor Eysenck Lie Score Across GPA Groupings Mean.Expected Course Grade by High and Low Pretest Performance Across Student GPA Level Groupings vi Page 239 242 245 256 259 264 Figure Figure Figure Figure Figure Figure Figure Figure Figure LIST OF FIGURES Explanation of Final Course Grade by Entering- Course Variables Mean Expected-Course Grade Plotted for Various GPA Groups by Entering-Course Aspiration- Expectation for Course Grade Mean Expected Instructor Grade Plotted by Instructor Course Load Level Across Student GPA/Midterm Groups Mean Expected Final Course Grade Plotted for various GPA, Midterm Performance Level Groups on End-of-Course Analysis by Entering-Course Aspiration-Expectation for Course Grade Mean Expected Course Grade Plotted by High and Low to Normal Student Test Anxiety Across Entering GPA Level Mean.Expected Final Course Grade by High and Lowthquerage Student Extraversion Plotted According to Various Groupings Mean Expected Course Grade Plotted by High and Low Instructor Authoritarianism.Across Student GPA Groups Mean.Expected Instructor Grade Plotted by Low and High Instructor Eysenck Lie Score Across Student GPA Groups Mean Expected Course Grade Plotted Across GPA Level by Pretest Performance Level vii Page 156 173 233 238 241 244 255 258 264 Chapter I THE PROBLEM Research on teaching and learning in the classroom has been criticized by both researchers and those who try to utilize findings from research studies in practical applications, Much of the criticism by researchers focuses on methodology related to problems such as failure in specificity of criterion, in maintaining homogeneity and independence, in clearly outlining research procedures, and in analyzing interactions, Those looking to research findings for aid in improving instruction in classrooms criticize research findings on the grounds that the results or implications are often difficult to generalize into the classroom setting and fail to lead to actual procedural efforts to improve instructional results, These criticisms are confounded by viewpoint differences on theories of teaching and learning as applied to classroom instruction, differences on the sc0pe of the problem to which a researcher should address himself, and differences on the purpose and nature of research strategies utilized in investigating problems, Theoretical differences are illustrated by the contrast between the viewpoint that a theory of teaching is just a variant of traditional theories of learning and hence explainable in terms of the latter, the view that research should focus on teacher Operations since learning is not necessarily the most important criterion of instructor effectiveness, and the view that although inextricably connected to theories of personality, learning, motivation and develOpment, theories of instruction must not be 1 2 develOped solely by inference from these.1 In addition, different views on the sc0pe of the problem to which a researcher should address him- self are due partly to the fact that many areas of education such as instructor effectiveness or creativity are complexes of many factors; hence views range from belief that these complexes present an Opportu- nity for new conceptualization of research designs to belief that the scOpe of the problem should be limited so it is amenable to traditional classical experimentation, These views in turn are compounded by different Opinions on the purpose and nature of research strategies in evaluating these complex factors. One extreme is the large-scale research project in which a curriculum is varied or special classes are established but lack of evaluative criteria or adequate design, because of its complexness, result in no evaluation of results.2 On the other hand, studies utilizing traditional experimental designs in which variables are manipulated and their effects upon other variables observed encounter difficulty in handling complex educational problems, In pointing to these difficulties, Stanley cites as two consequences the tendency to choose trivial problems that lend themselves to neat designs, the results being statistically significant but lacking import- ance, and the tendency in more complex studies to report main effects but with interactions which are left to conjecture so that absence 1N. L. Gage, "Paradigms for Research on Teaching," in N. L. Gage (Ed,), Handbook of Research on Teaching - A Project of the American Educational Research Association, (Chicago: Rand McNally, 1963), pp 0 94-14]- 0 2Frank B. Baker, "Experimental Design Considerations Associated with Large-scale Research Projects," in Julian Stanley (Ed.), Improving Experimental Design and Statistical Analysis, (Chicago: Rand McNally, 1967), pp. 214-217. 3 of significant differences in the findings is reported with monotonous regularity.3 The alternative prOposals for improving educational research on instruction reflect these differing viewpoints in terms of where critics see the need to focus strategies, Stanley cites the need to alter strategies toward two Specific problems in utilizing classical experimental designs. First, he criticizes the number of studies where main effects turn out to be of little or no interest and the interac- tions, vaguely explained, are of primary interest.4 He cites the need for the researcher to do Sufficient preliminary research and to incor- porate this into the design of his investigation, rather than discover- ing after doing the study that he doesn't think the main effects very interesting. His second criticism is leveled at what he refers to as the "frozen methodology" of those who "set up a nice design that (they) like and know how to analyze and that is balanced," and into this model force all theoretical models or problems, by drastic alterations on the latter for the sake of the former, with the end result that findings may be significant but not important or even relevant to the original 5 problem. He is equally critical of the converse of drastic alterations mithe problem, namely forcing the problem onto the statistical model by 3J. C. Stanley, "On Improving Certain ASpects of Educational Experimentation," J. C. Stanley (Ed.), Improving Experimental Design 229 Statistical Analysis, (Chicago: Rand McNally, 1967), pp. 32-33. 4J. C. Stanley, "Symposium: Interactions in Psychometrics and ExPerimentation," Educational and Psychological Measurement, 1961, 2.1. pp. 791-858. 5Stanley, op. cit., in Improving Experimental Design and §tatistical Analysis, pp. 32-34. 4 ignoring the statistical assumptions underlying the model. In either case, he stresses the fallacy of controlling for variables without stopping to consider the kinds of questions that can be answered in such a manner, or whether the kinds of questions to which answers are wanted are actually being posed. Stanley's assertion of the need to identify and explain interactions and to make them an integral part of the research design rather than shrugging them off, and the need to give careful considera- tion to the relationship between the research model and the questions being posed is extended by Box who feels "there tends to be much too much work in which peeple say, 'Given the model....' in the very first sentence of the paper and everything else follows, (when) usually the most important single question is the choice of the model in the first place."6 The need is to select the optimal design for a particular situation. The results of an experiment depend far more on what variables the experimenter decides to include, what range, how many levels, and so on than on anything else, even the data. Thus two equally knowledgeable researchers arrive at somewhat different designs. Box concludes that, "we must regard experimentation as an iteration -- an adaptive learning process, in which we run a series of designs, adapt- ing our strategy to results as they appear."7 Whereas Stanley stresses the importance of incorporating in the initial phase of an investigation 66. Box, "Bayesian Approaches to Some Bothersome Problems in Data Analysis," in Julian Stanley (8d,), Improving Experimental Design fiflyngtatistical Analysis, (Chicago: Rand McNally, 1967), p. 84. 7Ibid., p. 86 -‘_t --v . s 5 the search for interactions, Box adds the necessity that the initial phase provide the researcher with a picture showing the amount and nature of information obtainable at different points at which the research could continue. Although it is difficult to alter the habit of testing for one variable after another, if the data is full of interaction effects it is not meaningful to ask about the effects of one variable at a time. In addition, in the analysis of data derived from experiments involving large numbers of variables or susPected interactions, the argument can be made that significance tests are of doubtful value. Factors may ShOW’up as statistically significant which are not important in terms of reducing error.8 The necessity persists of first locating those variables which seem most important in accounting for the variation in some specified dependent variable and evaluating the relative importance of those variables.9 The problem of findings which may have statistical significance but lack importance is extended by Campbell to the problem of generalizing from experimental findings in laboratory settings to actual classroom settings, since any "extrapolation is never logi- cally certain but it is more plausible the greater the similarity of conditions and the less the distance in time and Space.10 He prOposes 8J. N. Morgan and J. A. Sonquist, "Problems in the Analysis of Data, and a Preposal," Journal of the American Statistical Association, 58, (June, 1963), pp. 415 - 35. 9H. M. Blalock, Jr., "Evaluating the Relative Importance of Variables," American Sociological Review, 26, (1961), pp. 866-874. 10D. T. Campbell, “Administrative Experimentation, Institutional Records, and Nonreactive Measures," in J. Stanley (Ed.), Improving Experimental Design and Statistical Analysis, (Chicago: Rand McNally, 1967), p. 260. 6 that since this implies the necessity to cross-validate in the actual classroom setting the findings from the laboratory situation, the experiments could begin in the classroom setting. This would provide a basis of experimentation where the extrapolation is small, and where the likelihood of valid extrapolation is great. Research is then part of the instructional setting, the measures being an integral element of the instructional process.11 Campbell prOposes the implementation of a research program within the already established classroom setting, the program to keep artificiality and obtrusive measures minimal by making such factors part of the setting, and over time to develOp by replication a body of data to be used as the basis for additional testing of hypotheses about instructional variables, cross-validation occurring by other programs in similar situations. Initially, this necessitates the evaluation of those variables in the instructional complex to identify the factors having the greatest importance for that given instructional situation. The initial purpose, according to Campbell, is the discovery of the structure of relations among variables in a specified instructional setting, and under what conditions and through what intervening processes this relationship OCCUIS . The comments of Stanley, Box, and Campbell summarize the basic assumptions of research procedures -- a model of normalcy and homoge- neity. Studies in the experimental tradition tend to look for dimen- sional differences, relegating everything else to either error or 111bid., p. 260 7 unexplained interaction; studies in the relational tradition tend to look for population relationships and prediction, with non-relationships or inaccurate prediction being classed as error. These assumptions are wholly justified under a tradition of parsimony, of favoring precision and stability so as to arrive at unified conceptions or comprehensive theory. These same principles of homogeneity and normalcy are aSSumed in much classroom instruction, without the traditions underlying their assumptions as in basic research. DeSpite the supposed innovations to provide for individual differences, the majority of classrooms still Operate on such features as singular instructional methods, uniform content, or grading on the curve. However, within the instructional setting a structure of relations exists such that many circumstances with inherent violations of this normalcy and homogeneity become obvious to the point where the instructional efforts are forced to deal with differences and heterogeneity. The purpose of this study is to discover the structure of relations among the variables in a given instructional setting, and under what conditions and through what intervening processes this rela- tionship occurs. While the focus of discussion is the relative import- ance of instructional setting variables in the classroom, rather than methodology, the necessity of dealing with problems of criteria, design, and analysis are inherent. Criteria considerations are based not on whether one mode of presentation is better than another but on which conditions are thought to Optimize the realization of instructional a 8 goals under clearly Specified and delimited conditions. Under such criteria it may be erroneous to make implicit assumptions of homogeneity and independence of empirical conditions in the original design. Since independence of variables cannot be assumed, the need is present to identify and explain interactions rather than leaving them to conjecture in the analysis. Purpose The purpose of this study is to determine the relative import- ance of potentially important variables to relevant course criteria or outcomes in a manner which is functional to improving instructional conditions. Iggortance refers to a given variable in terms of its weight or usefulness to a given instructional setting. In this context four characteristics of importance together provide the basis for evaluation of variables. The first characteristic associated with importance is expressed statistically, in terms of the relative amount of variance accounted for by the particular variable. While the relative amount of variance accounted for generally has been a primary criterion in virtually all research studies in either experimental or naturalistic settings, methodological difficulties continue to result in failure to account for variance that accumulates from an interaction. The second characteristic associated with importance is that of location, inferring in the sense of interaction or of multicollinearity that some variables account for§variance in Special locations. .For example, the researcher may be interested in the effects of test anxiety on student test performance and yet when trying to extrapolate what 9 would be significant findings, in this instance the fact that test anxiety has some negative effect upon the ability to take tests, he finds the compounding factor that students who are highly motivated may suffer certain forms of test anxiety but have found ways, because of the positive effect of grade reward, to compensate for whatever deficits this anxiety creates. In this case there would possibly be an interaction in which course motivation would "wash out" the weight of test anxiety at the upper levels of test performance. Location can also pertain to multicollinearity in the sense that there may be a small but significant correlation between the independent and dependent variable but the correlation is small because while it is consistent across the high range and low range of scores it is either the highest and lowest groups correlating or possibly just the éffect of the middle group minus the tails of the distribution. The third characteristic associated with importance is that of time, referring to the relative appearance or increase and decrease of the variable accounting for a prOportion of variance over the time setting of the course. For example, in the instance where students with knowledge or training in areas similar to course materials being used in a course they have just commenced, it could be that the previous training is an advantage carried consistently throughout the course, or it could be that the differences due to this increased advantage at the start would be lost over time. Or perhaps the vari- able appears simply in relation to one Specific point in time in the course. The fourth characteristic associated with importance is that of criterion, that is, accounting for variance in relation to the 10 nature of the dependent variable. For example, there would seem to be a considerable difference as to what factors account for variance in a variable to which a grade has been attached. The concept of importance embodying the characteristics of relative amount, location, time, and criterion can be discussed on two levels: in relation to the theoretical considerations which are the basis for the course or instructional process, and in relation to the actual outcomes of the course or instructional process. The first level raised the question of what, theoretically, is the relative importance of the instructional variables as embodied in the theoretical considerations behind the course or the instructional situation -- that is, the theoretical considerations which serve as the raison dlétre for the course or instructional situation, the theoretical considerations being expressed in the form of course criteria which themselves are defined by the course objectives. The second level raised the question of what, in fact, is the relative importance of variables derived out of the instructional setting in terms of the actual course outcomes, the outcomes themselves defined in terms of assessed performance. Course develOpment is a function of the evolvement of congruence between these two levels of importance. In this context of the study, variables derived out of the inst- ructional setting are referred to as instructional variables and include a set of learner intellectual and social characteristics, a set of environmental characteristics (including the teacher), and the learning itself in terms of changes of behavior toward specified objectives. The instructional setting is defined to include onlyths formal classrooms. ll Implications and Expectations The central question in this study is the relative impor- tance of instructional variables at two levels -- theoretically and in actual practice -- and the congruence between the levels. Implied in this question is the problem of assessing the internal consistency between course goals and procedures, and student evaluated performance. Thus, the research strategy must first develOp important combinations of instructional setting variables in describing different outcomes, and show how these descriptions change according to outcome by type and time. But equally necessary, in addition to homogeneous descrip- tions and relationships, the strategy must develOp the important heterogeneous explanations and the important exceptions to general prediction. . These implications suggest a number of expectations. First, it would be expected that the results will show coalescence occurring around certain classes of variables, for instance those variables which are the most important initial descriptions of performance. Second, it would be expected that the results will show differences among outcome variables (represented by assessed student performance) in the type of predictor variables accounting for variance and in the extent of the prediction of variance. Third, since measures usually are proxies for a specified theoretical construct, sometimes more than one, several measured factors may together represent a weighting of theoretical constructs in the form of interaction. Fourth, it would be expected that certain instructional setting variables operate in differing sets as barriers to student performance 12 in meeting course criteria and outcomes. Fifth, the results should show important isolated exceptions, normally relegated to measurement or statistical error, which represent real instructional problems. AAnd finally, given the compilation of the above in combination with the course theoretical underpinnings, it would be expected that the results will demonstrate the degree of congruence between course criteria and actual outcomes in terms of relevant variables.12 Theoretical Considerations Many of the criticisms by researchers of instructional studies focus on problems related to methodology and analysis. When the researcher has posed the problem and questions he wishes investigated, he then must ordinarily take into consideration two models. The first of these is the theoretical model upon which the design for the study is based. Whether the design is the type classed by Campbell and Stan- ley as "experimental" or one not involving the same assumptions as the former, the design is a representation of a theoretical model. There must be a functional representation between the questions being posed in the study and those the design is capable of answering.13 One of the reasons designs for instructional research in the classroom setting become complex is because of "the intransigence of the environment: because, that is, of the researcher's lack of 12The expectations are restated in Chapter 3 Specifically in terms of the instructional situation used for the study. 13Fred N. Kerlinger, Foundations of Behavioral Research, (New York: Holt, Rinehart and Winston, Inc., 1964), p. 370. 1 I I: 13 complete control."14 Violations of the theoretical assumptions underlying the design can occur. Two of those most commonly violated are the assumptions of independence and homogeneity. For example, one common method of investigating the latest innovation in instruct- ional procedure (such as the use of independent study) is to employ a design involving control-group comparison, comparing the "innovation" with some more "conventional" procedure. Analysis of the data from the comparison, using traditional approaches to analysis of variance, "assumes that each instructional procedure is independent of others and homogeneous within itself to permit its use as an independent variable."15 However, the first assumption, independent learning environments in the "experimental" and "control" groups is questionable if the teacher used in the two procedures, feeling that the method of presentation for the experimental group improves the quality of their treatment, inadvertently alters methods in the control group to enhance his teaching. Under the same conditions, the second assumption, that of homogeneity, can be questionable if the treatment and the control conditions are regarded as uniform independent variables. Designating the control group "conventional classroom", for instance, and failing to take into account the differences that exist among "conventional classrooms" limits generalizability beyond the specific samples in the study. 14Donald T. Campbell & Julian C. Stanley, Experimental and Quasi-experimental Designs for Research, (Rand McNally, 1963), p. l. 15Laurence Siegel & Lila Siegel, “A Multivariate Paradigm for Educational Researbh," Psychological Bulletin, 1967, 68, 5, p. 308. 14 The research design essentially embodies the methods of securing adequate and proper data to which to apply statistical procedures. One purpose of each major statistical technique can be regarded as to provide in some form an answer to a general problem: "how to abstract from the sum of squares, which is a mélange of error, that portion which represents useful information relative to the questions raised by the researcher in the beginning of his inquiry."16 While the statistical technique is inextricably tied to the research design, it also involves an underlying theoretical model. Many educational research problems outside the experimental laboratory setting have small resemblance to the precise models in the textbook examples. In the analysis of the data the researcher may find that there are statistical difficulties created by the data which make any results questionable or superficial. The data generated from the study variables may reflect any number of conditions which might conflict with the assumptions of a particular statistical model. For instance, use of a large number of variables creates the problem of handling data in the form of a wide variety of information on each individual. Or, the measure- ment of the variable may involve classifications such as a set of classes designating the fields of study in which students are enrolled, or perhaps the data is in the form of answers to attitu- dinal questions which may not really have a rank order. Or the problem 16Solomon Diamond, Information and Error, (New York: Basic Books, 1959), p. vi. 15 may be that even though the measure seems to be a continuous variable, such as test scores, the effect is not necessarily linear; time expended studying does not necessarily change uniformly with changes in test scores at either extreme of the test score range, for example. Or perhaps the problem is intercorrelations between many of the explanatory variables to be used in the analysis, making it difficult to assess the relative importance of various factors, since their intercorrelations get in the way; for example, high grade point average of a student tends to go along with higher level of aspiration and expectation for performance, with lower test anxiety, and so on. Or the problem may be interaction effects. For example, high test anxiety in test-taking situations may have a much more deleterious 'impact among students with average grade points than among those with high GPA'S when it comes to actually making that high grade needed (or wanted) on a test. Or the interaction may exist because the measured classifications are only proxies for the theoretical constructs, and several variables may have to interact to approximate the theoretical construct being investigated in the Study.17 There are a number of methods of handling the problems outlined in the preceding paragraph. The first of these, the use of many factors in the study can be handled by examining each factor one at a time keeping in mind the degree to which the one factor is intercorrelated with the others. A second method is to build combina- tions of factors arbitrarily or via factor analysis. In this latter technique the danger exists of combining factors on the basis of their 17Morgan & Sonquist, op. cit., pp. 415-435. A! l6 influence on the dependent variable may be in opposite directions, thus creating a factor with no correlation with the dependent variable. Or the problem of multiple factors can be handled by the use of multiple correlation techniques such as multiple regression analysis,18 as can the problem of intercorrelations among predictors, assuming the number of explanatory factors is not too large. However, an underlying assumption of multiple regression analysis is variables in continuous form, raising a problem from the preceding paragraph, that of variables in classification form, such as sex or college major. This necessitates construction of arbitrary scales or the use of dummy variables.19 The last of the problems mentioned above is that of the presence of interactions. The assumption that no interactions exist makes for a very efficient analysis procedure, but this assumption, according to Stanley, has been reaponsible for the lack 18Briefly, multiple regression analysis attempts to deter- mine the effect of a given independent variable on the given dependent variable, while holding constant or removing the linear effects of the other independent variables, and to determine whether these effects are significant after taking into account the intercorrelations of the predictors. Thus, if the researcher wished to predict the academic success of freshmen entering college he would attempt to ascribe the correct amount of explained variation in the dependent variable to each factor considered in the prediction, within the limits: of the linear and additive assumptions of the model. See N. Draper, Applied Regression Analysis, (New York: John Wiley and Sons, Inc., 1966). 19The use of dummy variables involves the assignment of a dummy variable to each class of a characteristic except one. It is called a dummy variable because it has the value 'one' if the individual involved belongs in that subclass, or a 'zero' if he does not. See Morgan and Sonquist,'gp,.gig., p. 422, orssee D. Suits, "The Use of Dummy Variables in Regression Equations," Journal of the American Statistical Association, 52, December 1957, pp. 548-551. iii 17 of significant findings related to main effects in many studies.20 One method of identifying an interaction is to perform a separate analysis of tables of residuals (theamount not explained by the regression equation) derived from the initial multiple regression analysis, which isolates some subgroup on a combination of character- istics.20 One dfiiiculty is that data containing a number of complex interactions requires repetitions of the technique to confirm susPiéions. .Another method involves restricting the total number of variables, using cell means as basic data, and using a variance analysis which looks directly for interaction effects. The difficulty is to avoid getting empty cells with very small numbers of cases involved in the analysis. A third method is that with some restrictions on the number of variables many feasible interactions can be built into the initial phase of the research. The difficulty is in knowing which interaction terms to introduce into the regression model in advance, particularly with factors unsupported by a body of data. The need in studies in the'hatural setting' of ongoing classroom activity where a complex network of variables is involved is to take into account the problems raised in the beginning of this chapter by Box and Stanley, that of reconciling the variables to the 20In the separate analysis of residuals derived from a mmltiple regression analysis for the purpose of identifying an interaction, "an estimate of the expected average of that subgroup «on the dependent variable can be derived by summing the multivariate czoefficients multiplied by the subgroup distributions over each of the lpredictors. Comparing this expected value with the actual average for t:hat subgroup indicates whether there is something more than additive effect." See Morgan & Sonquist, pp. 31.13., p. 424. 18 assumptions underlying the methodological and statistical models. Theaaproblems become essential in relation to the problem raised by Campbell, that of develOping research programs in the classroom setting as an integral part of that setting, and particularly acute in the initial phase, that of locating those instructional setting variables which seem most important not just to the outcomes Specified in the particular setting but those which appear important, theoreti- cally, to the improvement of instruction. Blalock points out that in such exploratory phases the job is to locate those variables which seem most important in accounting for variation and to attempt to evaluate the relative importance of such explanatory variables,21 "if only for the practical reason that both theorists and empiricists must limit themselves to a reasonable number of explanatory variables."22 The question arises of what should be the quantitative criteria to use in determining the relative importance of the various factors. One possible criterianis to compute some sort of measure of association between an independent and dependent variable. If there are several independent variables, their relative importance is assessed by comparing measures of association of each independent vari- able with the dependent variable, controlling for all the remaining independent variables. The measure could be some form of correlation coefficient in which case the respective partials are compared. Another 21The importance of a given explanatory factor is also always a function of the amount of variation in that factor. See H. Blalock, Jr., pp. cit., p. 867. 221bid., p. 866. n» t. ' hi 5“ . . a. .._ b. L‘. “ye “ *. 0;, .' .;t ' ET. 4 A '- --.u a \"\ ‘o;_ 19 possible quantitative criterion to use in determining relative import- ance is a measure in the form of a prediction equation in which slopes are used to measure the change in the dependent variable produced by a given change in the independent variable. If there are a number of explanatory variables, as in the case of multiple regression analysis, beta weights can be computed which indicate the change in the dependent variable produced by standardized changes in each explanatory variable, controlling for all remaining variables.23 However, the requirements of linearity and additivity assumed in multiple regression analysis makes the use of regression equations, such as prediction equations, difficult when the data involves many cases, much classificatory information, many intercorrelations, and many complex interaction effects. The more the theoretical and Statistical assumptions imposed on the data, the greater the reduction in the complexity of the analysis. The heart of a research strategy is its set of restrictions. However, in the initial phase of research in the natural setting of the ongoing classroom activities, attempting to identify and evaluate explanatory variables, variables appear in a range from classification to continuous, intercorrelations between explanatory variables make their effects difficult to assess, complex interactions among factors cannot be assumed away, and application of Statistical tests of significance, under the assumption of random Sampling models is questionable. 23See Quinn McNemar, Psycholpgical Statistics, (New York: Jlahn Wiley & Sonar, Inc., 1969), pp. 191-213, for discussion of beta c<>efficients and beta weights. 20 Under these conditions, the quantitative criterion for judging the relative importance of explanatory variables is not based on the researcher asking about the direct effects of one variable at a time all else held constant, but is more meaningfully based on the researcher asking what it is that is most critical to know in order to reduce predictive error a maximum amount. According to Sonquist and Morgan, the question is, given the units of analysis under consideration what is the relative capability of predictor variables to give maximum improvement in the researcher's ability to predict values of the "dependent" variable under consideration.24 Research as an integral part of ongoing instructional setting activity has been considered on the basis of two models the researcher must take into account in his study of the relative importance of instructional setting variables, one pertaining to methodological design, the other to statistical analysis. However, under such conditions there is a third model which must be taken into considera- tion, and that is the "model" or set of assumptions underlying the on- going process in the instructional setting chosen for the researcher's study. This "model" is not the components of the physical setting. It is the set of assumptions which delineates the conditions under which the instructional process operates in that particular setting. While this delineation lacks the precision of the research design and statistical models, nonetheless this model is a set of assumptions 2"JohnA. Songuist, Multivariate Model Building, a paper presented at the Conference on Multivariate Models for Data Analysis in the Social Sciences, Institute for Advanced Studies, Vienna, Austria, September, 1969, pp. 1-37. 21 about learner characteristics, environmental characteristics, and instructional process expressed (however implicitly) by the course designers. It is a reflection of what is supposedly "important" as judged by the designers (on whatever basis they have chosen to make these judgments). These assumptions are reflected in the criteria (goals, objectives, and so on) specified for the particular instructional situation. The instructional process and the outcomes as reflected in the learners (the outcomes judged by whatever means designated by the designers) are supposedly reconciled to the theore- tical assumptions in the designers' model. Since the researcher is going to operate within this given instructional setting, his research must include some assessment of the degree of congruence between the model of "important" variables and those which actually appear to be important in the ongoing instructional situation. Neither those of the model nor those in the ongoing classroom Situation are necessarily synonymous with those having "importance" to the improvement of instruction. The methodological criticisms by researchers of Studies on teaching and learning in the classroom are mirrored by those looking to research findings for aid in improving instruction in classrooms, this time on grounds that results or implications are often difficult to generalize into the classroom setting and fail to lead to actual procedural efforts to improve instructional results, Hilgard has reflected this concern in his statement that "any isolation of basic science from applied science, when it persists, is unfortunate. Over the years advances in science have occurred in intimate relation with advances in technology. If we were to apply (these) historical 22 lessons......we would expect an equal intimacy between theory and research in the basic process of learning and the applied aspects of instruction and training in the schools."25 Commenting on what instructional setting factors provide optimal value in classroom instruction, Hilgard notes that there are many relationships of practical importance derived from theories of learning. These relationships are held in substantial agreement by theorists.26 For example, one suggestion derived from a stimulus-reSponse theory base is that the learner should be an active, rather than a passive listener or viewer, stimulus-reSponse theory emphasizing the Significance of the learner's reaponses. Frequency of repetition is another, in terms of acquiring skill and in terms of sufficient over- learning to ensure retention. The roles of generalization and discrimi- nation in stimulus-reaponse theory point to the importance of practice in various contexts so that learning is appropriate to a wider (or more restricted) range of stimuli. The stimulus-reSponse theory notion of the importance of meaningfulness in learning and retention is found also in cognitive theory expressed in terms of the fact that learning with understanding tends to greater transferability and permanency than does rote learning. Another suggestion from cognitive theories is that the direction from simple to complex learning is not from arbitrary meaningless parts to 25Ernesa R. Hilgard & Gordon H. Bower, Theories of Learning, (New York: Appleton-Century-Crofts, 1966), p. 541. 26Ib1d., p. 562. 23 meaningful wholes, but rather from simplified to more complex wholes. The c0gnitive theory concept of figure-ground relatedness suggests the importance of the perceptual features associated with the problem the learner is working on in terms of providing directional signs as to the interrelatedness of what leads to what. Another Suggestion derived from cognitive theories is the importance of divergent think- ing in leading to inventive solutions of problems and the role of convergent thinking in arriving at logically corréct answers. In terms of motivation to learn, the importance of goal setting by the learner, and success and failure as determiners of future goal Setting, find basis in cognitive theory. Theories of personality and motivation also provide generali- zations which Hilgard considers offer teachers suggestions of what factors are important in the classroom setting. For example, the need for affiliation in contrast with the need for achievement as motivating forces, based in motivation theory, suggests that the same classroom learning situation may tap appropriate motives for one learner and not for another. Studies of the role of anxiety in learning suggest that the anxiety level of the individual learner may determine the beneficial or detrimental effects of certain kinds of encouragement to learn. Such personality-related factors as degree of rigidity in behavior or tendency toward authoritarianism or extent of introversion suggest factors of the individual learner which are affected by the atmOSphere in the classroom, such as highly competitive conditions, or those requiring considerable participation in group discussion activities. Whether derived from the conditioning and reinforcement 24 approach to learning Such as the contiguous conditioning of Guthrie or the Operant conditioning of Skinner, or derived from theories that treat learning more broadly, tending to take into account the way the learner perceives the situation, and variously characterized by such labels as cognitive, Gestalt, or field theory, or derived from corollaries of learning such as McClelland's view of the role of achievement motivation, the examples in the preceding paragraphs indicate the form in which suggestions as to important factors in classroom instruction are derived. A review of the suggestions makes it apparent that "laboratory knowledge does not automatically lead to its own applications."27 It is understandable since the theorists' concern is with matters of uncertainty important in establishing a firmer foundation for their theories, and secondarily if at all in relation to matters of classroom instruction.28 There have been successful applications to such aids as programmed instruction and simulators but within a limited context. AS to the relative importance of the suggestions to classroom instruction, "When the practical condi- tions of learning have to be arranged for particular learners, there is general agreement that attention has to be paid to the nature of the learners, to careful analysis of the tasks that confront them; beyond that, there is rather general agreement on some broad generali- zations from learning experiments and theory, but these are not 27Ibid., p. 564. ZgA. Raymond Cellura, "The Application of Psychological Theory in Educational Settings: .An Overview," American.Educational Research Journal, vol. 6, no. 3, May, 1969, p. 373. 25 uniquely bound to particular theoretical viewpoints, and are not very instructive in reapect to specific problems of improving efficiency of learning."29 However, Ausubel feels that while an adequate theory of learning may not be a sufficient condition for the improvement of instruction it is a necessary condition for the identification of important variables and he has develOped a cognitive-Structure theory of school learning dealing exclusively with meaningful verbal learning. He suggests that attempts to extrapolate from learning theory and its empirical evidence directly to classroom instructional problems has disastrous consequences because judgments of those factors from learning theory with important value to classroom instruction require additional research structured so as to take into account practical problems such as interactions in the instructional setting with factors not implicit in the learning principles.30 Another approach to determining the qualitative value of instructional setting variables is to identify those providing Optimal value in classroom instruction by studying the instructional complex from a particular point of view. For example, the instructional complex can be viewed with an emphasis on the perSpective of the learner. Such a conception focuses upon making explicit the behavior(s) or the characteristic(s) Of learners as the most important determiners of the ongoing instructional setting. The interests, the abilities, the 29Hilgard and Bower, loc. cit., p. 573. 30David P. Ausubel, “A Cognitive-Structure Theory Of School Learning," Laurence Siegel (ed.), Instruction: Some Contemporary Viewpoints, (San Fransisco: Chandler Publishing Co., 1967), pp. 207-259 26 achievements, the personality characteristics of the learner are Specified as the important criteria defining the instructional complex. Such studies of potentially important learner variables are legion. For example, in a review of studies which have attempted to specify the relationship Of personality variables to academic perfor- mance, Cattell lists no fewer than twenty-six personality variables (ranging from extent Of freedom from neurotic orientation to degree Of impulsivity or degree Of endurance) used as potential associates of learner academic performance.31 On the other hand, the instructional setting can be viewed with an emphasis on environmental variables in the classroom setting. For example, one particularly important facet in the learner's environment is the instructor. Questioning the assumption that effective teaching necessarily produces effective learning focuses upon making explidt the characteristics and instruc- tional behaviors Of teachers as the most important determiners of the instructional setting. Getzels and Jackson have annotated over eight hundred studies attempting to Specify the relationship of factors, ranging from degree Of authoritarianism to cognitive style,to teacher effectiveness, and were led to the conclusion that "deSpite the critical importance of the problem and a half-centry of prodigious research effort, very little is known about the relationship between teacher characteristics and teaching effectiveness."32 31R. B. Cattell and J. H. Butcher, Prediction of Achieve- ment and Creativity, (New YOrk: Bobbs Merrill, 1967), pp. 106-107. 32J. W. Getzels and P. W. Jackson, "The Teacher's Personality and Characteristics," N. L. Gage (ed.), Handbook of Research on Teaching, (Chicago: Rand McNally & Co., 1963), pp. 506-581. 27 Such difficulties in identifying those qualities of potentially important variables which would be most useful in the improvement Of instruction naturally result in criticism by those looking to research findings for help in improving instruction.33 Bruner summarizes the extreme Of such feeling in his statement that, "one is struck by the absence of a theory of instruction as a guide to pedagogy--a prescriptive theory on how to proceed in order to achieve various results, a theory neutral with respect to ends but exhaustive with reapect to means. It is interesting that there is a lack Of an integrating theory in pedagogy, that in its place there is principally a body Of maxims."34 Such criticism is excessive from the standpoint that the tradition Of research has been to seek par- simony in explaining relationships among variables, basing the search on assumptions of normalcy and homogeneity. When attempts are made to extrapolate findings directly to classrooms, many circumstances with inherent violations of this normalcy and homoge- neity reduce the effectiveness of parsimony. However, such criticism may be deserved as suggested by Campbell and Hilgard of the failure to provide consistent research within the classroom setting to serve as an intervening process between parsimony and application. Difficulties in identifying those dimensions of potentially 33Instruction is defined as the arrangement of external condi- tions of learning in such ways to Optimally interact with the internal capabilities of the learner, to bring about a change in these latter capabilities. 34Jerome S. Bruner, Toward a Theory of Instruction, (Cambridge, Harvard University Press, 1967), p. 31. 28 important variables useful in the improvement of instruction parallels the previously discussed difficulty of defining importance in meaningful quantitative terms. In cases where Such assumptions as homogeneity, lack of interactions, and independence cannot be made, parsimOny may prove to be inadequate, and importance may be more meaningfully defined qualitatively in terms of location, time, and criterion, (each Of these qualitative functions having been defined with examples in the earlier discussion of the purpose Of the study). Overview The position has been developed in the previous discussion that in the majority of courses of instruction in formal educational settings the weakness persists of failing to compensate in instruc- tional procedures for the dimensions of any factors beyond the homogeneous descriptions Of predominant characteristics and Of failing to take into consideration the important heterogeneous descriptions and exceptions to general prediction which create instructional problems. Those looking for help in improving instruction in such settings direct their criticism at research on teaching and learning, for failure to provide satisfactory means of dealing with "the real world". One criticism is that the models of methodological design and statistical analysis used in basic research are rooted in a set of assumptions and a tradition of parsimony not tenable in the complex Structure of the classroom. The second criticism is that attempts to analyze the classroom take the form Of summative evaluation, indicat- ing the weaknesses in the instructional complex but failing to suggest procedures for actual course develOpment or to provide a basis for 29 research as an integral Operation within the instructional setting to act as an intervening process between results of "basic" research and their application to improvement of instruction. This study investigates such a formative analysis. Perhaps nowhere else at any formal instructional level do more class settings qualify as exemplars of failure to compensate in instructional proce- dures for dimensions of factors beyond homogeneous descriptions of predominant characteristics than at the level of undergraduate college instruction. Such an instructional situation was chosen for this Study The following chapter provides background for the study at two levels. The first section is a review of pertinent literature con- cerning problems Of research on teaching and learning discussed in chapter one. The second section outlines the university instructional situation used in the study, and background under which it Operated. Chapter three develOps a rationale upon which the choice of factors included in the study was based, describes the measures used as proxies for the factors, and outlines the study procedures includ- ing: subjects; study design; analysis strategy. Then the expecta- tions raised in chapter one about the nature of instruction are re- stated in terms of the instructional situation used for the study. In chapter four the data are summarized and diSplayed to illustrate the analysis strategy used in interpreting the data, and the data are discussed in terms of the concept of "importance". In chapter five the data are intergrated and interpreted in a discussion of the issues and questions raised in chapters one and three with regard to classroom instruction. CHAPTER II REVIEW OF THE LITERATURE AND BACKGROUND OF THE STUDY SETTING Review of the Literature Many undergraduate college courses serve as exemplars of the failure to compensate in instructional procedures for the dimensions of any factors beyond the homogeneous descriptions of predominant characteristics in the instructional situation. Increas- ing numbers of students and increasing heterogeneity in their charac- teristics have been accompanied by apparent decreasing interest in the quality of undergraduate college courses on the part of faculty, and the movement of many colleges to more business-like production- cost accounting models of instruction such as the large lecture section of hundreds of Students. However, recently a great deal Of attention has been directed toward the subject Of improving instruction at all levels of higher education, particularly the instruction of undergraduates. Sugges- tions for improvement Often appear to have an amorphous quality because the factors discussed as relevant to improvement are expressed as generalities, or global programs. Calvin B. T. Lee documents a series of articles by professors very much interested in improvement of instruction, the majority of articles dealing in broad generalities such as the role of values in the classroom, or the use of television to decrease class size, or the professor and his roles.35 Suggestions 35Dalvin B. T. Lee (ed.), Improving College Teaching, (Wash- ington, D. C.: American Council on Education, 1967), pp. 57-71, 347-69. 30 31 from students also reflect this tendency to generalize about factors they consider important in the instructional setting. For example, the Carnegie Commission on Higher Education reports that ninety-one per cent of seventy thousand students interviewed would like their course work to be more "relevant" and ninety-five per cent consider teacher "effectiveness" Should be the primary criterion in faculty promotion.36 These examples from groups at two extremes are not uncommon. Although some overgeneralities might be assumed away as showing lack Of careful thought or lack of skill in communication, the number of such examples in the literature suggests that this lack of Specificity as to what dimensions Of the factors in the instructional setting are Of most relative importance in the improvement of instruction is partially a reflection of the criticisms of research on teaching and learning discussed in chapter one. Gage has documented the hundreds Of Studies in teaching and learning at all instructional levels.37 Since the setting for this study is at the college level, the review of the literature is confined to studies of instructional setting variables at that level. The studies chosen are examples of a number of studies reflecting the problems of methodology and application to ongoing instructional settings discussed in chapter one. The criticisms by researchers of methodology related to such problems as failure in Specificity of criterion, in maintaining 362Carnegie Commission on Higher Education, national survey of student Opinion, excerpted by New York Times, January 16, 1971. 37N. L. Gage, Handbook of Research on Teaching, (Chicago: Rand McNally & Company, 1963). 32 homogeneity and independence, in analyzing interactions, or stOpping to consider the kinds of questions that can be answered by the research procedures used,are reflected in the series of studies of the teacher as leader in the educational setting. These studies largely grew Out Of the conceptual framework Of Kurt Lewin, parti- cularly apparent in terms of climate-style research, and culminating in the research of Lippitt and White on social atmOSphere. Their series Of studies utilizing artificially-induced atmOSpheres (authoritarian, democratic, and/or laissez-faire) reported results on the tendency Of the same group Of people to behave in markedly different ways when operating under leaders who behave differently. In addition to the innovation of artificially-induced atmOSpheres in experimental design, the studies are illustrative of social interac- tion as a statistical analysis problem. These studies also illustrate a defect in later studies in which teachers manifest assigned patterns of behavior. It appears there is confusing of assumptions about what leadership "Ought to be" with research-oriented questions of "what produces what". Authoritarianism was equated with behaviors which could be classed as unfriendly or threatening in an a-priori fashion.38 The relevancy of these previous experiments is eSpecially pertinent to the series of studies that define leadership as points on an hypothetical continuum. Whether the hypothetical continuum ran 38As examples of the studies of Lewin, Lippitt, and White, see K. Lewin and R. Lippitt, "An Experimental Approach to the Study Of Autocracy and Democracy: A Preliminary NOte," Sociometry, 1938, l, PP. 292-300, and K. Lewin, R. Lippitt, and R. K. White, "Patterns of Aggressive Behavior In Experimentally Created Social Climates," Journal of Social Psychology, 1939, 10, pp. 271-299. 33 from authoritarianism to democratic, or student-centered to instructor- centered, or from directive to non-directive, the experimental differentiation between treatments was similar -- either/or. Two studies using this concept of the hypothetical continuum, one by Faw, another by Asch, reflect the same problems as those cited about the previous studies. The nature Of the two studies was similar, involving Similar group sizes and procedures, the general purpose being to evaluate the overall effectiveness of non-directive teach- ing Of an undergraduate course in general psychology as compared with the traditional lecture-discussion method. In each study, an attempt was made to more precisely deliuate the theoretical structure of the leadership role in terms of academic outcomes. In the first study, mean examination scores for students in the student-centered group were significantly higher, although the Students in the student- centered group had doubts about the amount of information and help they received under this method. The second study findings in terms Of academic achievement were in reverse to those of the Faw study. Those in the instructor-centered directive group did Significantly better on the final examination, but the students in the Student- centered group felt their class situation had been more helpful to them in learning subject matter for the examination than did those in the instructor-centered group.39 Ambiguity in the study findings is paralleled by findings of no significant differences. A study by Eglash was similar in form, 39V} Faw, “A Psychotherapeutic Method of Teaching Psychology," The American Psychologist, vol. 4, g, 1949, pp. 104-109, and M. J. Asch, "Non-directive teaching in PsychologY: an Experimental Study," Ps cho- lggical Mongggaphs: General and Applied, vol. 65, A; 1951, PP. 1-2 . 34 conceptual framework, treatment and subjects to those mentioned previously. However, there were no Significant differences between groups on examination performance, and there were no differences on attitudinal measures in general. There was a significant difference involving a "Quantitative Teacher Evaluation Form". The lecture class significantly felt that the course objectives were better met and the instructor's presentation did enhance learning. These results would indicate that there is a difference between actual achievement and what one's feeling Of achievement is, thus indicating that students' feelings Of satisfaction are a function of style Of presentation as well as the instructor's personality and course content, and suggests an interesting interaction.40 Another study also suggest- ing an interaction, this time between student ability and method of instruction, involved a similar design. However, in addition, subjects were subdivided according to measures of ability. Although there were no significant differences between the two groups, as a whole, there were Significant differences among the Sub-groups as to recall, recognition, and more understanding. The group method resulted in more understanding in learning, longer retention, and greater expression of individual differences. The lecture-demonstration method resulted in greater expression of individual differences, and longer retained, more understanding-type learning by the lower group.41 493. Eglash, “A group Discussion Method of Teaching Psychology," Journal of Educational Psychology, 45, 2, PP. 257-267. 41J. ward, "Group-study Versus Lecture-demonstration Method in Physical Science Instruction for General Education College Students," Journal of Egperimental Education,_1956,‘24, pp. 197-210. 35 Other studies have also attempted to add a further dimen- sion to the role Of the leader in terms of group behavior. For example, three separate studies over a ten year period (Deutsch, Smith, and Haines and McKeachie) each investigated competition and cOOperation as factors on group cohesiveness and achievement perfor- mance. Variables associated with develOping from group growth became apparent under conditions where members of the group were urged to cooperate, but no significant differences appear between cOOperative and competitive treatments in regard to achievement performance. This Suggests that perhaps certain results logically lend themselves to COOperative group structure, whereas other results are more easily interpreted in terms of a competitive classroom atmOSphere. It would appear that this could be logically interpreted as the natural outcome of the research designs. Situationswhere individuals are urged to cOOperate would be expected to produce stronger group orientation than conditionaunder which competition is urged. Lack Of significant difference of achievement between groups could be inter- preted as the natural Outcome, since in essence the cOOperative setting results in the effect Of compressed variance, whereas in the competi- tive situation the effect of normal variance due to individual dif- ferences in learning is Obtained. The results Of this experimental phenomena in terms of the analysis Of variance model is the accumu- lation of an abnormally large error term nullifying any potential results.42 42 M. Deutsch, "An Experimental Study of the Effects Of COOp- eration and Competition Upon Group Process, Human Relations, 1949, 2, pp. 199-231, A. J. Smith, "Productivity and Recall In cOOperative and 36 The studies cited in this discussion were chosen as examples of work Over a time Span of 30 years43 to illustrate how sequentially over time attempts have been made to increase the precision of the Lewin model both in terms of theory and methodology, and to achieve more Specific results, and that the degree of increase in stringent empiricism or more Specific theoretical structures is difficult to assess. These studies Of factors in group behavior, particularly in relation to the role of the leader, have been discussed at some length for two reasons. First, they are representative of groupings Of studies that have occurred around many factors believed to be of some importance in classroom instruction, and representative of the methodological problems in such Studies. Second, the conceptual framework of the role of leader and group behavior is particularly germane to undergraduate college instruction where many courses, particularly in the students' first years, involve an attempt to employ "small group discussion" in the form of a combination Of huge- lecture-small-discussion-group within a given course starring teaching assistants in the role Of "leader", nine times out Of ten. The methodological problems associated with the statistical analysis of data involving many variables, many subjects, presence Of data in nominal form, and so on, can be illustrated by the example Of Competitive Discussion Groups? Journal of Psycholggy, 1957, 41, pp. 193- 204, H. Haines and W. McKeachie, "cOOperative Versus Competitive Dis- cussion Methods In Teaching Introductory PSychOlogy," Educational Psychology, 1967, 58, QJ pp. 386-340. 4'3J. Kerrick, "Lecture Versus Participation in the Health Train- ing Of PeaaaCorps Volunteers," Journal Of Educational Psychology, 1967, 58, 5, pp. 259-265, as a recent study in the same vein. 37 the number of studies, over time, attempting to identify and assess the relative importance of factors believed to be important in predicting college success of a student. Astin attempted to identify those factors associated with Students who drop out of college, in 44 a four year longitudinal study of 6660 Students, while Sgan attempted to identify the importance Of the Scholastic Aptitude Test in rank 45 in graduating college class, while Barger and Hall attempted to establish the relative importance of personality patterns on the Minnesota Multiphasic Inventory zto achievement in college.46 In a review Of such studies, HOpfenSpirger concludes that, "in Spite of the many studies which have been made in order to find accurate predictions Of college success, little progress toward improved predi- ction has been achieved. Current studies reflect findings similar in level and precision to those made thirty years ago. DeSpite inno- vations in statistical techniques it is unusual to find a correlation between college grades and other measures above the level of 0.60, and most correlations reported fall in the range of .45 to .55.47 The 44A1exander W;.Astin, "Personal and Environmental Factors Associated with College DrOpouts.Among High Aptitude Students," Journal of Educational Psycholpgy, 52d April, 1964, PP. 219-224. 45Matthew'R. Sgan, "An.Alternative Approach to Scholastic Aptitude Tests as Predictors of Graduation Rank at Selective Colleges," Educational and PSychOlogical Measurement, 24, February, 1964, PP. 347-50. 46Ben Barger and Everette Hall, "Personality Patterns and Achievement in College," Educational and Psychological Measurement, 23, February, 1964, pp. 339-341. 47P. HOpfenSpirger, "An Investigation of three statistical Techniques and Their Applicability in Prediction of College Success," (unpublished Master's thesis, George washington University, 1967), p. 4. 38 predicting methodology has been, traditionally, linear regression using as few independent variables as possible to produce an estimate of the dependent variable, with the evaluation of precision typically being the amount of variation in the dependent variables "explained" by the independent variables. HOpfenSpirger concluded that the statistical technique Of regression analysis was capable of produc- ing useful predictions, providing the data could be adjusted to meet the criteria of the regression model without excessive trans- formations which themselves could well affect results. However, under rigid assumptions of the linear models Sufficient trans- formations made results dubious.48 Criticisms by those trying to use research findings to im- prove instruction are reflected in research difficulties of identifying and assessing the relative importance Of instructional setting factors. However, Hilgard's contention that learning theories can provide only global structures for instructional improvement is challenged by Ausubel on grounds that direct extrapolations fail but suggestions related to teaching course materials are best derived from learning theory.49 According to Ausubel, cognitive learning theorists maintain that meaningful verbal learning is the "human mechanism best suited for acquiring and storing the vast quantity of ideas and information represented in any body Of knowledge."50 Whereas rote learning and retention are influenced primarily by the 48HOpfenSpirger, loc. cit., p. 31. 49Hilgard, loc. cit., pp. 571-572. SOAusubel, 10c. 313., p. 219. 39 interfering effects of similar rote materials learned immediately before or after a learning task, meaningful learning and retention are influenced primarily by the properties of the relevant subsuming ideas in cognitive structure with which they interact.51 In his theory, new meanings are acquired when potentially meaningful Symbols, concepts, and so on are related to and incorporated within the cognitive structure on a nonarbitrary, Substantive basis. He has attempted tO incorporate Such learning theory propositions as cueing, overlearning, retroactive inhibition, and so on, into a framework applicable to classroom experimentation. For example, in a study on the learning of college text-type materials, Ausubel, Robbins, and Black demonstrated that memory of a passage on Buddhism was not affected adversely by later study of a passage on Christi- anity, and in fact when passages repeated points from the earlier material, memory was actually facilitated, if the passages were learned in accordance with his structure of "meaningfulness".52 Skinner's applications of principles of operant condition- ing to programmed instructions, and of Lumsdaine's, who attempted to show that Guthrie's views more accurately account for what is done in programmed learning, tend to refute criticisms Of those who question the value Of learning theory in instructional improvement.53 51Meaningful verbal learning is simply that kind Of learning which takes place when potentially meaningful verbal material is substantially related to or subsummed under existing knowledge already held by an individual, and done in a nonarbitrary way. This pre- supposes a meaningful learning set on the part of the learner. 52D.Ausubel, P. Robbins, C. Lillian, and E. Blake, "Retro- active Inhibition and Facilitation in the Learning of School Materi- als, Journal of Educational Psychology, l957,,fl§, PP. 334-343. 40 The question becomes, however, what Should be the relative importance attached to such factors within the context of the classroom. Ausu- bel has confined his structures principally to less complex verbal learning rather than problem solving tasks, and the differential effects of learning set are still Open to question. In the case of the relative importance Of programmed instruction represented via teaching machines and workbooks, they may have many theoretical advantages over lecturing or other instructional methods of a conventional nature; however, little experimentation has been done at the college level to determine their usefulness within the interactive nature Of factons affecting college student behavior.53 Another approach reflected in the literature to determine the qualitative value Of instructional setting variables is to identify those providing Optimal value in classroom instruction by studying the instructional complex from a particular point Of view, such as focusing on characteristics of the learner or on certain environmental characteristics (including the teacher) which affect the learner. However, at the college level, Nevitt Sanford points out that until very recently the college Student has been "a familiar hero of psychological literature---intelligent, cheerful, and above all available, long a favorite subject of experiments designed to es- tablish general truths about human behavior. But he has not Often been investigated as a student; in other words, it is not often that 53w. J. McKeachie, "Research on Teaching at the College Level," in N. L. Gage, (ed.), Handbook of Research on Teaching, (Chicago: Rand McNally and Company, 1963), p. 1155. 41 psychologists or other social scientists have studied his behavior as a reaponse to the demands Of the role of student."54 But over time a number of studies have indicated student traits, abilities, attitudes, and so on, which appear important to student behavior in the classroom. Past performance as a factor has been investigated both in terms of differences among students in mental abilities and differ- ences in past academic performance. A series of studies by Cronbach, Henry, Dyer, Fishman, and Whitla indicate test Scores on college "entrance" examinations such as college boards correlate .35 to .55 with college grade-point average.55 Large-Scale studies of the general intelligence Of undergraduate students (Educational Testing Service, Fosmire56) infer that the average intelligence test scores of major groups of students regularly fall into an order in terms of major field of interest chosen for Study, although wolfle indicates that the interpretation of these findings can be misleading in that such 54Nevitt Sanford (ed.), The American College, (New York: John Wiley and Sons, Inc., 1962), p. 531. 558ee a series Of studies: Lee J. Cronbach, Essentials of Psycholpgical Testing, (New York: Harper & Bros., 1949), Ervin R. Henry, "Predicting Success in College and University," Handbook of Applied Psychology, Douglas H. Fryer et al (eds.) (New York: Rine- hart & Co., 1950), PP. 449-453, Henry S. Dyer, College Board Scores, (New York: Colllege Entrance Examination Board, 1955), J. A. Fishman, "1957 Supplement to College Board Scores, No. 2", (New York: College Examination Board, 1957), Dean K. Whitla, Handbook of Measurement and AssessmenE in Behavioral Sciences, (Reading, Mass: Addison-wesley, 1968). S6See Educational Testing Service, Annual Report, 1951-1952, F. R. Fosmire, "Generality of Some Academic Reputations, Science, 1956, 124, PP. 680-681. 42 fields as business or education, looking poor in terms of mean intel- ligence scores or overall score distribution, have about the same share of the tOp students as do the more academic fields.57 The role Of grade-point average in determination of current performance is complicated by the fact that GPA is itself a complex indice, involv- ing such facets as student motivation and teacher grading practices.58 Achievement motivation as a factor in academic performance has been investigated both by questionnaire and projective technique. Studies involving the Edwards Personal Preference Schedule indicate a correlation between achievement motivation and academic perfor- mance;59 the studies involving the Thematic Apperception Test report 60 61 inconsistent findings, those Of McClelland and of Weiss found a positive correlation, those by Haber62 and by Farquhar and Krumboltz63 S7D. WOlfle, America's Resources of Specialized Talent, (New York: Harper Bros., 1954). 58David E. Lavin, The Prediction of Academic Performance, (New York: Russell Sage Foundation, 1965). 59Dean K. Whitla, "Research in College Admission," in Educational Evaluation: New Roles, New Means, Ralph w. Tyler (ed.), Chicago: National Society for the Study of Education, 1969), p. 90. 60David C. McClelland, "Issues in the Identification of Tal- ent," in Talent and Sociepy, David C. McClelland et al. (eds.), Princeton: Von Nostrand Co., 1958). 61Peter weiss, "Achievement Motivation, Academic Aptitude, and College Grades," Educational and PSychOlogical Measurement, 22, (Winter, 1959), pp. 663-666. 62Ralph N. Haber, "The Prediction of Achievement Behavior by an Interaction of Achievement Motivation and Achievement Stress," Dissertation Abstracts, Vol. 17, 1957, pp. 2686-2687. 63William W. Farquhar and John D. Krumboltz, "Reliability and Validity of the N-achievement Test," Journal of Consulting PSychOlogy, El, (1957), pp. 226-228. 43 found no correlation. Motivation by fear and motivation by hope were investigated by Atkinson and Litwin, who found male college students high in test anxiety were first to complete exams and did poorer work than during the course, as opposed to those students with positive motivation to succeed.65 Studies using general measures of anxiety not related specifically to test-taking have not shown significant relations to measures of academic performance.66 Degree of interest in the subject being taught as a factor in academic performance is not clearly understood. Cronbach found such measures to have a low correlation with grades alone.67 Studies of interest-test scores in relation to Specific occupations are unclear. In studies Of medical and engineering students, interest- test scores were unrelated to overall performance in that curriculum when ability was controlled, although related to some Specific course aSpectS.68 Differences in attitudes among undergraduates has been viewed largely in terms of positions pppularly regarded as "liberal" or "conservative". With some consistency, the most conservative groups are in applied rather than academic fields. Students in education 65J. W..Atkinson and G. H. Litwin, “Achievement‘Motive and Test Anxiety Conceived as Motive to Approach Success and‘Motive to Avoid Failure," Journal Of Abnormal Psychology, 99, 1960, PP. 52-63. 66Dean K. Whitla, in Ralph W. Tyler, loc. cit., p. 90. 67Cronbach, loc. cit., pp. 108-115. 68V. H. Hewer, "Vocational Interest-Achievement-Ability Interrelationships at the College Level," Journal of CounselingyPSych- 9.1.931. &. (1957). pp. 234-238. 44 are difficult to classify. Those in secondary education reflect the attitudes of their prOSpective teaching fields, and those in elem- entary and physical education tend to be among the most conservative groups.69 The studies do not differentiate those with internalized attitudes from those whose attitudes are largely external "trappings" nor is it clearly Specified as to the relationship of Such attitudes in classroom interaction. The evidence indicates that major changes in attitudes or values of college Students do not take place in 70 Jencks and the short Space of time within a given course setting. Riesman contend that the docility registered in the majority of classes by students is simply a confirmation of the student attitude of the educator's irrelevance to contemporary culture.71 Authoritarianism as reflected in held beliefs has been investigated in a series Of studies by Stern, Stein, Bloom, and Pace. Students high on a variable akin to authoritarianism (rigid, uncritical of held beliefs, as two examples) were found to gain more when taught in an homogeneous group and to prefer lecture conditions.72 More 69Carl Bereiter and Mervin Freedman, "Fields of Study and the People in Them," in The American Collegs, Nevitt Sanford, (ed.), (New York: John Wiley and Sons, Inc., 1962), p. 568. 70Paul L. Dressel, "The Role Of Evaluation in Teaching and Learning," Evaluation in the Social Studies, Harry D. Berg (ed.), (washington D. C.: National Council for the Social Studies, 1965), p.5. 71C. Jencks and D. Riesman, The Academic Revolution, (New York: Doubleday, 1968), pp. 163-192. 72G. G. Stern, "Environments for Learning," in The American College, N. Sanford (ed.), (New York: John Wiley and Sons, Inc., 1962), pp. 690-730. .5” x;- 45 complicated are the findings of Koenig and McKeachie that flexible nonauthoritarian women participate more in small groups than do rigid women, but the relationship is reversed for men.73 Extent Of student independence as a factor in classroom instruction has been investigated by WiSpe. Students differenti- ated as "insecure" on TAT-like measures had unfavorable attitudes toward permissive teaching situations; those differentiated as "satisfied" had favorable attitudes toward instructors and fellow students and either permissive or directive teaching methods; "independent" students had moderately favorable attitudes toward fellow students and the instructors, but wanted more permissive teaching, regardless of method used.74 The studies cited in relation to learner characteristics represent examples of the foci of interest in such research. Over the past ten years reviewers in compendia of these studies, attempt- ing to put them in perSpective in terms of improving instruction, have noted similar problems: in many studies the differences among groups are very small and tentative; among similar studies, inconsis- 75 tent findings. Similar results are reflected in the studies with primary 73K. Koenig, and W. J. McKeachie, "Personality and Independ- ent Study," Journal of Educational Psychology, 1959, 59, pp. 132-134. 7('1. G. WiSpé; "Evaluating Section Teaching Methods in the Introductory Course," Journal of Educational Research, 1951, 45, 75See Sanford (1962) pp. gi£., Gage (1963) pp. cit., Lavin, (1965) pp. cit., Whitla (1968) pp. cit. ‘A H. 'I. 46 focus on teaching method. Studies of size Of lecture class or size of discussion class in relation to exam performance, or instruc- tor-centered versus student-centered teaching approaches in relation to exam performance and attitudes often reflect inconsistent findings among studies and nonsignificant differences due to the fact that Optimal factors for some students are detrimental to the achievement of Others. Some analyses of teaching methods take such individual differences into account. For example, Siegel's findings suggest a tendency for high ability students to gain more in course-related attitudes in small rather than large sections;76 Calvin, Hoffman and Harden found less intelligent students did better in problem-solving Situations conducted in an authoritarian manner rather than a permissive manner.77 Use Of results Of studies focusing on teacher characteristics have also encountered difficulties. For example, a number of Scales have been develOped to investigate teacher attitudes. Scores tend to be a function of grade level Of teaching and Of major field of interest. However, studies in relation to the role of these attitudes in the classroom tend to assume teaching involves a unitary attitude, and pool the data Obtained from such Scales. In addition, findings suggest that teacher attitudes vary in validity in terms of teacher effectiveness in the classroom according to the values held by students 76L. Siegel, "Students' Thoughts During Class: A Criter- ion for Educational Research," Journal Of Educational PSychOlogy, 1963, 23, pp. 45-51. 77A. D. Calvin, F. K. Hoffman, and E. L. Harden, "The Effect of Intelligence and Social AtmOSphere on Group Problem-solving Behav- ior," Journal of Social Psychology, 1957, 52, PP. 61-74. 47 who are interacting with the teacher.78 Similar results are found in studies related to values held by teachers. Findings Suggest that in at least two of the value areas measured by Such inventories (economic and social) teachers might, as a group, be distinguished from the general population. Even more important to the understand- ing of teacher performance are the differences between various teach- ing Specialties. However, the usefulness of such findings, in isola- tion, in terms of teacher performance in the classroom has not been demonstrated.79 Another focus of studies of teacher characteristics is that area referred to in the literature as "personality" -- factors such as degree Of introversion or extent of femininity. Such studies seem to fall into two major categories: studies attempting to dis- criminate between teachers and nonteachers with scales such as those 80 and studies Of the Minnesota Multiphasic Personality Inventory; attempting to assess the relationship of Scores on such scales to measures of teacher "effectiveness" in the classroom.81 A question raised by many of these studies is whether the same teacher, no matter 78Getzels and Jackson, pp. cit., pp. 515-523. 79R. D. Mann, Interpersonal Styles and Group Development, (New YorkL Wiley, 1967), pp. 89-116. 8oAllen L. Edwards, The Measurement of Personality Traits, New York: Holt, Rinehart and Winston, Inc., 1970), pp. 167-190. 81Getzels and Jackson, pp, cit., pp. 534-551. 48 what his personality, will affect all Students the same way. The assumption is made that there is one ideal teacher equally effective with all students. Studies of how intellectual ability contributes to teacher performance in the classroom have had limited success. Findings from the numerous studies in which intelligence tests have been adminis- tered to teachers in an attempt to assess its relation to performance are inconsistent. The role played by different types of cognitive functioning such as divergent thinking and by possible attitudinal correlates Of such abilities in the performance of the teacher has not been explored in interaction with Students in the classroom.82 The previous discussion Of findings in the literature serves as an example of the basis upon which those looking for help to improve classroom instruction make charges that the results or implications of studies are Often difficult to generalize into the classroom setting and fail to lead to actual procedural efforts to improve instructional results. However, it is evident that from whatever particular point of view the researcher has chosen to view the instructional complex (however narrowly defined), the difficulty in many cases is not inadequate conception or develOpment in the study but that the results taken in isolation explain a very small portion of behavior when viewed within the entire instructional complex. Accordingly, there have been some attempts to define conceptual frameworks for instructional research which are based on 82Philip W..Jackson, Life in Classrooms, (New York: Holt, Rinehart and Winston, Inc., 1968), 146-155. 49 the assumption that factors comprising the instructional complex interact. An example Of a study based upon such a framework is that Of Siegel and Siegel, who have conceptualized what they refer to as the instructional gestalt -- the educational process as viewed in terms of broad frameworks which "give appropriate recognition to the variety Of instructional settings, teaching procedures, simultaneously exposed learners, and multiple criteria of effectiveness without sacrificing either the essential flavor of the instructional process or Specificity of its conditions."83 The Siegels' particular framework focused upon four clusters of independent variables, identified as instructor, learner, course, and classroom environment clusters. Specific variables within each of these four clusters, chosen on the basis of previous research findings, were Specified as the most "potent determinants of what transpires within the instructional gestalt."84 Alternative criteria for assessing instructional effectiveness were Specified, as was the expectation that components of the gestalt probably would interact differently for the different criteria. Under these condi- tions there would be no one best instructional method. The best method would be a function of the educational Objectives, the cir- cumstances, and the participants. Interactions would occur between the variables within a given cluster and between variables across the §3Laurence Siegel and Lila Corkland Siegel, "A Multivariate Paradigm for Educational Research," Psychological Bulletin, 1967, ss, 5, pp. 306-326. 84Laurence Siegel and Lila C. Siegel, The Instructional Gestalt in Televised University Courses, Number 609, United States Department of Health, Education, and Welfare, Office of Education, p. 27 50 the four clusters. For example, the variable academic ability, from the learner cluster, supposedly would be in interaction with selected environmental, instructor, and course variables from the remaining clusters. Whereas specific aSpects of the instructional complex may Stimulate bright students, the very same elements may threaten or discourage students of less ability. Thus the model must focus on the interactive relationships between variables comprising the instructional complex, as well as upon the summary or main effects. The assumption underlying this framework is that effects of various kinds of instruction within a given course could be conceptualized and empirically studied in relation to variation in learning environments, learner characteristics, and the relevant activities of the instructor, with the hope of discovering combinations of learner, instructor, environmental, and course variables Optimizing desired educational outcomes. In making the transition from the generalized paradigm to the application of it in research in the classroom the difficulties encountered in the reported study raise again the problem discussed by Campbell and Stanley and Box in chapter one, the reconciliation of models, In the factorial design with analysis of variance the multivariate paradigm was of necessity simplified. Nine independent variables were used: two from the environmental, two from the ins- tructoral,and five from the learner cluster. The three variables from the course cluster were eliminated from the analyses by developing separate analysis matrices for each course, preventing accumulation of data pertaining to the interactions between these variables and the (I) *7] 51 others explored. The learner variables though continuously distri- buted were dichotomized, excluding the middle forty per cent of each learner distribution. The complex analysis of variance model does identify when significant interaction is present, but unless the interactions are clearly hypothesized at the Start, this approach does not provide a basis for empirical interpretation of the nature of these interactions. Rather, significance observed at one of the levels of interaction serves only to demonstrate that main effects are confounded. However, the design did serve as the investigators suggest, first, as a means for sensing interactive conditions probably present in the instructional setting, and second, if interactions are not present, to add strength to the generalization Of main effects. Background Of Study Setting The background of the setting for this study mirrored the problems discussed in relation to the Studies reviewed in the literature -- namely, classes Of undergraduate Students with increas- ing numbers and increasing heterogeneity in characteristics, diffi- culty in assessment of the relative importance Of variables in the instructional setting in a manner functional to improving instruc- tional conditions, and difficulty in utilization of findings from research studies,for any combination of reasons examined by Campbell, Stanley, and Box in chapter one. The setting of the study was the first course in the undergraduate teacher-preparation sequence, which can be categorized as a lecture-discussion group, multi-section, survey-type course. The instructional content of the course included, secondarily, an introduction to the teaching profession and to the program Of teacher 52 education, and primarily an introduction to the subject matter in the area of educational pSychology. Anywhere from seven hundred to one thousand students were enrolled in the course in any given term. The course itself was part of the curriculum offered in the College of Education at Michigan State University. Trow has outlined the predominant characteristics of large state-Supported institutions (in which category this university falls) as being: first, a relatively high student to faculty ratio; second, a research- oriented faculty with a possibly genuine but limited interest in undergraduate teaching; third, a Student body Of great heterogeneous academic ability and motivation.85 The principle method of compensating for these character- istics has been to use senior faculty primarily in graduate instruc- tion and research and in highly Specialized upper-level undergraduate areas, while limiting the contribution of such faculty in the early educational phase of undergraduates to large lecture settings. The instructional gap thus created is typically filled by the teaching- assistant (usually a graduate student enroute to his doctorate) in a classroom structure Of either the lecture-discussion section approach or the multi-section, independent-instructor approach with common curricular examinations.86 The expected consequences of such instructional situations, if large numbers of students are enrolled and if there is failure to compensate for the heterogeneous 85Martin Trow, "The Undergraduate Dilemma in Large State Universities," Universities Quarterly, December, 1966, pp. 17-18. 86Ibid., p. 30. 53 student pOpulation, would be: "the good students becoming lost, bored, and remaining unchallenged; the middle students -- and all Of the stu- dents to some degree -- disassociating the material (theory and research) from their lives and their futures and from implications other than related to the disci- pline; the poor Students becoming totally uninvolved because they cannot compete with the good students, and because within this setting, they cannot compensate in any way for the entering differential."87 Sources of suggestions for improvement of undergraduate instruction at Michigan State University parallel those sources for undergraduate education in general, mentioned in the introduction to the review of the literature in the previous section of this chapter. The generalized Student comments on "relevant" education in the Carnegie Commission report are mirrored by comments in the editorials Of the campus newSpaper which in a discussion of the Uni- versity's commitment to undergraduate education, referred to "under- grads getting ‘inferior' education from graduate assistants."88 The University Committee on Undergraduate Education based its recom- mendations for improvement of undergraduate education on the following stated principles: interaction between teacher and student as the primary consideration in implementing programs; programs would be judged on learning and scholarship; quality of instruction and the teacher as a model were to be the main concern of both individual faculty members and undergraduate programs; liberal education in terms of human develOpment was to be enlarged as the foundation of 87Ibid., p. 35. 88The State News, University campus neWSpaper, May, 1969. 54 the institution's approach; educational Opportunity was to be a part of instruction so that entering characteristics of Students would not be the main determiner of degree of instructional impact.89 The College of Education, of which the course used in this study was a part, also provided suggestions for improvement of undergraduate education. The report of the College's Curriculum Review Committee Stated, Providing quality programs to large numbers of students is one Of the most challenging tasks faced by large universities.....In order to give undergraduate instruction the emphasis which we believe it deserves, the College of Teacher Education has designed a plan which will bring our undergraduate Students in con- tact with outstanding senior faculty members and at the same time allow for student identi- fication with a small group where his indivi- dual questions and comments can become mean- ingful. The actual physical representation of this plan to provide a'huality program to give undergraduate instruction the emphasis....it deserves? in terms of the course used in this study, was a lecture-discussion section format involving a number Of composite parts: one course coordinator and evaluator, one teaching internship advisor for the teaching assistants, eight faculty lec- turers, sixteen graduate teaching assistants, and seven hundred to one thousand-plus Students (mainly SOphomores, but some juniors and 89Committee on Undergraduate Education,_lpproving Undergrad. Education: The Report of the Committee, (East Lansing, Michigan: NUohigan State University, 1967). 90Curriculum Review Committee, "Organizational Plan For Foundation Courses," Repprt on Undergraduate Curriculum Revision, mameograph, (East Lansing: College of Education, Michigan State University, April, 1965), p. 1. 55 a few seniors), plus two different hours during the day at which times the format was offered to the students, using the same instructional personnel. At whichever time chosen to take the course the student met on alternate days with other Students for one large lecture section or in his discussion section of approxi- mately twenty-five to thirty students. Briefly, the role of the course coordinator was to "be the person reSponsible for bringiqgunity to the course....(and to) also be reSponsible for the coordination of the discussion sectionsJ' The role Of the teaching internship advisor for the teaching assis- tants was to coordinate a content seminar for the teaching assistants and conduct a seminar in the methods and development of college teaching, in which the assistants were required to participate. The role of the course evaluator (in this instance also the course co- ordinator) was to construct examinations because ”Examinations in a course enrolling large numbers of Students are doubly important. The construction of evaluation instruments which will prOperly mea- sure the progress students make toward the predetermined goals of the course is a most important function."91 The role of the senior faculty can best be explained by the fact that the content of the course was to center around the main interests of the faculty, that is, their areas of Specialization, in this instance, primarily areas within educational psychology. Senior faculty members who have a particular compe- tence in some portion of the course content and who 911bid., p. 2. 56 are known to be Outstanding lecturers will be invited to give lectures on that portion of the course which falls within their particular area of interest and competence. Each senior faculty mem- ber thus will be involved with only from three to six lectures covering a particular unit Of the course. Other times he will be free to pursue other research or instructional reSponsibilitieS.. ...we expect that it will become much easier to employ competent new Staff whose reSponsibility it will be to only make Specialized contributions to the instruction in the undergraduate basic courses.92 The role of the graduate teaching assistant under this format became the following: The students will meet in small (under thirty-five) discussion sections which will be conducted by graduate assistants (TA'S) employed for this purpose. During these sessions it is expected that the student will have an opportunity to have his questions clarified and be identified as an individual within the total group. It is expec- ted that the discussion sections will be related closely to the lectures and that here the student will be concerned with the application to class- room problems Of Sge concepts covered during the lecture sessions. In chapter one Campbell advocated the implementation of research programs within already established instructional set- tings, on the premise that Over time via replication a body of data would be develOped to be used as the basis for additional testing of hypotheses about the nature of instructional variables. (The extension of this, Of course, was that ideally cross-validation would be occurring by other such programs in similar situations.) Under such a program the initial purpose would be to discover the 92Ibid. 93Ibid. D9 A the 57 Structure of relations among variables in a Specified instructional setting, and under what conditions and through what intervening processes this relationship occurs. Under such research conditions not only would it be necessary to take into consideration the reconciliation of research and Statistical models as Stressed by Stanley and Box, but to take into consideration a less precisely delineated "model". This latter model is not the components of the physical setting but the set of assumptions about learner charac- teristics, environmental characteristics, and instructional process expressed (however implicitly) by the course designers and is a reflection of what they supposedly judge to be "important" factors. These assumptions are reflected in the criteria (goals, objectives, and so on) Specified for the particular instructional situation. The instructional process and the Outcomes as reflected in the learners during and by the end of the course are Supposedly recon- ciled to the theoretical assumptions in the designers' model. In the course used for this study the "implied" model, composed of a set of assumptions about the Structure of relations among variables in the instructional setting,was reflected in the Review Committee's plan to implement the goals of the University and the College of Education to improve undergraduate education. The overall goal was to "provide quality programs for large numbers of students.....in order to give undergraduate instruction the empha- sis....it deserves." Quality programs themselves would appear to be strategies -- an organization of planned interventions, sequences, and sets of standards to be used in the develOpment of the Students. Any given course within these programs would implicitly be providing 58 a level Of instruction which would guarantee the expected "quality". So, within the course setting used in this study these same goals would be reflected.94 One such goal provided that educational Opportunity was to be a part Of instruction so that entering characteristics of Students would not be the main determiner Of degree Of instructional impact. This implies recognition of the heterogeneity of students. However, the role of senior faculty was confined to mass lecture sections. Therefore it would appear that the assumption was that the teaching assistant within the discussion section setting was able to make discriminations among students on the basis of this heterogeneity and to provide the Opportunities for Students to overcome barriers created by differential entering characteristics. The course also provided for common, multiple-choice exams, but the final grade was to be a weighted composite of sixty per cent exam grades, forty per cent instructor grade (determined on whatever basis the instructor considered important). So, a second assumption of this model was that the teaching assistants'grades made discriminations on import- ant instructional setting variables that the common exams did not discriminate upon; a third assumption would appear to be that the the common exams themselves did not create an atmOSphere based mainly on competition for the high exam Score alone. Another goal provided that a course be judged on the the basis of the learning and "scholarship"generated among students 94The University and College of Education goals were Speci- fied by detailed quotations on pp. 53 - 56. 59 by it. The first assumption appears to be that the content pro- vided by the faculty from diverse areas of Specialization would be viewed by students as a sound basis for organizing their approach to the role of educational psychology in their own future teaching, and not be viewed as diSparate entities. The second assumption about what the nature of the learning experience in this course should be is reflected in the list of competencies to be develOped by this course Specified by the Curriculum Review Committee and appearing in detail in Appendix A. With two exceptions the key word in each objective is either "understands" or "recognizes" or "dis- tinguishes" which can be interpreted to mean the Student is able to recall some piece of information for a multiple-choice exam or can apply a principle to a multiple-choice question. A third assump- tion about the nature Of the learning experience in this course is that the more complex problem solving thought process was to take place in the discussion section where "the student will be concerned with the application to classroom problems of the concepts covered during the lecture sessions." Implicit here is the notion that the teaching assistants possessed the skills to help students to integrate diSparate lectures and readings in a problem-solving approach to future teaching experiences. An additional goal of the course was embodied in three sapects. First, interaction between teacher and student was to be a primary consideration in implementing a course of instruction. Second, not only quality instruction but also the teacher as a model were to be the main concern of faculty members. Three, a primary 60 Objective of any course in this sequence of teacher preparation was to contribute to the develOpment of those "basic understandings, skills, and attitudes which characterize a teacher who can reSpond competently to all situations within which he must function."95 The concept involved in this three-part goal was that over a sequence Of courses the models represented by teachers would have meaning to students, due to the interaction of faculty with students. Students were tO be exposed to eight senior faculty members in nine weeks via large lecture sections. The assumption would seem to be that any interaction which might result in students perceiving their teacher as model in some form would be within the discussion section. A consideration underlying each assumption in the "implicit model" would appear to be that a decrease in faculty participation was possible on the basis of projected teaching-assistant performance. Trow has questioned the validity of this assumption, asserting that: teaching assistants are often poor instructors, less competent, less able in the content field, and poorly motivated to be an assistant; since the design of the course is established by the senior faculty, the formal content reflects their academic interests and any impli- cations or relevance Of the material to the students' lives is left to the TA; since the TA is generally weak in content, rarely does true integration take place, and examples, discussions, and activities within the TA's instruction are involved with his beliefs and experi- 95Leland Dean, Associate Dean of the College of Education, Memorandum to the Educational DeveIOpment Project, as Preface to the College Of Education Curriculum Review Committee's Rsport on Under- graduate Curriculum Revision, mimeograph, (East Lansing, Michigan: Michigan State University, April, 1965). 61 ences rather than with content, skills as representative of the content, and the true implications or questions related to the material.96 However, it would appear that the assumption was made that by providing a content seminar in which senior faculty dis- cussed their projected lectures with the TA'S, and by providing an additional seminar concerned with the methods and development of college teaching, the problems raised by Trow could be averted. The above assumptions constitute the "model" underlying the instructional process in the natural setting chosen for the study. It is the set of assumptions which delineates the conditions under which the instructional process Operates in that natural setting and cannot be ignored if the researcher plans to alter the structure of those relationships in any way. Neither those variables implied by the assumptions of the model nor those in Operation in the ongoing situation are necessarily synonymous with those having actual "importance" to the improvement of instruction. Summary Many undergraduate college courses serve as exemplars of the failure to compensate in instructional procedures for the dimensions of any factors beyond the homogeneous descriptions of predominant characteristics in the instructional situation. However, recently a great deal of attention has been directed toward the subject of improving instruction at all levels of higher education, particularly the instruction of undergraduates. 96Trow,.pp. cit., pp. 19-23. 62 Suggestions for improvement often appear to have an amorphous quality because the factors discussed as relevant to improvement are expressed as generalities. Although some overgener- alities might be assumed away as Showing lack of careful thought, the number Of such examples in the literature suggests that this lack of Specificity as to what dimensions of the factors in the instruc- tional setting are Of most relative importance in the improvement of instruction is partially a reflection Of the criticisms of research on teaching and learning discussed in chapter one. The first section of chapter two constituted a review of the literature. Since the setting for this study was at the college level, the review Of the literature is confined to studies of instructional setting variables at that level. The studies chosen are examples of a number of studies reflecting the problems of methOdOIOgy and application to ongoing instructional settings discussed in chapter one. The criticisms by researchers of methodology related to such problems as failure in Specificity of criterion, in maintaining homogeneity and independence, in analyzing interactions was illus- trated from the literature by a series of studies which investigated factors in group behavior. The studies cited were chosen as examples Of work Over a time Span of thirty years to illustrate how sequen- tially over time attempts have been made to increase the precision of the early Lewin model both in terms of theory and methodology, and to achieve more Specific results, and that the degree of increase in stringent empiricism or more Specific theoretical structures is 63 difficult to asses; In addition, this group of Studies was illus- trative of groupings of studies that have occurred around many fac- tors believed tO be of some importance in classroom instruction. Further, the conceptual framework Of the role Of leader and group behavior is particularly germane to undergraduate college instruction where many such courses attempt to employ lecture-discussion formats. The methodological problems associated with the statistical analysis of data involving many variables, presence of data in nominal form, and so on, was illustrated by the example of the number of Studies, Over time, attempting to identify and assess the relative importance of factors believed to be important in predicting college success Of a student. Current studies reflect findings similar in level and precision to those made thirty years ago. Criticisms by those trying to use research findings to improve instruction, to the effect that the results or implications anecdifficult to generalize into the classroom setting and fail to lead to actual procedural efforts to improve instructional results was illustrated by a series of studies from the literature conducted from the point of view of learning theorists, in the classical mold, from the point Of view Of learner characteristics Specifically rela- ted to instruction, and from the point Of view of instructor chara- cteristics. Among all studies,inconsistent findings and lack Of Significant differences made forming any pattern difficult. For example, the studies related to characteristics of the learner suggest that: academic performance seems higher for students with more favorable attitudes toward school, greater interest in the 64 subject area, and greater degree of achievement motivation; who have a more positive self-image, less anxiety in test-taking situations, have greater cognitive flexibility, are less defensive, and are more conforming. The difficulty is that the differences among the groups in the studies are very small and tentative. It is evident from the literature that from whatever particular point of view the instructional complex is chosen to be viewed by the researcher, the difficulty in many cases is not inadequate conception or develOpment in the study but that the results taken in isolation explain a very small portion of behavior when viewed within the entire instructional complex. Accordingly, there have been some attempts tO define conceptual frameworks for instructional research which are based on the assumption that the factors comprising the instructional complex interact. As an example of such an approach, the paradigm Of L. Siegel was discussed, which conceptualizes what Siegel refers to as "the instructional gestalt". This particular framework focused upon four clusters Of independent variables, identified as instructor, learner, course, and classroom environment clusters. Specific variables within each of these four clusters were Specified as the most "potent determinants of what tranSpires within the instructional gestalt". Interactions would Occur between the variables within a given cluster and between variables across the four clusters. Thus the model must focus on the interactive relationships between variables comprising the instructional complex, as well as upon the summary of main effects. In making the transition from the generalized para- digm tO the application of it in research in the instructional setting 65 the difficulties encountered in the study raise again the problem discussed by Campbell and Stanley and Box in chapter one, the reconciliation of models. In the factorial design with analysis of variance the multivariate paradigm was of necessity simplified. In addition, the complex analysis of variance model does identify when significant interaction is present, but unless the interactions are clearly hypothesized at the start, this approach does not provide a basis for empirical interpretation of the nature Of these inter- actions. The second section of chapter two outlined the background for this study to be reported in the following chapters. The setting was the first course in the undergraduate teacher-preparation sequence, which can be categorized as a lecture-discussion group, multi-Section, survey-type course. The instructional content of the course included, secondarily, an introduction to the teaching profession and to the program Of teacher education, and primarily an introduction to the Subject matter in the area Of educational psychology. The actual physical representation of this course involved a lecture-discussion section format made up of a number of composite parts: one course coordinator and evaluator, one teaching intern- ship advisor for the teaching assistants, eight faculty lecturers, sixteen graduate teaching assistants, seven hundred to one thousand- plus students (mainly SOphomores, but some juniors and a few seniors), plus two different hours during the day at which times the format was Offered to the students, using the same instructional personnel. In chapter one Campbell advocated the implementation of 66 research programs within already established instructional settings. Under such a program the initial purpose would be to discover the structure Of relations among variables in the Specified instructional setting, and under what conditions and through what intervening processes this relationship occurs. Under such research conditions not only would it be necessary to take into consideration the reconciliation of research and statistical models as stressed by Box and Stanley, but to take into consideration the "model" underlying the instructional process in the natural setting chosen for the study. This model is the set of aSSumptions which delineates the conditions under which the instructional process Operates in that natural setting. In the course used for this study the "implied” model, composed of a set of assumptions about the structure Of relations among variables in the instructional setting, was reflected in a plan to implement the goals of the University and those of the College of Education to improve undergraduate education. The over- riding goal was tp provide "quality" instruction to large numbers of students while at the same time decreasing the participation of senior faculty in providing this quality instruction. The assumption was made that this decrease in participation was justified on the basis of projected graduate teaching-assistant performance. Under these conditions, the teaching assistant would provide the only means for individualization of instruction or for participative instruction; due to the impersonality of lectures and exams, students would tend to identify the course with the discussion section; because of 67 constantly shifting Subject matter -- the result of changing. faculty lecturers, the TA'S would be reSponsible for clarification and continuity; a primary goal of the course was to act as a first step in the develOpment of the students toward becoming ”competent" teachers, and it rested with the teaching assistant to be a model of teaching competence broadly exhibiting the attitudes, skills, and understandings theoretically Specified as necessary for this role. The above comments summarize the assumptions which constituted the "model" underlying the instructional process in the natural setting chosen for this study. Neither those variables implied by the assumptions of the model nor those in operation in the ongoing situation are necessarily Synonymous with those having actual "importance" to the improvement of instruction. The following chapter, chapter three, is divided into three main sections, each section having subsections. The first section: develOps a rationale upon which the choice of the factors included in the study was based; then presents a discussion of the factors themselves. The second section is a discussion of the instrumentation -- the various measures chosen to represent each of the factors in the study; the factors are grouped by two main cate- gories -- outcome (reward and performance outcome factors) and predictor (student and instructor factors). The third section is an outline of the procedures for the study, including: the participants and the setting; the design of the study; the strategy of data analysis. An additional fourth section restates the questions raised in chapter one about the nature of instruction, in terms of this course. CHAPTER III RATIONALE FOR CHOICE OF FACTORS; INSTRUMENTATION; PROCEDURES The review Of the literature has served to Show that there are a number of ways to view teaching and learning and, on the basis of whichever viewpoint chosen, to analyze any given component Of the instructional situation. While difficulty in utilizing the results of some studies is due to methodological inadequacies in conception and develOpment, in many other cases the difficulty is that the results taken in isolation explain a very small portion of behavior when viewed within the entire instructional complex. No one method, no one style, no one personality factor is more important than another, independent of the realities in the classroom. In this study the purpose is to attempt to unravel the components by discovering the structure of relations among pot- entially important variables in the instructional situation, and under what conditions and through what intervening processes this relation- ship Occurs, in order to determine the relative importance of the variables to course criteria or outcomes in a manner functional to improving instructional conditions. This method is a way of partitioning the diverse multi- faceted factors that constitute the relationships in the instruc- tional situation and of providing a basis for going further into the analysis of the educational process. Since the structure of relations among the variables in this study is derived from the interaction process in a particular instructional situation, their impact upon a 68 69 given student in another instructional situation would not neces- sarily be expected to be precisely the same. However, the underlying concern is the Options which, as a result of the structure of rela- tions among the variables, are Open to students working their way through any given instructional Situation. One question being asked is in what ways the structure Of relations among variables seems to be related to the paths students choose to (or seem compelled to) pursue in the instructional situation when compared with other stud- ents, without inferring these choices are inevitable. Theoretically, no matter what formal instructional Situation is being viewed by the researcher, there is some Overan."job” going on in the situation, and this "job" is a process of "doing what needs to be done" to get to some goal. That is, there is some overriding goal (which may be composed of a series of subgoals) to which all the variables in the instructional situation are supposedly geared. The form this "job" takes is defined by these variables, such as the characteristics of students, the content of the course, and so on. This "job” cannot be understood without taking into account a number of considerations. One consideration which must be taken into account is the impact of different types of subgoals, that is, various phases of the overriding goal. For example, if a stated sub- goal is "to develOp values" (whether or not this is possible is a mute point in this discussion), its impact as a goal must be viewed in relation to the other subgoals for which the instructional situation was created and exists. A second consideration which must be taken into account is that while the form Of the "job" or "task" is defined by the variables, 70 certain variables, or more precisely Specific dimensions of these variables, may carry more relative importance or weight in this definition. Coming together in the instructional Situation are various dimensions of a number Of variables which form a structure Of rela- tions within this context. However, over time these various dimen- sions of the variables exert varying degrees of pressure on the "job" or "task" of reaching the Specified "goal". For example, the students and teacher (or teachers) have entered the instructional situation with different conceptions Of what the "job" or "process” should be like, and over time it is possible that interpersonal pressures created by these differences could interfere with the phase of the task which is goal-oriented. Therefore, at any given moment in the instructional situation the "job" actually has two components: the part which is goal-oriented (in many cases academically-oriented), and the part which deals with the pressures on the instructional situation at that moment. For example, if scores on a test have caused anxiety among students and this anxiety is interfering with the goal-oriented phase of the "job", the dissipation of this anxiety is at that moment a part of the "job". Thus, a third consideration which must be taken into account is how, over time, things are going in the instructional situation in relation to the goals. A fourth consideration which must be taken into account is that among the many factors in the instructional situation the students are usually Of particular interest. For example, even if the focus of interest in a study is the teacher, his role, his attitudes, and so on, inevitably the students (or learners or whatever label is 71 applied) become a focal point in terms of their role, or performance, or any number of other possibilities, in their relationship with the teacher. It is for them the instructional situation was created and exists. In the case of studies within the ongoing instructional situation the structure of relations among the variables creates circumstances with inherent violations of normalcy and homogeneity such that the focus of attention is not only what students have in common but also how students vary and what various subgroups exist. There are a number Of alternative explanations for what goes right, or wrong, in terms Of the "job" or "process" taking place in the instructional situation, that is, for what happens instruc- tionally, depending on one's viewpoint of teaching and learning. When the researcher enters the instructional situation there exist many factors with potential "importance" defining this "job", and depending upon his interests, the type Of study he chooses to do, and so on, he selects, or introduces additionally, those variables relevant to his study. The question in this particular Study is which factors in the instructional situation appear to make a real difference for the purpose Of course develOpment. This necessitates including not only potentially important entering-course character- istics but those which over time play a role. The core Of any investigative strategy is its set of restrictions. Eventually, some restrictions must be put on the strategy, such as limiting the number of factors or variables used, or limiting the freedom with which they are permitted to Operate. In this study the problem becomes that Of including those factors which not only have relevance but also potential importance to the improvement 51: 1e: 51; ac: 'va CI" 72 of instruction. Several considerations must be taken into account. Campbell cautions against initially rushing in after variables unrepresentative of the instructional Situation and learning process but beloved of the researcher, although this does not preclude their later value and use.97 A second consideration is that those who have designed the course or sequence of instruction in a given instructional situation supposedly do not operate in a vacuum independent of know- ledge Of teaching and learning, and factors they have chosen to empha- size may have potential importance. This necessitates taking into account the "model" or set of aSSumptions of the course designers in addition to the actual instructional process, Since some of these factors may be emphasized in the "model" and may or may not be in actual Operation. A third consideration which must be taken into account is the suggestions from research findings and from theory which support Special attention to variables that seem indicative of critical instructional conditions. Similar considerations are necessary in the choice Of measures; these measures are proxies or stand-ins for the factors chosen for inclusion in the study as potentially important to criti- cal instructional conditions. According to Campbell, one of the things necessary in such research is to use measures (to tap factors chosen for inclusion as outlined in the previous paragraph) which intrude as little as possible into the instructional situation.” Two sources of 97Campbe11, loc. cit., pp. 263-269. 981b1d., p. 265. be re 63‘. tc St in Su th; PEI tic tic 73 such nonreactive "proxies" are those derived from instructional measures which are already a part of the natural Situation, Such as past performance factors represented by grade-point average and scores from Standardized tests of general ability, or course outcome factors represented by course test-scores. The remaining factors considered to have potential importance must be represented by new measures. Where a number of measures are available to represent a given factor, all else being equal, choice Should include considera- tion Of the degree of the measure's unobtrusiveness, and its possi- bilities for future use as a permanent part of the research situation based on a conception of its relationship to the instructional process, rather than being seen as introduced just because of a conSpicuOus new experiment.99 Table l (p. 74), lists the factors potentially important to relevant course criteria or outcomes chosen for inclusion in this study, based on the earlier discussion of considerations to be taken into account in choosing variables. Beginning at the left, each subsequent list contains those factors in the preceding list plus those included over time, divided into "classes" by headings, for the understanding of the reader and explanatory purposes. Factors concerning students were chosen from the areas of performance (both past and in the context of the instructional situa- tion in the study), attitudes, personality, and personal characteris- tics. Studies in the review of the literature suggested that prior 991bid., p. 266-267. TABLE 1: ENTERING-COURSE (AID I) Personal Sex Age Major Past Performance GPA OQT Personality Extraversion Neuroticism Social acceptance Test anxiety Pretest anxiety Attitudes Learning set Reason enrolled Pre-attitude Course-Specific motivation at pretest 74 SUMMARY OF COURSE PREDICTOR FACTORS MID-COURSE QAID 11) Personal Sex Age Major Credits earned Current load Past Performance GPA CQT Personality Extraversion Neuroticism Social acceptance Test anxiety Pretest anxiety Midterm anxiety Attitudes Learning set Reason enrolled Pre-attitude Course-Specific motivation at pretest Course Performance Pretest score Instructor Extraversion Neuroticism Social acceptance Defensiveness Rigidity Authoritarianism END-OF-COURSE QAID III) Personal Sex Age Major Credits earned Current load Past Performance GPA OQT Personality Extraversion Neuroticism Social acceptance Test anxiety Pretest anxiety Midterm anxiety Final exam anx. Attitudes Learning set Reason enrolled Pre-attitude Post-attitude Discussion att. Course-Specific motivation at pretest Course-Specific motivation after midterm Success expect. Accuracy judgment Course Performance Pretest score Midterm score Instructor Extraversion Neuroticism Social acceptance Defensiveness Rigidity Authoritarianism Risk Course load Teaching exper. 75 knowledge and ability Of learners possibly play a role in explaining behavior in the instructional process. The possible role Of past performance also happened to be an assumption of the instructional situation "model" (as discussed in chapter two to the effect that educational opportunity was to be a part of instruction so that enter- ing characteristics of students would not be the main determiner of degree of instructional impact). Prior knowledge and ability are reflected by a number of factors: past academic performance in instructional situations with an underlying institutional similarity to that Of the instructional situation under study, and represented by the institution as a gradepoint; past performance in situations ostensibly indicating potentiality of general intellectual ability, not only Of verbal ability but also conceptual and problem-solving ability, and represented by test scores from standardized measures of such abilities; past performance on content (facts, concepts, and so on) with a possible underlying similarity to content which is a part Of the instructional process in this situation, and reflected in the content of the pretest; past performance on content of the course as related to later performance in the course, and represented by the midterm test as related to end-Of-course performance. The review of the literature suggested that while the role of learner attitudes and values on broad, general categories is: difficult to interpret in relation to classroom instruction, a number of Specific aSpects of attitudes possibly are important in explaining behavior in particular contexts. Two such points made were: no major changes in attitudes and values in broad categories could be expected over the short Space of a given instructional situation although 76 changes in attitudes related to Specifics of that situation were possible; student indifference and docility are Symptomatic of student attitudes of the irrelevance of educators to students' lives. The importance of student attitudes in relation to the learning process was also an assumption of the course "model" of which one objective was that the instructional process contribute to develop- ment of attitudes which characterize a teacher, and another was that course content provided by faculty would be viewed by students as a sound basis for organizing their approach to the role of educational psychology in their own future teaching. A third suggestion from the review of literature was that students differed in their prefer- ence for factual and conceptual learning experiences, not necessarily based on extent of ability to deal with conceptual learning tasks. The importance of providing Opportunities for students to reflect these differences was also an assumption of the "model". These various Specific SSpects of attitudes were reflected by a number of factors included in the Study. One set of attitudes reflected qualitative judgments about the instructional process or "job" itself, of what students expected of the course and what they felt had occurred, of how they valued instructional activities and the course as part of their teacher training. A second set of atti- tudes pertained to the sepirations and expectations students hold for themselves in terms of their performance, how successful they feel they've been and how accurate their assessment. Such attitudes may serve as preconditioners of student attitudes toward future involve- ment with Such content. A third set of attitudes pertained to academic "set" or preference for factual or conceptual learning, 77 such attitudes possibly Operating as selective tendencies under certain instructional conditions. The review Of the literature was not encouraging in Sugges- tions of personality factors with potential importance to the instruc- tional process. However, it would seem that certain facets of some variables which represent entering differences among Students might Operate as selective tendencies under certain instructional condi- tions. Furthermore, it would also seem more realistic to explore only factors with which, in interaction, students and teacher could logically be expected to come to grips in the instructional process. Factors reflecting both these conditions were chosen which seemed to have importance to the instructional process under both large group and small group conditions. The first set of factors reflects anxiety related to academic Situations in general, and Specifically to the "job" in the instructional situation in which evaluation is involved._ The second set of factors wane those considered to have the most potential relevance to performance in the discussion section. In a "discussion section" which really does not qualify as a small group, a person Of more reticent nature may be at a disadvantage because of more recalcitrance to make themselves known and heard. Or possibly, those who experience a high degree of anxiety not necessarily related to test-taking may find it constraining if what is demanded of them involves personal commitment beyond that which they feel able to make. The review of the literature suggested that while personal factors such as sex, age, degree Of involvement in the content area, pressure from other activities and so on do not seem to have a significant impact in terms of the overall instructional process, they 78 are possibly related to Specific aSpectS of the instructional process. The following Such factors are represented in the study: sex, age, major field of interest, credits earned academically, and current academic course load. After reviewing over eight hundred studies of teacher charac- teristics as factors in the instructional process, Getzels and Jackson expressed a general pessimism about the usefulness of such factors in explaining the instructional process in the classroom. Their sugges- tion was that a possible fruitful area of concern appeared to be different types of cognitive functioning in addition to those currently assessed by tests of general ability. In addition, the tentative prOposal was made that attitudinal and behavioral correlates of cogni- tive ability might be more important in understanding teaching success than are the abilities per se.100 The possible importance of such factors to the teacher's role in groups using "discussion method" as contrasted with "lecture method" was raised by McKeachie in his review of teaching methods in which he pointed out that a number of Studies suggest the possibility that discussion in the classroom encourages and develops critical-thinking and problem-solving ability in students but the results are not apparent, Since talking does not necessarily mean the talkers, including the teacher, are evidencing prediSposition toward, or engaging in, critical thinking and problem solving.101 In a study of teachers in classrooms, Jackson noted that Such prediSposi- tions might not necessarily be in evidence in the teacher during 100J. W. Getzels and p, w. Jackson, lOc. cit., pp. 574-576. 101w, J. McKeachie, in N. L. Gage (ed.), 122; Slim: P- 1122' 79 interaction with Students, but it seemed to be important in what he called the "preactive phase" of teaching -- that is, during time Spent contemplating what has or will take place in their instruc- tional situations, among teachers considered superior by admini- strators and colleagues.102 In a study of the components and pro- cess involved in inquiry behavior,103 Shulman found that supervising teachers tended to rate more highly those student-teachers classed as dialectic (that is, those who focused on problems rather than solutions, interacted frequently with students through discussion, easily roamed widely over course materials, and always allowed for the unexpected)than those student-teachers classed as didactic (that is, those who strove for immediate, certain and unambiguous closure, avoiding unpredictability), irrespective of whether the supervising teacher himself (or herself) might be so classed.104 These studies raise two suggestions: first, the possible importance to the inst- ructional process in general of such factors as evidenced by teachers, and second, their importance in this study as evidenced in the 1ozPhilip W. Jackson, lOc. cit., pp. 144-155. 103Rather than using a term such as problem solving, Shulman uses the term inquiry process, basing his concept of inquiry on Dewey's description of that process, and dividing the process of inquiry into four parts: problem sensing, problem formulation, search, and resolution. 104Lee S. Shulman, Michael J. Loupe, and Richard M. Piper, Studies of the Inquiry Process, United States Department Of Health, Education, and Welfare, Final Report, Project 5-0597, (East Lansing: Educational Publication Services, Michigan State University, 1968), PP. 148-155. 80 instructors who interact with the students who themselves are even- tually to assume the role of teacher in the instructional process. This latter suggestion Seems relevant on the basis of the goals and assumptions of the course "model" and the expected role of the graduate assistants as instructors, as described in chapter two. Briefly, the goals were: first, interaction between student and teacher was to be a primary consideration; second, one result of this interaction in the discussion section was that students were to be assisted in integrating materials from lectures and readings, and in applying these integrations in a problem-solving approach to future teaching roles; third, another result of this interaction was that Students were to be assisted to integrate their course experience into not only basic Skills but also attitudes and understandings which supposedly characterize the competent teacher; fourth, another reSult of this interaction was to be that Opportunity be provided for hetero- geneity among students in terms of differential characteristics; fifth, not only quality instruction for the students but also the teacher as model were to be the main concerns Of faculty members. A considera- tion which seemed to underlie each of the above points was that a decrease in regular faculty participation was possible on the basis of projected teaching-assistant performance in relation to each of the goals. Thus the roles of the teaching-assistant instructor appeared to include: moderator, and where necessary, instigator of discussion; assistor in Student integration of seeming diSparate course content and in usage of such content in a problem-solving approach to issues of teaching and learning; contributor to development of attitudes 81 of students toward their future role as teachers; discriminator of differential characteristics among Students, and evaluator of their role in student performance; and lastly, the role Of model. As was discussed in chapter two, under the time-Span of the typical college course the effect of a single course Of instruction upon long-held values of students is questionable. Further, what this role of model was supposed to entail for the faculty, let alone the teaching-assis- tant instructors,was not clear. However, two assumptions are tenable. Possibly the instructor was to be viewed as a model Of the "attitudes, skills, and understandings which characterize a competent teacher" -- a goal discussed in chapter two. More realistically, the Students might view the instructor as a model Of a Stance taken toward the activity in which the students themselves were being asked ito engage -- a model Of a stance, or prediSposition toward, a critical- thinking, inquiry approach to issues raised through the course content and its application. The suggestions in the studies cited above and the assumptions of the course "model" seem to stress the importance of the teacher's ability to engage in inquiry, critical thinking, and SO on, not only in the preactive phase Of teaching as suggested by Jackson but in the interactive phase also -- that is, on the classroom firing line. Getzels and Jackson pointed Out the general paucity of studies which suggest possible characteristics relevant to such abili- ties in the instructional process.105 In this study, the question is 105J. W. Getzels and P. W..Jackson, in N. L. Gage (ed.), loc. cit., pp. 574-576. ‘1’ 144 .‘J 82 confined to which such factors or prediSpositions (rather than behaviors themselves) characteristic of teachers might most support students in their choice to use their competencies to engage in "problem-solving" activities. In his discussion Of inquiry learning Shulman points out the amount of current research focused on enhancing such processes in children and the results of such research suggesting the importance of such processes in the development Of the cognitive functioning of children. He suggests that these processes are very likely to be dampened without adequate participation and modeling on the part Of teachers, and that research on pupil inquiry may be 6f limited value without equal attention to teacher inquiry. In his study of inquiry behavior in student-teachers Shulman indicated certain factors which seemed important in such behavior. Those evidencing a high degree of inquiry prediSposition also tended to evidence high associational fluency, high cognitive complexity (preference for ambiguity, assyme- try, the unexpected, as Opposed to preference for the regular, the predictable), high verbal problem-solving ability, low expressed test anxiety, willingness to risk on a test of logical thinking, and "lib- eralism" in political values.106 The inquirer was: prediSposed to balance the risk Of inquiring against the risk of not so doing, and the greater the degree to which the individual was comfortable with complexity, the less the perceived risk involved; prediSposed to become personally involved in problems and the process of resolution; pre- 1063. Shulman, M. Loupe, and R. M. Piper, loc. cit., PP. 53-97. 83 diSposed toward flexibility in thought and in usage of the indivi- duality Of students in the learning process.107 In a study of concepts not unrelated to those investigated by Shulman, the importance which Shulman attached to such processes in the develOpment Of the cognitive functioning of children is extended by Perry. He stresses the importance of similar processes, but in particular their cruciality during development taking place in late adolescence and young adulthood, and he tentatively posits such development as a "stage" of development, similar to the Piagetian sense of "Stage". In his longitudinal study of a group of students throughout their college years, Perry postulated that the ultimate purpose of students in the period of late adolescence and young adult- hood is "to find those foams through which they may best understand and confront the human condition". His study focused on how such young peOple "orient themselves in a relativistic world through the content and style Of ongoing acts of commitment, and the forms of the: ae Options through which some students appeared to withdraw or retrench at various points in the develOpment." The study reSultS enll>'l:1nasized the interweaving Of hierarchies of values with hierarchies 0f thought (what Perry termed interweaving meta-valuing with meta- th3"~‘l:7nking).108 The develOpmental shheme depended heavily on Piaget's cm31c:epts Of assimilation of an experience and accommodation of struc- \—-—_ 107Ibid., pp. 154-155. 108William G. Perry, Jr., Forms Of Intellectual and Ethical agglmment in the College Years: A Scheme, (New York: Holt, Rine- hart and Winston, Inc., 1968), pp. 8-54. 84 structure through transformations and recombinations which can result in new and more differentiated structuring Of experience, and the scheme also reflected processes ascribed to Piaget's stage of formal Operations. Perry's postulated "stage" and the Study results Sugges- ted that powers of detachment and Objectivity consequent on the ability to "meta-think" make it possible for the individual to address the environment in Such a way as to be able to move from the moral to the ethical, from the formal to the existential. The study Suggested that during this "Stage" of development the most difficult moment for students in the instructional process seemed to occur "at the transition from the conception of knowledge as a quantitative accretion of discrete rightnesses to the conception of knowledge as the qualitative assessment of contextual observations and relationships".109 In discussing those teachers from whom they felt the greatest sense of support in their attempts to make this transition, Students inferred a number of prediSpositions charac- teristic of these teachers. One such prediSposition was what might be classed as a type of "risk" -- that is, a willingness under ambi- guous conditions to live with these ambiguities and based upon them, to express judgments which might deviate from the status quo, yet expressed with confidence, willingly exposing themselves to criticism rather than playing it cozy or equivocating. An associated prediSposi- tion was reflected in less need to make socially desirable reSponseS (that is, those desired either by the "establishment" or the Student 1091bid., p. 210. 85 population), less need for social acceptance. An additional predis- position was "an openness -- a visibility in their (the teachers') Own thinking, grOping, doubts, and styles of commitment".110 The previous theoretical considerations in relation to teacher traits led to the inclusion in this study Of factors on the following continuums: extent of conservatism and uncertainty in judgments in ambiguous contexts; extent of conservatism in beliefs and extent of proneness to control events and to settle for the Status quo; extent of rigidity in interaction with Students; extent of need to reSpond in socially desirable ways and for social acceptance; extent of reticence or recalcitrance to initiate interaction with students; extent of anxiety created by demands for personal commitment beyond that which the individual feels able to make. The focus of interest in this study was upon the relative importance of potentially important variables to relevant course criteria and outcomes in a manner functional to improving instructional conditions. The course criteria were defined by the course goals and objectives. The impact of the instructional process upon Students in terms of certain Of these goals (such as beginning develOpment of atti- tudes toward the future role of teacher) was assessed indirectly by Specific questions in attitudinal measures. The study was concerned with what, theoretically, was the relative importance of the "predi- ctor" factors in terms of the assumptions underlying the course, and focused directly on what, in fact, was the relative importance of these factors to actual course outcomes -- the Outcomes themselves 1101616., pp. 209-215. f5] 1 led 86 defined in terms of assessed performance. The "outcome" factors thus defined were included in the study in the form of Student scores and grades, derived from course tests and instructor assessments. Reference must be made to two limitations on factors included in the study. The list of "predictor" factors indicates that the factors have been limited to those characteristic of students and instructors. McKeachie, after reviewing literature on environ- mental variables in the instructional situation such as books, films, and teaching machines, indicated that Studies of the effects of such factors alone have not yielded definitive results, and sugges- \ted that their importance depends upon the Objectives of a particular instructional Situation, the characteristics of the particular stud- ents, and the excellence of the materials.111 In the instructional Situation used in this study such materials were standard and consist- ent across all classes. Consideration of such factors was confined to assessment of student reaction to their usefulness, reflected in the judgments made by Students on Specific questions in their evaluations of the course and the instructors. The second limitation was the exclusion of the senior faculty. Previous surveys of Students in this course of instruction (under conditions of a new faculty lecturer approximately every week, seen three times) had indicated that students were unable, at the end Of the course, to distinguish among the faculty without assistance in identifying the presentations, and that students identified the course experience with the discussion section. It was felt that useful assessment of the faculty contribution would be der- ived through the condensed means of end-of-course judgments of students 111w, J. McKeachie, loc. cit., pp. 1173-1258. 87 concerning the faculty role in their course experience and indirectly reflected in exam performance on recall and applied portions. These limitations were imposed to avoid overloading the analytic model and were judgments based on the review of the literature in chapters two and three which Suggested that the core of any given instructional situation is the interactive relationship between students and teacher. The following section outlines measures (chosen via the cri- teria discussed previously) used as "proxies" for factors in the Study. 11. INSTRUMENTATION The term "outcome variables" referred to those factors consi- dered to be tangible representations of the theoretical assumptions of the course "model". These outcomes were meSSured by course examina- tions and grades. Each outcome was analyzed on the basis of a set of predictor variables. The term "predictor variables" (themselves mea- sured by tests and questionnaires) referred to those student and inst- ructor characteristics considered important to the Outcome variables. OUTCOME VARIABLES The Outcome variables involved two forms: performance and ' reward. Course performance factors were represented by examination scores. Reward variables were represented by: grades assigned student scores on examinations; grades assigned by instructors; final grade assigned in the course, a weighted composite Of all other course grades. Reward variables The term "reward" was applied to the grades because they represented the value placed on Student performance by a highly competi- tive system. Represented in student grade were components of student effort, student ability toward whatever criteria were being judged, and “Co. .010. 88 a component Of instructional bias in terms of what was chosen to be judged. For purposes Of analysis, reward data was coded l - 5 (F - l, A I 5), the "F" being assigned the real number value of "1" for analysis as it represents an actual range of performance Scores at the lower tail of the curve. Reward variables included the final grade in the course, the midterm grade, the final exam grade, and the instructor grade. Course grades as valid measures are always questionable. Judgments concerning their validity depend upon those elements of performance which are measured, and the reliability not only of this measurement but of the cut-Off points upon which the grades are based. The reliability between the midterm grade and midterm exam (r = .918) and between the final exam grade and the final exam (r = .932) suggests the grade for examinations was representative of performance scores on the tests. There was a relatively low relationship between midterm grade and final exam grade (r = .502). The midterm grade and final exam grade were grade representations of the midterm exam and final exam scores. It was these grades rather than the scores themselves which were used in calculation of the final course grade, by means of a weighted transformation procedure in which the midterm grade counted twenty per cent, the final exam grade forty per cent, and the grade awarded by the discussion section instructor forty per cent. The mean final course grade was 3.633, with standard deviation .821. The instructor grade was even more difficult to assess. weighted to account for forty per cent of the final course grade, it accounted for almost that much in a part-whole correlation with course grade 89 (r = .642, r2 = .412). Its correlation with midterm and final exam grades was not high (r = .372 and r = .307), and was .371 with Student grade point average. Instructors were given virtually a free hand in the composition and assignment Of this grade, and with the excep- tion of three instructors, did not Specify their various bases for judging student performance. Course Performance Variables The term "course performance variables" was applied to the actual results Of Student performance as measured by examination scores. These examinations took the form of multiple-choice tests over content and simple application of content from the readings and lectures in the course. Course performance variables included the pretest, the midterm examination, and the final examination includ- ing both recall and application sections. The pretest consisted Of forty items, Supposedly the best performing items available to test student entering-course knowledge and ability in the course. Chosen as representative of the various test-item areas in an item-pool being develOped to form a comprehensive exam for the course, the items had at least face validity in terms of their relationship to the course content. The internal reliability of the pretest was r = .763 (Kuder-Richardson Formula #20).112 Mean performance was 22.205, with standard deviation 3.366, as indicated in Appendix B. Appendix C indicates that its relationship to the midterm 112Robert L. Thorndike, "Reliability", Educational Measurement, E. F. Lindquist, ed., (washington, D. C.: American Council on Education, 1963), pp. 586-597. 9O examination and the final examination was low but consistent for each (r = .290 and r = .297). The forty-item multiple-choice midterm examination had a mean of 30.716, with standard deviation 4.046. The items were directed to the readings and lectures included in the first half of the course, and was primarily a recall exercise. The internal reliability of the test was .847, using the formula cited in the previous paragraph. Appendix C indicates that the relationship between this exam and the final exam was .526. The final examination was composed of eighty items divided into two sections. The first section dealt principally with content and lectures in a recall format. The second section consisted of series of questions related to simulated actual settings outlined in the test and tO which the Students were to apply the information and principles presented in the course in order to select the correct answer. The final exam mean performance was, for the total exam, 53.801 and standard deviation 7.157. Based on the Kuder-Richardson formula previously cited, the reliability for the total exam was .710. Appendix B indicates that the exam was negatively skewed (-.685) and also leptokurtic (Kurtosis = 5.183). However, this was attributable almost in its entirety to the extremely high kurtosis on the applied section of the exam (Kurtosis a 16.815). Since the first half Of the final exam tapped student recall of course material, it was not unexpected that the relationship between this section and the midterm, also a recall exercise, was greater than that between the application section of the final and the midterm (r = .503 and .412 respectively). 91 The applied section was more closely related to student entering ability aarici test-taking ability than was the recall section, as indicated by a touch higher correlation between the former and the Qualifying Test- Total (r = .420 as compared to r = .302). PIRIEDICTOR VARIABLES The term "predictor variables" was applied to those student and iristructor characteristics considered to have importance to the course Otitcomes, and used in their analyses. The predictor variables were Of troo forms: those involving student characteristics and those involving irrstructor characteristics. S_tudent Charac teristic 3 Previous research Such as that discussed in Chapter II has iridicated the relevance to classroom instruction Of factors in Such caitegories as past performance, attitudes, and personality. Student Cllaracteristics considered to have not only relevance but importance tc> the outcome variables were utilized from such categories. Past Performance variables have importance not only as indicators <>f past accomplishment but in the sense that this past accomplishment is a partial determinant of future performance on the basis of ability, motivation, and the degree others judge student effort on this past basis. The past performance measures take two forms: the results of standardized tests of ability and background, and accumulated Student gradepoint for previous academic work. Both are Standard procedures in many universities. The standardized test of entering ability and background was the College Qualifying Test, university-administered, which provides both verbal and quantitative scores in addition to the total score. So that the range of 92 scores met the requirements for a particular computer analysis, the rrezsults were divided by two such that a score indicated as twenty-five in t:t1£a Study results of the next chapter would actually be a standard score at the fiftieth percentile. Possible truncating effects in the zarnaalysis were not noticeable by inSpection. The verbal, quantitative, zarrd total Scores are referred to in the analysis as CQT-V, -Q, or -T. Tfine Other measure, grade-point average, was self-reported by the students A; high relationship between self-reported GPA and university-calculated (SPA was found on a randomly selected sample of sixty-three students in ttmastudy (r = .945). Previous research indicated students reliably report their GPA ' s .113 Course Performance variables involved the pretest and midterm Exxam scores, in some of the analyses of course outcomes. The pretest [Irovided an indication Of previous background in the content Of the caourse. The midterm Scores provided not only a measure of performance £18 judged by the course planners but has inherent in it a motivational component for outcomes late in the course. Statistics pertaining to ‘these measures were discussed in a previous section of this chapter. Attitude variables selected for inclusion were those relevant to the caurse outcomes as interpreted by the course designers. One set Of attitudes reflected qualitative judgments about what students expected of the course and what they felt had occurred, about how they valued instructional activities, and the value placed on the course as part of their training. Such variables were represented by measures of pre- and post-course attitude and motivation, attitudetoward the dis- 113O. M. Davidson, "Reliability of Self-reported High School Grades," unpublished research report, American College Testing Program, Iowa City, Iowa, 1963. 93 cussion section experience, and reason for enrolling in the course. A second set of attitudes pertained to the aSpirations and expectations students hold for themselves in terms of their performance and eventual "reward" in the course, and have importance through the degree Of congruence between these attitudes and actual performance. In a course which serves as an introduction to an area, such attitudes may serve as preconditioners of student behavior and of attitudes toward future involvement in the content. A third set of attitudes pertained to academic ”set", and represented entering differences between Students which might operate as selective tendencies under certain instructional conditions. For instance, in a course in which integration Of ideas into a conceptual framework is stressed rather than recall Of Specifics, students oriented toward or with a preference for factual learning might find themselves at a disadvantage. Three instruments were involved in the first set of attitudes. In indicating reason for enrolling in the course, students reSponded to one of five reasons on a one to five Scale moving from negative to neutral to positive -- (#1. The course was required or I would not have enrolled, to #5. The course was not required but I wanted to take it.). The reliability of the scale was not established. Its relationship to to the other two instruments was low (r = .394 and r = .174 with pre- and post-course attitude, and r = .091 with discussion attitude). (Appendix C). Student pre- and post-course attitude and motivation toward the course itself were measured by instruments adapted from the instructional rsearch of Siegel and Siegel, and involved a series of Statements 94 related to the instructional setting and the conduct Of instruction.114 1% yes or no reSponse to each item indicated whether the Student felt that item indicative of the course. Statements moved from very negative to very positive, with the students' score the median yes reSponse. The developers of the instrument reported the procedures in its develOp- ment and reported Split-test reliability of .780. Neither the Siegel results nor those Of this Study Showed Significant relationship to past performance measures, Suggesting that course-Specific attitude rather than general academic motivation was being measured.n5i The original fifty-seven items were used in the past-sense in the post- course instrument. The test-retest estimate of this adapted version ms .633. Student attitude toward Specific discussion section experience was represented by the sum Of positive reSponses on an instrument Of 23 statements, related to procedures, qualities, and a ttitudes of the discussion section instructor. The items used were Chosen from a series of items being develOped at the University for a n all-University student-attitude-toward-instruction scale. The :L 1: ems had undergone three administrations and analyses and were C: omsidered internally consistent with ability to discriminate. The Sp lit-half reliability of the instrument used in this study was r = 317.116 114Laurence Siegel and Lila C. Siegel, The Instructional Gestalt Melevised University Courses (mimeographed research report, Miami University, Oxford, Ohio, 1966), p. 49 115Ibid., p. 52 116Linquist, 10¢. cit., pp. 586 - 597 95 The second type of attitudes, pertaining to the aSpirationS and expectations students hold for themselves in terms Of performance, has only recently been incorporated into educational practice in any Systematic way. Lewin, Festinger and Sears directed attention to this concept. Lewin described level of aSpiration as a phenomenon Operating in a situation involving choice Of a future Objective, and as a complex of three factors -- pursuance Of sucess, avoidance of failure, and subjective probability judgment. The strength Of these forces depends upon the particular way the individual sees his past experiences and cultural referentSFD In this study, scores for high, actual, and low expectation for the first two administrations were determined by presenting students a Scale of potential grades from F to A+ (numbered one to twelve) and asking them to circle the grade expected. The correlation between administrations one and two was .465 (Appendix C). In a study utilizing measures at the beginning, middle, and end of this course utilizing larger numbers of students, the correlations between one administration and the next of the same variable-type ranged from .500 to .757. The first two administrations were in the above form, in the third administration the student was presented the same grade list but asked to indicate the chances in one hundred he had of receiving each grade with the provision that the total chances indicated add to one hundred for all estimates. The score was determined using the same twelve-point scale and awarding the scale score for the grade showing 117K. Lewin, T. Dembo, L. Festinger, P. Sears, "Level of ASpiratiOn," Personality and the Behavior Disorders, J. McVicker Hunt (ed.), (New York: Ronald Press, 1944), p. 376. 96 fifty per cent level Of probability as determined additively. There was a constant, relatively high relationship between theSe measures and past performance, course examinations, and grades.118 The third form Of attitudes, pertaining to student tendency to prefer factual or conceptual learning, involved the use Of the Learning Set Scale develOped by Siegel and Siege1319This scale provides an index, moving from factual to conceptual, of Student preference toward type of learning. The instrument was composed of thirty-one items each consisting of three statements -- one more conceptually oriented, one more factually oriented, and one judged a compromise. The student selected the condition most preferred and the condition least preferred in any item, and his score was determined by assigning a minus one to each factually-oriented choice, plus one to each conceptually- oriented choice, and no Score to neutral items. Thus, for any given item if the student reSponded positively to the conceptual statement and negatively to the factual statement, the score was plus two; converse reSponseS would be minus two. All statements were scaled for degree of factual or conceptual orientation, the weighting Of the scale determining the score in any given item. Split-test reliability, r = .90, and test-retest reliability, r = .92 (using a five-day interval) was reported by the instrument authors. In develOping scale validity, the authors report no relationship to standard measures of ability such as 113J. T. Parmeter, "An Evaluation of an Introductory Course in Teacher Education," (unpublished doctoral thesis, Michigan State University, 1970), p. 78. 119Siegel, pp. cit., p. 41. 97 college entrance examinations or a measure of creativity (Guilford's creativity tests)¥z'ln an earlier Study the authors discussed validity in relation to performance and indicated that under classroom conditions controlled for such factors as teacher-student interaction and ability, set discriminates on performance between classes oriented either factually or conceptually.121 Personality variables, while a rather global group, have import- ance in that certain facets Of some variable which represent entering differences between Students might Operate as selective tendencies under certain instructional conditions. Variables were chosen which seemed to have importance to classroom instruction under both large group and small group conditions, and to the questions raised in Chapter I. For example, a student with actually a good graSp of the subject matter but with high anxiety toward test-taking situations could be at a disadvantage under conditions of the highly competitive, large-group multiple-choice-test situation, or the student tending to be more introverted, less willing to commit himself verbally, could be at a disadvantage under conditions in which his grade depends upon class participation. The first set of measures attempted to tap those factors associated with anxiety related to the academic Situation. The first measure, the Alpert-Haber Test Anxiety Scale, was used to provide an indice of general test anxiety. The items pertain specifically to anxiety experienced 12oIbid., p. 47. 121Laurence Siegel and Lila C. Siegel, "Educational Set: a Deter- minant of Acquisition," Journal of Educational Psychology, 56, 1966, p. 1-12 98 during the taking Of examinations, based on the concept that while some anxiety may be facilitative, debilitative anxiety can depress performance}:22 In this study the facilitating and debilitating Scales were combined into a single index of debilitative anxiety. The scale authors indicate that general anxiety Scales and scales of general test anxiety measure different attributes and that the latter are more useful in predicting academic performance. The test anxiety scales accounted for variance in academic performance over and above that accounted for by ability and aptitude measures. The reported test-retest reliability over a ten-week interval was .87 for the debilitative Scale, .83 for the facilitative scale, and Over an eight-month interval, .76 and .75 reSpectively. There were consistent negative correlations between the Albert-Haber and measures of performance in the study. (Appendix C) The second instrument used to assess academic anxiety was a meaSure Of test anxiety to Specific instances. Students were asked to indicate degree Of anxiety before beginning the pretest, midterm and final exams by choosing one of five alternatives moving from very little to very much. There was almost no anxiety reported before the pretest and virtually no correlation with other variables, and anxiety increased over successive exams, the midterm and final exam anxiety scores correlating positively with the Albert-Haber Scale (r = .314 and r = .322), and low negative correlations with the examination Scores. In addition, students were asked to indicate 122Richard Alpert and Ralph Haber, "Anxiety in Academic Situations," Journal of Abnormal and Social Psychology, vol. 61, 2, 1960, pp. 207-15. 99 how successful they felt their performance had been, selecting one of five choices moving from very unsuccessful to very successful, and to indicate the degree to which they felt the exam reflected their knowledge of the course content, selecting one of five choices moving from very inaccurate to accurate. The second set of personality variables were those considered to have the most possible relevance to performance in the discussion section. For example, among those students of lesser ability or those who, to the extent that GPA represents a degree Of test-wiseness, have not mastered the skill, a more Outgoing nature, more carefreeness, may be an advantage in a discussion section that really does not qualify as a small group. Or, possibly those who experience a higher degree of personal anxiety, not necessarily test anxiety, may find it constraining if what is expected of them is not made clear, or if what is demanded involves personal commitment beyond that which they feel able to make. In choosing the instrument, consideration was given to the fact that what was wanted was not detailed information on a great many personality variables for diagnosis, but an indicator of dimensions that seemed most important to actual differences between students in terms of course outcomes as Specified by the course planners and the processes whereby these outcomes are supposedly attained. The measure chosen was the Eysenck Personality Inventory which is considered to measure two major personality dimensiOns, extraversion-intraversion and neuroticism- stability.123 Consisting of two twenty-four item Scales which provide 12311. J. Eysenck and Sybil B. G. Eysenck, Eysenck Personality Inventory (San Diego, Cal.: Educational and Industrial Testing Service, 1963). 100 estimates of degree Of individual outgoing, carefree, social inclina- tions and anxious, overresponsive, unstable prediSpositions, it also includes a nine-item "lie scale" to identify those subjects showing "desirability reSponse set," this latter adapted from the MMPI. Test-retest reliability coefficients range from .80 to .94, Split-half estimates of item intercorrelations for each scale range from .75 to .90. American college norms are based on data by Gideon, Gordon, Jensen and KnappEZi The concept Of validity to a large measure rests on the particular criterion one uses in defining any given personality variable. The reader is referred to the discussion of this problem by Jensen, Lingoes, Stephenson, and vernon in Buros, The Sixth Mental Measurement Yearbook,”S and to the theoretical bases for these dimensions in Eysenck, Personality Structure and Measurement}25 Personal variables was the term applied to student characteristics such as sex, major, age, credits completed in university program, and current load. In certain instances because of Space requirement,the factors are referred to by a code number in the diSplay of the data in Chapter IV and the Appendix. For example, social science majors are referred to by the code number "5" where Space necessitates. The code reference appears in Appendix A. 12“Eysenck and Eysenck, pp. cit., addendum 125Oscar Buros (ed), The Sixth Mental Measurements Yearbook, (Highland Park, N. J.; The Gryphow Press, 1965), pp. 286 - 295. 126Hans J. Eysenck and Sybil B. G. Eysenck, Personality Structure and Measurement, (San Diego, Cal.; R. R. Knapp, 1967). 101 Instructor Characteristics The measures served as proxies or stand-ins for the factors specified for inclusion in the Study, and therefore were chosen to approximate the theoretical considerations emphasized in the dis- cussion of choice Of instructor factors. Any measure is itself defined by those theoretical bases used in its development. In addi- tion, the great demands upon the instructors' time both in terms of their teaching duties and their own studies were taken into account in restricting the extensiveness and time requirements of the measures. Encouragement of problem-solving behavior (or inquiry, or however such processes are defined by their bases in various theories) was suggested in the literature as: an important consideration in teaching at all levels; particularly critical in late adolescence; an important consideration in the training of future teachers. It was also a critical explicit assumption of the "model" of the instruc- tional situation used in this study. The discussion of teacher prediSpositions possibly relevant to encouragement of such behavior in students led to inclusion in the study certain factors previously Specified. One such factor suggested was a prediSposition given by students the title of "risk"127 and therefore so labelled in this study. However, the term as defined in the context of that suggestion did not pertain to the commonly used definition of actual decisions made under a "payoff" situation. Rather, it was defined in terms of cognitive 127Berry, loc. cit., p. 211. 102 judgments involved in a problem-solving stance, and consisted of two aSpects: first, the teacher exhibited tolerance of uncertainty, of inexactness, of ambiguity, avoiding the making of extreme and simplistic judgments to reduce his own uncertainty, and avoiding being insensitive to errors caused by such a simplistic approach; but in addition, where ambiguity and possible error were evident to all, the teacher seemed willing to make judgments which would be considered less conservative, less safe in terms of conforming to what the maj- ority viewpoint might be, without fear of censure and with acceptance Of being exposed as incorrect or unwise. This prediSposition has relevance for several reasons. First, Schwab has drawn attention to the reverse of the first of these aSpects -- that is, the tendency to reduce ambiguity and uncertainty by simplistic and extreme judgments and positions -- as a common characteristic among undergraduate students.128 Second, in his study Of elementary teachers, Jackson drew attention to the reverse of both of these aSpects, noting a not dissimilar simplistic tend- ency; that is, the teachers in their preactive phase of teaching tended to reduce ambiguity and uncertainty by extreme and simplistic 'positions of one to one causality, and where ambiguity and their pwn ‘possible error must be considered, to be more conservative in their judgments and Opinions, to preclude being exposed as wrong.129 Third, 128Joseph J. Schwab, College Curriculum and Student Protest, (Chicago: The University of Chicago Press, 1969), pp. 3 - 36. 129Jackson, lpp. cit., pp. 143 - 147. 103 Shulman noted that those elementary student-teachers who were classed as dialectic (more likely to adOpt a "problem" approach) evidenced greater willingness to take risks on a test Of logical reasoning.130 The measure the underlying basis of which most closely approximated this factor as defined in the context of this study was the measure of judgment extremity and confidence develOped by Brim and by Kogan and wallach in their investigation of the degree of extremity exhibited in rendering probability judgments about ambiguous external events, subject to influence by the mechanism Of tolerance for versus reduction of uncertainty. A basic point in the formulation of Kogan and wallach is that cognitive-judgmental tasks ostensibly deal with problem-solving performance as Opposed to decision-making procedures, and that the risk element is more or less covert, "emerg- ing implicitly in terms of the strategy the subject employs". Since no one tells the Subject he has been correct or incorrect, the risk element is based on the subject's assessment "of his own tolerance for error".131 "When judgmental extremity serves as an expression of tendencies toward reducing general uncertainty in ambiguous situations, then the individual is not sensitive to the greater potential for error that lurks in extreme judgments. If judgment extremity does not serve uncertainty-reduction, then a person is more likely to respond as if aware of the greater chance for error that extremity entails. Under Such conditions, greater judgmental extremity characterizes persons who 130 Shulman, loc. cit., p. 188. 131 Nathan Kogan and Michael A. wallach, Risk Taking: a Study in Cognition and Personality, (New YOrk: Holt, Rinehart and Winston, 1964), pp. 2-7. 104 take greater risks than persons who are more conservative in pay-off situations."132 The instrument consists of fifty items requiring judgments about the likelihood of variOus events, each so ambiguous as to prevent Specification of a correct answer. After making each judgment the subject then Specifies his level of confidence in his judgment. The theoretical basis for the development of the measure and its role in their investigations are fOund in Kogan and wallach, Risk Taking: a Study in Cognition and Personality.133 A second prediSposition with potential importance to the instructional process appeared to involve another conception of conservatism; that is, the extent of the teacher's conservatism rela- tive to a number of attitudes held, which, while dissimilar in some aSpects, seem to have an underlying common base in terms of the approach taken toward the "job" or learning process in the classroom. As with the previously discussed conception of extremity and of degree of conservatism in judgments, this conception of "conservatism" would seem to be important to an instructional process based on aSSumptions Of the "problem" or "inquiry" approach -- in instructional situations in general, in particular in instructional situations involving young adults trying to make the transition from the conception of knowledge as the quantitative accretion of discrete rightnesses to the concep- tion of knowledge as qualitative assessment of contextual relation- ships, and Specifically in instructional situations involving future teachers who themselves eventually will be enmeshed in a relationship 1321bid., pp. 197-198. 1331bid. 105 with students in the learning process. In his study of elementary teachers on the job, Jackson has drawn attention to one aspect of this conservatism in what he calls "pedagogical conservatism" -- that is, "an accepting attitude toward educational conditions as they presently exist, with interest in educational change typically restricted to ideas about how to rearrange the room or regroup the students. This acceptance of the status quo appeared to be a part of the general myOpia typifying the classroom teachers' intellectual vision."134 Examples illustrative of how this "conservatism" or acceptance of the status quo Operates in various areas of the "job" or learning process in the classroom can be derived from the literature. An example from one area is reflected in the attitude taken toward the entire concept of encouraging stu- dents to think not just convergently but divergently, of less rote memory for memory's sake. In discussing the failure of this concept to become a reality in the classroom the Panel on Educational Research and Deve10pment pointed out that while lip service is given to this by teachers, such teaching has to be "something more than (students) answering intelligent questions intelligently; it is creating the situations in which intelligent questions are likely to be asked (by students)," requiring an overhaul of the teacher's established patterns of teaching, a commitment most seem reticent to make.135 An example from a second area is reflected in the attitude taken 134Jackson, loc. cit., p. 148. 135Panel on Educational Research and Deve10pment, Innova- tion and Experiment in Education, (Washington, D. C.: U. S. Govern- ment Printing Office, 1964), p. 6. 106 toward the focus of "control" in the classroom. While teachers reject the stereotype of the teacher-autocrat and eSpouse a more egalitarian approach to control, nonetheless "control" in the majori- ty of classrooms still appears to mean power and a teacher-centered learning process, ranging from authoritarian discipline to a less anti-democratic approach but one still reflecting reluctance to provide for variation in behavior which deviates from set routines.136 An example from a third area is reflected in the attitude taken toward the importance of the educational process as a means of reflecting and preserving "those wishes and values that are to be found in the mainstream of the pOpulace."137 Any number of possible consequences of stressing preservation of the status quo in this regard can be listed. For instance, one such is that sometimes consciously but more often seemingly unconsciously there results in what might be termed an "ethnocentrism" in the designation of Students as members of "ingroups" and "outgroups" on the basis of how visibly the parti- cular individual manifests these desired values. Studies of Such groups range from those investigating the effects of the seeming easier acclimatization of young girls than of young boys to life in the classroom, to Studies of disadvantaged Students about whom the teacher, in working with them with the best of intentions, nonetheless tends to have a set of negative expectancies about their behavior which 136Herbert M. Kliebard, "The Observation of Classroom Behavior," in The Way Teaching IS, Association for Supervision and Curriculum DeveIOpment and National Education Association, (Washing- ton D. C.,1§966). Michael A. wallach and Nathan Kogan, Modes of Thinking in Young Children, (Néw York: Holt, Rinehart & Winston, 1965), p. 319. 107 often appears to act as a self-fufilling prophecy.138 A second consequence of the tendency to stress preservation of the Status quo in terms of attitudes, values, and behavior is that in the dilemma of whether "education will function as an arm of the majority in our society, reflecting those wishes and values that are found in the mainstream of the populace, or function as a minority voice, holding the general culture up to the mirror of constant appraisal," the former triumphs at the expense of the latter.139 This dilemma extends the concept of "pedagogical conser- vatism" -- that is, an accepting attitude toward educational condi- tions as they presently exist -- to a second aSpect, that of a general conservatism in attitudes -- that is, as a component of the teacher's lifestyle. Such historians as Richard Hofstadter, in writing of the role of conservative beliefs in the lifestyle of individuals, have alluded to this prediSposition in teachers' lifestyles Spilling over into teaching Stance, tending to result in teachers as a group being seen as followers rather than as in the vanguard, Showing the way to 140 "the mainstream of the populace." Reports Such aS that by the Office of the United States Commissioner of Education on the State of 138E. Riessman, "Teachers of the Poor: a Five-Point Plan," Proceedings of the 17th Annual State Conference on Educational Research, (Burlingame: California Teachers' Association, 1965), Mimeo. 139Wallach and KOgan, loc. cit., p. 319. 140Richard Hofstadter, Anti-Intellectualism in American Life, (New York: Alfred A. KnOpf, Inc., 1963), and The Paranoid Style in .American Politics and Other Essays, (New York: Alfred A. Knapf, Inc., 1967). 108 the education professions, under terms of the Education Professions DevelOpment Act, support this contention.141 Such reports point to the general all-pervasiveness of these attitudes, extending into various aSpectS of the lifestyle such aS an ideology of political conservatism -- at its most liberal, an ideology at the extreme conservative end of the liberal Spectrum. The Study of the Carnegie Commission on the current Status of education at all levels indicates that such predispositions are not confined to elementary teachers but are also a distinguishing feature at the secondary level.142 This conception of "conservatism" appears to have import- ance for a number of reasons. Initially, it has relevance at all educational levels. First, while exceptions can be cited as justi- fiable refutations of this conception, the literature Suggests its existence as a common generalized prediSposition. Second, an individual's prediSposition to this "conservatism" Should not neces- sarily be assigned a negative value. Rokeach points out the danger in placing a positive value on change and a negative value on non- change.143 For example, there is positive value in preservation of 141Office of the United States Commissioner of Education, report on the State of the education professions, authorized under the Education Professions DevelOpment Act, Annual Report, 1967, United States Office of Education. 142Study of the Carnegie Commission, discussed in Crisis in the Classroom. Charles E. Silberman, (New York: Random House, 1970). Results of the Study suggest that Such prediSpositions are characteristic of all levels of educational instruction. 143Milton Rokeach, The Open and Closed Mind, (New York: Basic Books, Inc., 1960), p. 10. 109 certain values and beliefs deemed important by society (although which ones Should be so preserved is open to debate); the question centers on the emphasis placed on this as a goal. Further, review of the literature in chapter two indicated that to be prediSposed to be more authoritarian or to be more teacher-centered does not necessarily infer lack of concern for Student welfare, or that more student-centered approaches necessarily improve Student performance or attitudes; how- ever, the question iS not the degree to which Such an atmOSphere may be repressive but the degree to which it may be oppressive and soporific. Third, the relevance of this conception of "conservatism" does not rest on what Levinson terms "reactionary conservatism" typi- fied by assertions of the need for a drastic change to turn back the clock to the old verities;144 rather, the relevance of this concep- tion rests on what Dewey terms "teachers as Students of teaching". 'The accepting attitude toward conditions as they presently exist :suggeStS a tendency not to initiate change, but not necessarily to be aadverse to change. Dewey's fear is that when a change is adopted, Ilabels Such as "progress" may mean only perfecting and refining Skills alt this new level, achieving a new plateau of acceptance of the Status <1uo, devoid of initiative and a reflective approach about new ways to juxtapose seemingly dissimilar ideas about teaching.145 Second, this conception of "conservatism" seems to have 144T. W;‘Adorno, E. Frenkel-Brunswick, D. J. Levinson, and R. Nevitt Sanford, The Authoritarian Personality, (New York: Harper and Bros., 1950, PP. 151-182. 145John Dewey, "The Relation of Theory to Practice in Educa- tion," in M. L. Borrowman (ed.), Teacher Education in America, (New York: Teachers' College, Columbia University, 1965). 110 particular relevance at the educational level of adolescence and young adulthood where, as Piaget suggests, the student is trying to make the transition from a world assessed in terms of discrete entities to a world assessed in terms of qualitative contexts, and is working at "injecting himself into adult society...by means of projects, life plans, theoretical Systems, and ideas of political or social reform".146 In his Study of this period of develOpment, Perry pointed out that in identifying those teachers from whom they felt the greatest sense of Support in their attempts at this transition, Students pointed to not only the prediSposition to express judgments which might deviate from the Status quo, willingly exposing the self to criticism rather than equivocate, but also the prediSposition of "openness" --a visibility of the teacher's own doubts and grOpings, and of his commitments in education and lifestyle.147 Third, this conception of "conservatism" appears to have special relevance at the educational level involving Students who are themselves to be future teachers. The review of the literature :ln chapter two pointed out that differences in attitudes among under- graduates have been Studied largely in terms of positions regarded .aS "liberal" or "conservative", and that with some consistency the Inost conservative groups are in applied rather than academic fields. Students in elementary and physical education tend to be among the most conservative groups. Those in secondary education reflect the 146Jean Piaget, Six Psychological Studies, (New York: Vintage Books, 1968), pp. 60-73. 147Perry, 10c. cit., p. 210. lll attitudes of their academic Subject fields, although as a group perhaps more conservative than those in their fields not working toward teacher certification.148 Further, in his Study of the inquiry process among elementary education majors, Shulman found that among the determinants of those prediSposed to inquiry, one determinant was a tendency to be more politically liberal. The politics Score was derived from three sources: political party preference of the Student's parents, the Student's own political identification on a Scale from conservative to liberal, and Student rank ordering of preference for four possible presidential candi- dates. Shulman pointed out that the results seem to lend support to "the notion that the same dynamics that underlie choice of poli- tical and social values predispose one to dialectical cognitive functioning149 -- a finding congruent with the work of Harvey, Hunt, and Schroeder."150 Fourth, this conception of "conservatism" would seem to Thave particular importance to students in the teacher-education course involved in this Study. One assumptions of the course "model" «iiscussed in chapter two was that problem-solving thought process (activities were to take place, and were Specifically a goal of the discussion section. Among the teaching assistant's roles in the discussion section was not only the role of assisting students in 148Bereiter and Freedman, loc. cit., p. 568. lagShUIMn, 10cc Citc’ pp. 86 - 88' 1500. J. Harvey, D. E. Hunt, and H, M. Schroeder, Conceptual Systems and Personality Organization, (New York: Holt, Rinehart, & Winston, 1963), as reported in Shulman, loc. cit., p. 88. 112 their attempts to engage in such activity, but also the role of teaching assistant as contributor to develOpment of Student attitudes toward their future as teachers, and also the role of teaching assis- tant as model -- presumably both as model of "attitudes, skills, and understandings which characterize a competent teacher," and as model of a stance taken toward the activity in which the students themselves were being asked to engage in the discussion section. A number of considerations were taken into account in the choice of the measure to serve as a proxie for the conception of "conservatism" aS discussed in the previous pages. One considera- tion, mentioned earlier, was a time factor, this measure being only one of several requiring the time of the teaching assistants, already hard-pressed. A second consideration was that Since this measure was only one of several representing factors considered potentially important, this necessitated confining each proxie to a Single measure to avoid overloading the analytic model. Third, aS Rokeach points Tout, alternate ways of thinking about change, ingroups and outgroups, ;authoritarianism, and so on means that the particular way Such factors tare thought about involves implicit value judgments, influencing the :aperational definitions employed, and their interpretation.151 Of interest in this Study was the conception of a set of attitudes forming a stance or prediSposition reflected in a defined "conservatism" in lifestyle, an extension of which was reflected in various aSpectS of "pedagogical conservatism" in the classroom. The 51 Rokeach, loc. cit., p. 11. 113 Single measure which seemed most approximate as a proxy for the various aSpectS of the above defined conception of conservatism was the F-Scale. Rokeach has criticized use of the Scale in Studies of authoritarianism outside those of the developers of the scale on grounds that "the F stands for fascism and the scale was designed (first) to be an indirect measure of prejudice without mentioning Specific minority groups, and (second) to measure underlying persona- "152 However, lity prediSpositions toward a fasciStic outlook on life. although this had been the original point of inquiry, Adorno et al point out, in discussing the results of their studies,"....That we have achieved the second purpose underlying the F-scale -- to cons- truct an instrument that would yield an estimate of fascist recepti- vity at the personality level -- has yet to be demonstrated."153 Further, with regard to the first purpose, Adorno et a1 assert, "It seems that the F Syndrome is actually more closely related to general «ethnocentrism than to anti-Semitism," (which they used as an example caf prejudice as differentiated from ethnocentrism), indicating statis- tical relationships to this effect.154 In addition, in the discus- sion of the various related concepts in their body of work, the point is made that Fascism and Marxism as polarities on a right-left Scale 152Rokeach, loc. cit., p. 12. 153Adorno, Frankel-Brunswick, Levinson, and Sanford, loc. cit., p. 279. lSaIbid., pp. 264-265. 114 "do not find active Support on the American Scene," and necessitated consideration of conservatism and liberalism as the prevalent right- wing left-wing ideologies.155 Rokeach is justified in contending that the concept of authoritarianism cuts across, and is possibly independent of, ideo- logies, and he posits the greater potential usefulness of the concept of dogmatism as a measure of general authoritarianism.156 In this study, the conception of "conservatism" refers to an accepting atti- tude toward the Status quo, defined in terms of "pedagogic conserva- tism" and lifestyle conservatism", involving aSpects Such as conven- tionality and a tendency to resist or at least Show little iniative toward change, "authoritarianism" as reflected in the need for con- trol in order to maintain this status quo, and an "ethnocentrism" as reflected in a provincialism toward those who do not reflect these aSpects. The concept of dogmatism involves as a function those who are prediSposed to demand change, across all ideological dimensions, and less closely approximates the various aSpectS of the defined concept in this Study. In addition, in his Study of con- cepts with potential usefulness as determinants of prediSposition to inquiry, Shulman found that the concept of dogmatism as represented by Rokeach's Scales did not serve as a determinant of teachers who might be less inclined to engage in inquiry.157 To the extent that 155 Adorno et al, loc. cit., pp. 1-8, 150-180, 260-279. 156 Rokeach, 10c. cit., pp. 1-18, 121-131. 157Shu1man, loc. cit., pp. 43-92. 115 Rokeach points out that the California F-Scale meaSures "right authoritarianism (and is correlated with conservatism in ideology 8 the review of the literature indicates that and politics),15 teachers as a group have an over-all mean-item Score below that of other adults of similar Statu8.159 However, the inference iS not necessarily that teachers aS a group are at the egalitarian end of an authoritarian-egalitarian continuum or at the liberal end of a liberal-conservative continuum, but possibly that as Opposed to Levinson's "reactionary conservative" they more typify his concep- tion of either "true conservatism" or of "passive liberalism".16o In discussing use of this Scale in Studies of teacher attitudes Getzels and Jackson criticize Studies justifying the F- Scale through correlations with other attitudinal measures, compounding any acquiescence set, but they give credence to McGee's Study of the F-scale in relation to behavior observed in the classroom which indicated that teacher Scores were consistent with their classroom behavior, and that a positive relationship existed laetween this "measure of anti-democratic potential and a measure <>f teacher's overt authoritarian behavior in the classroom" ‘Various modifications of the Scale emp10ying indirect nonideological 158Rokeach, pp. 121-122. 123. gig. 159H.'M. McGee, "Measurement of Authoritarianism and its Relation to Teachers' Classroom Behavior, Genetic Psychol- Monogr., 1955, 52, pp. 89-146. 169Adorno et al, loc. cit., pp. 152-180. 161Getzels & Jackson, in N. Gage (ed.), 122..£££-: P: 523' 116 items have been used in Studies, Such as the Inventory of Beliefs develOping out of the work of Stern, Stein, and Bloom.162 In discussing their entire body of work, Adorno et 81 point out that the work was guided by the following major hypothesis: "that the political, economic, and social convictions of an individual often form a broad, coherent pattern."163 The F-Scale develOped as an extension of premises on which studies of ethnocentrism and con- servatism were based. The relationship of the F-Scale to the con- cept of ethnocentrism (defined as provincialism or rejection of the culturally unlike whereas prejudice involved feelings of dislike against individual groups)164 was previously discussed. The concepts of conservatism and liberalism had been previously investigated with regard to trends Such as support of the Status quo and resistance to social change. 165 As contrasted with the dogmatism Scale, Rokeach showed in this studyer the extent to which the F-Scale correlated with liberal-conservative political meaSures.166 Nine variables such as conventionalism, authoritarianism (submission and aggression), and power were derived and defined, and together were considered as a single F-Syndrome. The theoretical basis for their derivations and definitions, the Scale and its use in their studies are found in 162G. G. Stern, M. I. Stein, and B. S. Bloom, Methods in Personality Assessment, (Glencoe: Free Press, 1956). 163Adorno et a1, loc. cit., p. 1. 164Ibid., pp. 260-279. 16SIbid., pp. 153-207. 166 Rokeach, loc. cit., pp. 119-122. 117 Adorno et al, The Authoritarian Personality.167 A third prediSposition with potential importance to the instructional process was that of rigidity. Rokeach has drawn attention to the theoretical differences between concepts such as dogmatism and that of rigidity. While both refer to resistance to change, the latter more closely approximates the resistance to change of habits or sets. For example, a person is said to perform a task rigidly, not dogmatically or conservatively.168 This prediSposition would seem to have relevance to all educational levels involving even minimal interaction between student and teacher. For example, in his Study of elementary education student teachers, Shulman noted that degree of flexibility in behaviors such as capitalizing on individual Student characteristics as they arose during the instruc- tional process characterized Student-teachers classed as dialectic, as contrasted with those classed as didactic.169 In addition, it would seem to have particular importance in instructional Situations Specifically created to insure interaction between Student and teacher in a non-lecture approach to the instructional process. The measure the underlying basis of which most closely approximated this factor as defined in the context of this study was the measure of rigidity developed by Cough and Sanford. The Scale is comprised of twenty-four items the referents of which appear to be Specific tasks and habits, 167Adorno et a1, loc. cit. See also R. N. Sanford, "The Approach to the Authoritarian Personality," in J. L. McCary (ed.), Psychology of Personality, (New York: Logos Press, 1956). 168Rokeach, loc. cit., pp. 182-195. 169Shulman, loc. cit., pp. 53-148. 118 rather than attitudes as defined in the previous discussion. Bal- anced for rigidity-flexibility keying, it is now scale Fx in the California Psychological Inventory.170 A fourth prediSposition with potential importance to the instructional process was that of defensiveness as reflected in the need for social approval. Kogan and wallach in their studies of decision-making prOpertieS indicated the importance of this factor relative to the individual's image-maintenance, in which Such a defensive concentration contributed toward the individual's adoption of posture's consistent with this image relative to the nondefensive individual's more casual approach, allowing the latter to be more sensitive to the properties of the problem confronting him.171 This prediSposition would appear to have relevance at all educational levels in Such instances as in decisions concerning the "fate" of individuals, for example, grades (particularly the subjective compo- nent of Such), in the role of "authority" in a subject area and in Supposed professional skills (as differentiated from the previous dis- cussion of "authority"), and in the interactive relationship establi- shed with students. It would appear to have particular importance in instructional situations involving students attempting to alter their conceptions of learning in terms of "rightness" versus "quali- tative assessment" and who identify teacher "openness" in his own doubts and commitments as Supportive of their efforts (as earlier 1700, K. Buros (ed.), The Sixth Mental Measurements Year- book, (Highland Park, New Jersey,Gryphon Press, 1965). 171Kogan and wallach, 10C. cit., PP- 159'185- 119 discussed relative to Piaget and Perry), and eSpecially in the instructional Situation in this Study, a Situation specifically designed to give Students opportunity to engage in this activity. The measure chosen to represent this concept was the Scale develOped by Crowne and Marlowe to measure degree of reSponse set of self des- cription in a socially favorable light, and conceptualized by the designers as a meaSure of "need for social approval". "Item Style and content indicate the instrument has 'lie Scale' properties, and is useful as an index of 'defensiveness' -- an index of the tendency to deny personal traits that, although moderately undesir- able, are possessed by virtually everyone and to accept traits that are highly desirable but possessed by virtually no one."172 The theoretical basis for develOpment of the measure, the instrument, and Statistics pertaining to it are found in Crowne and Marlowe, "A New Scale of Social Desirability Independent of PSychOpathology'.‘1Z3 Additional prediSpositions which seemed to offer potential importance to the instructional process were extent of reticence to initiate interaction with Students, and extent of anxiety created by demands for personal commitment beyond that which the individual feels able to make. These factors would appear to be important at all educa- tional levels, but in particular at the level of instruction involving young adults engaged in thought processes discussed earlier in this 172KOgan and Wallach, 10c. cit., pp. 23-24. 173D. P. Crowne and D. Marlowe, "A New Scale of Social Desi- rability Independent of PSychOpathology," Journal of Consulting Psy- chology, 1960, VOL 24, 4, pp. 349-354. 120 chapter (and Summarized in the above discussion of social approval), who find teacher "Openness" and "willingness to commit self" sup- portive of their own gropings. In addition, in instructional Situa- tions Such as that in this Study, where "discussion approach" is the Specified format for the instructional process and eSpecially where the discussion section fails to qualify on any criterion as a small group, the conceptions of a more reticent versus a more out- going temperament, and of excessive anxiety and instability versus stability under pressure would seem to have particular relevance. The instrument which served as proxy for these conceptions was the 174 Its choice was Personality Inventory of Eysenck and Eysenck. based on two considerations: it is conceptualized to measure two major personality dimensions, extraversion-introversion and neuroticism-stability; and its use correSponded to its usage as a proxy for Specified Student characteristics. An extensive dis- cussion of the rationale for use of this Scale in this Study andcf information pertaining to the scale itself was outlined in the presentation of Student characteristics.175 Two additional instructor factors, pertaining Specifically to this instructional situation, were: first, the course load being carried -- that is, their academic load in purSuing their degrees; and second, previous teaching experience. In instructional Situations at all levels,extent of involvement outside the Specific instructional 174B. J. Eysenck and Sybil B. G. Eysenck, loc. cit. 1758ee pp. 99-100 in this chapter. 121 situation and extent of teaching experience are two criteria empha- sized by "evaluators" (Such as principals) as important to the success of the instructional process. In the particular instruc- tional Situation in this study, a course "model" based on a set of assumptions existed; a number of these assumptions pertained to the various roles of the teaching assistant, and were themselves based on the assumption that a decrease in faculty participation was feasible on the basis of projected teaching-assistant performance. A review of the extensiveness of these roles (see chapter two) would seem to indicate that the first of the above factors, course load, would have relevance not only in the initial phase of the course, but increasing relevance over time. The factor of previous teaching experience would seem to have particular relevance on the same basis -- with roles Such as model of attitudes, Skills, and under- standings possessed by competent teachers. Further, teaching experience had been an important criterion in the choice of teaching assistants. In addition, an instructor questionnaire was constructed composed of two parts: first, a series of questions concerning such factors as the instructor's graduate Studies, background in the Sub- ject area of the course, occupational aspirations, and some personal data (sex, age, marital Status, and so on); second, a series of questions, open-ended, involving instructor value judgments concerning the instructional situation, relative to what the instructor perceived the course model to be (that is, the aSSumptions and objectives under- lying the course), the degree of congruence between these assumptions 122 and the "real thing", his own objectives for the discussion section, his methods and procedures for achieving these objectives, his perception of the nature of the behaviors Students should be able to exhibit as a result of Successful completion of the course, and perception of how the course could be improved, and Specification of his procedures for determination of students' discussion section grades. The instructor questionnaire was not part of the analytic model. No instructor names were involved; a number code was used to match instructor and students, the matching being done by other than the individual conducting this study. III. PROCEDURES Setting and Subjects The setting of the study was the first course in the undergraduate teacher-preparation sequence for future teachers, which can be categorized as a lecture-discussion group, multi-Section, survey-type course. The instructional content of the course included, secondarily, an introduction to the teaching profession and to the program of teacher education, and primarily an introduction to the subject matter of educational pSychology. Participants in the study included five hundred and thirty-two students enrolled in the course, fifteen graduate teaching-assistant instructors, and indirectly, eight faculty lecturers, one course coordinator and evaluator, and one teaching internship advisor for the teaching assistants. The course was conducted over a ten-week period, five days per week, fifty minutes at a time, at two different hours during the day at which time the same format was offered to the Students using the same 123 personnel. At whichever of the two hours designated at the beginning of the course, the Student met with all other Students in the course in one large lecture section by a faculty lecturer three days a week (Mondays, Wednesdays, Fridays), and two days a week (Tuesdays and Thursdays) met, along with the other twenty-five to thirty students in his discussion section, with his graduate-assistant instructor. Design for the Study From the discussion at the Start of this chapter concerning the instructional Situation, focusing on the instructional process or ”job", it was possible to derive several premises about the instruc- tional Situation. First, Students enter into the instructional situation with a set of characteristics which potentially may affect the instructional process and the instructional outcomes. Second, as the students continue in the course, certain events occur which may have importance to both the process and the outcomes. These events include a) contextual factors and b) reward and performance factors. Third, the course outcomes are various (and weighted), and contin- gent on a combination of entering factors and continuing factors. The structure of the design was develOped to approximate the course by extracting data at different points in time as the instructional process progressed. That is, the design Structure approximated the instructional Situation Structure, of factors which had importance at a given time to students in the instructional process, as it develOped. The extractions of data were made concern- ing two categories of factors: outcome factors and predictor factors. Predictor-factor data was extracted at three points, each time recombining the new data with the previous data extracted. That 124 is, predictor-factor data was extracted at the following three points: entering-course, mid-course, and end-of-course. Table 2 Summarizes the predictor factors concerning which data was derived. GPA Success expect- CQT ation after Personality midterm perf. Extraversion Accuracy judg- TABLE 2: SUMMARY OF STUDENT PREDICTOR FACTORS ENTERING-COURSE MID-COURSE END-OF-COURSE recombine recombine Personal Personality Personality Sex Midterm anxiety Final exam anxiety Age Attitudes Attitudes Major Course-Specific Post-attitude Credits earned motivation Discussion section Course load after midterm attitude Past Performance performance Anxiety ment of final Social acceptance grade, after Test anxiety mid. perf. Pretest anxiety Course Performance Attitudes Midterm Score Learning set Reason enrolled Pre-attitude Course-Specific motivation at pretest The terms "entering-course, mid-course, and end-of-course" designating extraction points are relative. That is, the terms refer to just before, along with, or juSt after the day of a given test which was part of the course -- Specifically, the pretest, the midterm, and the final. The measures concerning the instructor predictor factors (risk, "conservatism", rigidity, defensiveness and social approval, anxiety, extraversion, course load, and teaching experience) were given, along with the instructor questionnaire, between the entering-course and mid-course extraction points. 125 Chapter one presented Campbell's discussion of the diffi- culties in extracting data unobtrusively in the natural setting, and of the Strategy of using data already available, data derived as an ongoing part of the instructional Situation, and data obtained from intrusions as logically a part of the natural setting as pos- sible. Certain data was obtained from sources already available, Such as the OQT-Scores, derived from the entering-university testing program. Student evaluation of the discussion section experience was derived from an all-university procedure of the natural setting, that of evaluating courses. A procedure not uncommon in various courses as part of professorial license was to ask for personal information concerning Student major, sex, level (here represented by credits completed), and course load; in this instance, obtained along with GPA on the face-Sheet of a measure in a battery discussed below. Some brief measures, one or two questions, not normally part of the course Structure appeared along with measures normally part of the instructional process -- for example, data of anxiety relative to a given Specific course exam was obtained as an addendum to that exam. Data from additional Similar brief measures was obtained in lecture sessions -- materials handed to Students entering, completed in the brief time prior to the lecturer's appearance, kept during lecture, tossed into barrels on departure. The course post-attitude measure followed the taking of the final exam; because of the Scheduling of the exam in the evening with no other course exams to follow (a regulia practice of the course), students were able to Spend the extra few min- utes necessary. The entering-course factors represented by the pretest, the learning set Scale, the measure of general test anxiety, and of 126 extraversion and anxiety were administered in what could be called a "battery" form in the initial phase of the course. The pretest was a normal part of the course setting, the others were not, requiring two periods of course time. Table 2 (p. 124) indicates that following this initial "battery", any additional continuing data was obtained by the means outlined above. "Outcome" data was extracted at the same three points as was the predictor data discussed above. The first extraction included pretest-Score data; the second extraction, data concerning the midterm exam scores and grades; the third, data concerning final exam Scores and grades, discussion section grade, and final Summary course grade. These were part of the instructional Situation struc- ture. The purpose of the entire project, of which this study was a part, was outlined to Students the first day of the course. Stu- dents were assured: no names would be involved, only student number (all measures bore this number); no work would be done with the data until after their completion of the course; Student reSponseS were privileged information and no one in instructional or administrative roles in the course would be allowed to examine the individual stu- dent reSponseS. A question was included in the post-course assess- mentamking students to indicate the extent to which they felt the project had affected either the instructional process or their parti- cipation in it. ReSponses to this item indicated 43 per cent believed not at all, an additional 26 per cent believed very little if at all, 10 per cent did not know, 13 per cent felt perhaps some- 127 what, three per cent felt considerably, and five per cent were omits.176 Other than handling those measures normally a part of the instructor role, no additional participation was expected of the teaching assistants. All students enrolling in the course chose one of the two times at which the course was offered, as part of university procedures. Prior to the start of the course, students in either of the two time periods were randomly assigned (via a table of random numbers) to one of the discussion sections in that time period; on the same basis, teaching assistants were then assigned the sections. A second list of random numbers was generated to handle late additions to the course. Students dropping the course were replaced by late-enrolling Students. No section changes were allowed. Later draps from the course could not be replaced; however, the discussion sections terminated with approximately the same number of students in each section. Strategy of Data Analysis The data analysis paralleled the research design, involv- ing analyses represented by the three extraction points in the design. In chapter one the discussion of the purpose of the Study indicated that the analysis questions revolved around two major categories: a) instructional process; b) instructional outcomes. Interest cen- tered on what happens to the explanation of outcomes when data is added from the time perSpective of the research model. That is, 176J. T. Parmeter, loc. cit., pp. 101-102 128 given entering-course characteristics, the data appeares to be explained thus; given the addition of mid-course characteristics, then the addition of end-of-course characteristics, how does this appear to change? An extension of this point was the interest in the explanation of results as they occurred sequentially in the course. That is, what factors appeared to have the most importance to entering-, mid-, and final-course performance. Table 3 indicates the attempt to maximize explanation by looking at changes in struc- ture as well as changes over time. TABLE 3: MODEL OF COURSE STRUCTURE AND TIME SEQUENCE Analysis Extraction 1 2 3 I pretest final course (enter- grade ing char; II midterm Score final course (mid- midterm grade grade course) final exam-l final exam-2 . 111 final exam-T final course (end’°f' final ex gr grade course) disc. sec.gL. Each factor in Table 3 represents an "outcome" factor. With the exception of the pretest, each represents an "outcome" relative not only to course Structure but also to course instructional process. "Pretest" is included as an "outcome" on one dimension; that is, it was not an outcome influenced by the instructional process per se. Its importance as an "outcome" in the analysis rests on its contrast with the "outcomes" involving instructional process, when examined 129 in terms of the predictor factors. Table 3 (p. 128) indicates that the first-stage analysis involved two separate analyses and one set of predictor factors -- analysis of entering-course characteristics first to the entering- performance indice, that is, to the pretest; then second, to the final Summary course outcome, that is, to the final course grade. Of interest was not only what entering-course factors were of import- ance to pretest performance, but also what entering-course factors would appear to be of importance to final course grade without the contending contextual course factors develOping from the instruc- tional process. Table 4 lists the entering-course factors involved in the two analyses in this first stage. TABLE 4: PREDICTOR FACTORS IN STAGE I ANALYSES Personal Personality Attitudes Sex Extraversion Learning set Age Anxiety Reason enrolled Major Social acceptance Pre-attitude Past Performance General test anx. Course-Specific GPA Pretest anx. motivation CQT'S at pretest Table 3 (p. 128) indicates that the second-stage analysis involved three separate analyses and one set of predictor factors -- that is, analysis of all predictor factors extracted by mid-course (including a set of contextual factors recombined with those from the stage I analysis) first to midterm exam grade; second, to midterm exam score; third, to final course grade. Table 5 lists all factors to mid-course used in the analyses of the three outcome factors. 130 TABLE 5: PREDICTOR FACTORS IN STAGE II ANALYSES Personal Attitudes Sex Learning set Age Reason enrolled Major Pre-attitude *Credits earned Course-Specific *Current load motivation Past Performance at pretest GPA *Course Performance CQT'S Pretest score Personality *Instructor Extraversion Extraversion Anxiety Anxiety Social acceptance Defensiveness General test anxiety Social acceptance Pretest anxiety Rigidity *Midterm anxiety "Conservatism" *Not involved in stage-I analysis Table 3 (p. 128) indicates thazthe third-stage analysis involved six separate analyses. The Six outcomes involved included: final examination score-part one; final examination score- part two; final examination Score total; final examination grade; discussion section grade; final course grade. One set of predictor factors was used, composed of three parts: a set of factors which, following the midterm examination feedback, were considered to have potential importance to end-of-course outcomes (factors Such as their reconsideration of expectations for their course performance, their adjusted level of motivation, Specific additional instructor factors as course pressure increases in the instructional process, and so on) recombined with the factors considered to have importance initially in the course, and those develOping potential importance by midterm examination time (that is, recombined with all previous predictor factorSL Table 6 lists all factors to end-of-course used in the analyses of the six outcomes. 131 TABLE 6: PREDICTOR FACTORS IN STAGE III ANALYSES Personal Attitudes Course Performance Sex Learning set *Pretest Score Age Reason enrolled **Midterm score Major Pre-attitude Instructor *Credits earned **PoSt-attitude *Extraversion *Current load **Discussion att. *Anxiety Past Performance Course-Specific *Social acceptance GPA motivation *Defensiveness CQT'S at pretest *Rigidity Personality **Course-Specific *"Conservatism" Extraversion motivation **Risk Anxiety after mid. **Course load Social acceptance **Success expec. **Teaching exper. General test anxiety after mid. Pretest anxiety **Accuracy judg. *Midterm anxiety after mid. **Final exam anxiety *Not part of stage I analysis **Not part of stage II analysis In summary, the research strategy was to try to approxi- mate the theoretical course model through the design and analysis models -- that is, through the data gathered, the timing of the data gathering, and through the combinations of data used in the analysis stage. That is, the purpose of the study was to identify and evaluate;, in what could be considered the initial phase of research in a Specific ongoing instructional Situation, those instructional situation factors which seem most important to the Specified instruc- tional situation outcomes. Since the objective was confined to loca- ting and evaluating factors and identifying problems in the Structure and relations of these factors which would reduce the validity of Studies in this situation under more stringent conditions, the research design outlined in the previous section was so Structured as to extract from the instructional Situation as many as possible of 132 the factors which were judged to be "candidates for importance" to the instructional process on the basis of considerations derived from the literature and the assumptions of the instructional Situa- tion designers. That is, within the limitations of these considera- tions, an attempt was made, by inclusion of a considerable number of factors in the design, to maximize chances of including those with the most potential importance, on the assumption that it would probably be more difficult to correct problems arising from failure to see the importance of a factor than it would be to correct those arising from incorrectly including a factor. In addition, since interest was not in comparing instructional processes and their contingent factors but rather in maximizing the potentiality of the factors across the instructional situation, subjects were considered as a Single group. The assumptions of the study research design put both Specifications and constraints on the analysis Strategy that could be used. As indicated by the outline of the analysis design in the previous pages, the Strategy for the analysis paralleled that of the study design. Therefore, the Strategy of the analysis was in essence a first Stage in an inductive "model" building in this particular instructional Situation, as discussed in chapter one. That is, in the analysis of each Specified outcome the question of interest is: what must be known most in order to reduce predictive error the most; given the results, what additional information would help reduce this more. (The ultimate aim, of course, as Campbell and Stanley Suggest, would be that through a number of iterations the point of serendipity is reached in which the assumptions underlying the most stringent 133 models become tenable in complex instructional settings.) In addi- tion, through the Strategy of analysis in this instance, estimates of the explanatory power of the model Should be available relative not only to the total variation of a given outcome factor but also to the total explanatory power of given predictor factors. Moreover, the Study design put a number of additional constraints on the potential analysis strategy. For example, the variables extracted appeared in a range from classification to continuous; application of Statistical tests of significance under assumptions of a fully random sampling model was questionable at best; the probability of the presence of complex interactions could not possibly be assumed away (the presence of which, according to Stanley, Should not only be revealed, but identified and located.) Under the considerations in the previous paragraph, the analysis strategy adapted was an analysis process which accounted for variance in a Specified outcome factor by Optimal Splits of a Specified set of predictor variables. Paralleling the Study design, the analysis strategy also, in the beginning, regarded Subjects as a Single group. This initial group of Subjects was then divided, through a series of binary Splits, into a mutually exclusive set of subgroups, the subgroups chosen so that their means accounted for more of the total sum of Squares (that is, reduced the predicitive error more) than the means of any other set of Subgroups. Briefly summarizing the entire strategy: the first decision is what single division of the initial Single group into two groups (on the basis of the most important predictor at that point) will do the most good in reducing predictive error (that is, providing the largest reduc- 134 tion in the unexplained sum of squares; the next step is to determine which of the two groups of Subjects thus created has the largest remaining error Sum of squares and should be looked at next for possible further subdivision (the other group being a possible candidate for examination later in the analysis). Whenever a further subdivision of a group will not reduce the unexplained sum of Squares in that group by a Specified percentage of the total Sum of squares (usually at least one per cent), the particular group Splits no further and is regarded as a final group. If, however, the group does meet the criterion, and is still a candidate for subdivision, the group must be composed of at least a Specified number of subjects (uSually a minimum of twenty, to provide Stability); if not, the subgroup is regarded as a final group. If both these criteria are met, the group is still a candidate and must meet the criterion of accounting for a Specified percentage of the original sum of squares, (one-half per cent reductionlin error before a Split is allowed). If the criterion is not met, said group is regarded as a final group. When no groups exist to meet these criteria, the final set of Sub- groups exists, with each subject in the original group now a member of only one subgroup; that is, a final Subgroup. The detailed theo- retical considerations underlying the Strategy, the statistical algor- ithm, the rationale for establishment of the parameter setting, its use in a number of Studies, and its development over time have been documented by Sonquist and Morgan.177 177J. A. Sonquist and James N. Morgan, The Detection of Inter- action Effects, (Ann.Arbor, Michigan: Institute for Social Research, 1964) Pp. 5-1400 135 Sonquist points out that in an initial analysis where the purpose of the data is not only to locate and evaluate potentially important factors but also to produce information useful in determin- ing the extent to which problems exist in the data and what variables are involved, parameter settings for the criteria are indicated which will insure that the maximum of information is extracted from the data.178 Thus as few constraints as possible should be placed on the partitioning process, the parameters set to permit the creation of as many final groups as is feasible. "In addition to providing a maximum 179 amount of information, the total explained variation provided by this analysis estimates the amount that could be explained by a con- (1."180 For purposes of this figuration model of the data being use analysis, stopping criteria were purposely relaxed slightly in order to provide as many leads as possible: the criterion relative to total sum ofsquares before a group is examined was set at .0001 (a Specification used by the designers of this strategy); minimum group size was set at twenty,. “and the criterion relative to sum of Squares before a group is examined was set at .006. Each of the outcome factors was also examined on the basis of the specified set of predictor factors (as outlined in the design for analysis) using multiple linear regression analysis. These 178J. A. Sonquist,'Mu1tivariate Medel Building? paper presented at the Conference on Multivariate Medels for Data Analysis in the Social Sciences, Institute for Advanced Studies, Vienna, Austria, September 26/27, 1969, pp. 1-37. 179By "explained" is meant that the means of the two halves are used for predicting rather than the over-all mean. 18oSonquist and Mergan, loc. cit., pp.22-23. 136 analyses were performed to provide information relative to the points raised in chapter one concerning the use of this strategy with Such data, and were not part of the design of the Study. There- fore, results of these analyses appear only in summary form in the following chapter, and are referred to in the text only in regard to points raised in chapter one. IV. EXPECTATIONS AND IMPLICATIONS The supposition has been put forward that within any given instructional Situation there is, at any given moment, an overall "job" going on, and this "job," represented by the instructional process, is to reach some objective or goal. That is, there is some overriding goal (which may be defined by subgoals) to which all factors in this given situation are geared. The particular foam taken by the instructional process, therefore, is defined by these factors. A further supposition is that in order to understand this process it is necessary to take into account several considerations. First, underlying this instructional process is a "model" -- a set of assumptions (held by those creating the instructional situation) about the nature of the instructional process, and therefore about the nature of the factors defining the process. Second, while the form of the instructional process is defined by the factors, Specific dimen- sions of these factors may carry more relative importance or weight in this definition. Third, at any given moment, the instructional process is a process of balancing two components -- that which is goal-oriented, and that which deals with pressures that have arisen in instructional situation at that moment. Fourth, different types 137 of subgoals suggest that for each subgoal a different combination of factors involved in the instructional process will explain Student performance in the achievement of that subgoal. Fifth, as the instructional process develops, the interest is not only on what Students have in common but also on how they vary. In this Study, the purpose iS to attempt to unravel the component parts of a given instructional situation by exploring the structure of relations among potentially important factors involved in the instructional process, and the conditions under which these relationships occur, in order to determine the relative importance of these factors to course criteria and outcomes. The concept of "impor- tance" refers to the weight or usefulness of a given factor in this situation, defined by four characteristics: amount, location, time, and criterion. The considerations underlying the concept of "the instruc- tional process" and those underlying the concept of "importance" suggest a number of expectations about the structure of relations among the factors at any given moment and over time, in the particular instructional Situation used in this study. 1. Coalescence will occur around certain classes ofypredictor variables inytheir description of a given outcome variable. the coalescence differing according to conditions surrounding the various outcome variables. This first expectation relates to the contextual character- istics which surround a given outcome -- that is, the nature of the particular outcome, and the pressures upon the instructional process at a given time. The two predominant contextual characteristics in the instructional situation in this Study which Should result in 138 coalescence of predictor factors, but differing in nature, are: first, the dimensions of competition versus non-competition; second, conditions relative to different points in time throughout the course. For example, the pretest as described by choice of items and means of administration is Similar to the exams determining final course grade. However, the nature of the two outcomes (pretest and final course grade) on the dimension of competition leads to the expectation that coalescence would occur around different classes of predictors in explaining the outcomes. In the same sense, Specific factors arising from the pressures within the context of the instructional process (for example, anxiety Specifically toward the midterm exam) lead to the expectation that the structure of relations among factors present early in the course would be altered over time. Therefore, the Specific questions to be considered in this study are: 1. IS the structure of relations among classes of entering-course variables similar for the pretest (noncompetitive) and the final course grade (competi- tive)? 2. IS the Structure of relations between classes of predictors of the final course grade Similar for different points in time -- entering-course, mid- course, and end-of-course? 2. The reSults will Show differences amongythe outcome variables Qppessed student_performance) in the type of predictor variables accounting for variance and in the extent of the explanation of variance. AS compared to the previous expectation which was concerned with differences in explanation attributable to differences in contex- tual conditions, this expectation is primarily concerned with explana- tory differences between outcomes in terms of the inherent tasks underlying the outcomes. First, it is expected that there will be 139 overall differences in explanation among the outcomes because of the differing terms of each analysis (for example, the overall vari- ance of the outcome variable, and the number of predictor variables included in the analysis), Second, it is expected that comparisons between analyses reasonably comparable in terms of overall variance in the outcome variables and identical in terms of the number of predictor variables, but different in terms of the underlying task, will result in different types of predictors emerging and in the extent of variance explainable by the different predictors. More Specifically, it is expect that there will be broad differences in explanation between the final exam Score-l, primarily a content-recall exercise, and the final exam score-2, an applied, problem-solving exercise. Another comparison which should illustrate differences attributabb:to underlying task is that between final exam grade, representing students' preparation for the exam in combination with other factors, and the instructor grade, representing the students' involvement and performance in the discussion section over the length of the term, in combination with other factors. Therefore, the Specific questions to be considered are: 1. What accounts for the largest amounts of variance across the range of the different outcome variables? 2. Are there differences between the levels of the dif- ferent outcome variables in: a. type of predictors accounting for variance? b. the extent of prediction by both number of splits and amount of variance explained? 3. Are there differences between outcome variables Similar in analysis conditions, but different in underlying 140 task, Specifically: a. between the final exam Score-l (recall) and the final exam Score-2 (applied)? b. between the final exam grade and the instructor grade? 3, Measured factors may together represent a weighting of their underlying theoretical constructs in the form of an interaction. Almost monotonously the literature reporting investigations of the instructional setting cite Significantinteractions which were either uninterpretable or for which there was no provision allowed in the theoretical model. The reality of the instructional setting is such that interactions Should be expected, and may, in fact, be the ideal result of an instructional Strategy. If the instructional outcomes result in interactions which improve the performance of students off the expected, then those interactions are highly desir- able. Given the number of Students involved in the instructional situation under study, the highly heterogeneous nature of these Students, the large number of instructional personnel, and the exten- sive number of objectives to which the course was directed, the likelihood of interactions occuring is even more probable. Thus for this study, a number of interactions would be expected and because of the lack of homogeneity and independence involved, it would be expected that of the interactions which emerge there would be varia- tions of those dimensions: (1) interactions involving levels of stu- dent characteristics to course outcomes and interactions involving levels of instructor characteristics to course outcomes; (2) the inter- actions involving positive instructional effects (results greater than expected) and negative instructional effects (results leSS than 141 expected); and (3) the interactions involving different degrees of interactiveness -- that is, interactions might emerge independently of multicollinearity. Because interactions were expected in the study and made a part of the analysis strategy, they will be used throughout the description of results, and therefore, the following questions will be directed only toward examples of the above dimen- sions: 6. Are there examples of interactions between instructor characteristics and course outcomes? Are there examples of interactions between student characteristics and course outcomes? Are there examples of positive interactive results (students performing above the expected) on course outcomes? Are there examples of negative interactive results (Students performing below the expected) on course outcomes? Are there examples of interaction with no collinearity present? Are there examples of interaction combined with collinearity? 4. Certain instructional process variables operate differentially as facilitators or barriers to student performance in meeting course criteria and outcomes. From the preceding discussion of interactions, one of the dimensions developed was the quality of the effect of the interaction on Student performance -- that is, positive interactions as the resultof performance above the expected and negative interactions as the reSult of performance below the expected. That discussion was but part of a larger discussion directed solely to the nature of interactions, but the idea of facilitators and barriers is significant enough to merit a concern by itself. In terms of the instructional setting as a 142 means to course goals, it is expected that there are a number of different factors which might act either as barriers or facilitators Of student performance beyond those that just appear as interactions. Certain factors are obviOus and consistent. For example, the nature of the overriding conditions to which the instructional setting was geared was strongly competitive. Therefore, the role Of factors associated with competition in similar situations, Such as the role of the grade-point average to course outcomes,would be expected to operate consistently across all levels of Students over time. However, it is also possible that the idea of barriers and facilitators to performance is more selective than the obvious, consistent predictors. Instructor characteristics possibly operate aS such in different ways. It is also possible that student characteristics may Operate selec- tively across different ranges of past and present performance to increase or decrease outcome results. In addition, part Of the rationale behind this Study recognizes the importance of both time and type of Outcome as important contributors to differences in Student performance. As an extension of this rationale, it is possible that certain factors work selectively with reSpect to either time or type of outcome is facilitators or barriers. Therefore, the Specific questions to be considered in relation to this expectation are: 1. What different factors Operate consistently across all levels Over time as facilitators or barriers to student performance? 2. DO instructor characteristics differentially Operate as barriers or facilitators to student performance? 3. Are there differing sets of predictors that act as barriers or facilitators for the different ranges of student performance in the course? 143 4. Are there differing sets of predictors that act selectively as barriers or facilitators to student performance at different points in time? 5. Are there differing sets of predictors that act selectively as facilitators or barriers to Student performance according to the type of course outcome? 5. The results will Show isolated exceptions. normally relegated to measurement or Statistical error. which rgpresent either important exceptions to general prediction.yor real instructional problems. This expectation is an extension of the concept of charac- teristics as barriers/facilitators, discussed above. In this instance, the expectation is that there will be isolated exceptions of Small groups of Students on the barrier-facilitator dimension which, unlike other small groups that might be considered created by chance fluctua- tion, appear consistently. Therefore, the Specific questions to be considered are: 1. Are there Small groups which represent actual exceptions to general prediction when compared with other Students on the same dimension of this characteristic? 2. Are there small groups which appear to represent real instructional problems? 6. To some degree the data will provide an estimate of congruence between course goals and conduct of the course. The goals and the assumptions underlying the processes by which the goals were to be achieved were discussed in chapter two. Condensed here, there appear to have been three major goals. First, educational Opportunity was to be a part of instruction so that entering characteristics of students would not be the main determiner Of degree of instructional impact. The implementation of this goal was based on three aSSumptions: a) the capability of the teaching assistant to make discriminations on student heterogeneity and to provide Opportunities for Students to overcome barriers created by differential entering characteriStics; b) common exams would not 144 create an atmOSphere based mainly on competition for the high exam score; c) TA grades would make discriminations on important variables that common exams did not discriminate upon. The second major goal was that the course was to be judged on the basis of the learning and "Schokarship" generated among students by it. The implementation of this goal rested on the following assumptions: a) the content provided by the faculty from Speciality areas could be used by Students to organize an approach to educational pSychology and problems in the classroom; b) the list of competencies to be develOped by this instructional process and expressed as a set of objectives Opera- tionalized this major goal; c) teaching assistants possessed the skills to help students to integrate diSparate lectures and readings in a problem-solving approach to future teaching experiences. The third major goal embodied three aSpectS: a) Student-teacher interac- tion was to be a main consideration; b) not only quality instruction but also the teacher as a model were of primary importance; c) an Objective of the course was to contribute to develOpment of attitudes characterizing a "competent" teacher. The implementation Of this goal rested on the aSSumption that a decrease in faculty participation was possible on the basis of projected TA performance in these areas. To get at course congruence necessitates both a reintegration of the results correSponding to the previous questions and a further examination of the data, as applied to the following questions. 1. was educational Opportunity a part of instruction SO that entering characteristics Of students would not be the main determiner Of degree of instructional impact (performance Outcomes)? 2. To what degree did the course prepare the students to 'use disparate information in an integrated approach to 145 solving classroom problems? 3. To what extent did the course provide models Of the instructional behavior Suggested as goals for students? Summary Within any given instructional situation there is a "job” being performed, represented by the instructional process. That is, there is some overriding goal to which all factors in this situation are geared; therefore, the particular form taken by the instructional process is defined by these factors. .In order to understand this process it is necessary to take into account several considerations. First, there is a "model" -- a set of assumptions -- underlying the instructional process. Second, while the form of the instructional process iS defined by the factors, Specific dimensions of these fac- tors carry more weight in this definition. Third, the instructional process involves the balancing Of two parts -- one part Specifically goal-oriented, the other part dealing with pressures that have arisen in the instructional Situation at a given moment, affecting goal progress. Fourth, the impact of various subgoals must be taken into account. Fifth, as the instructional process develOpS, the interest is not only on what Students have in common but also on how they vary. In this study, the purpose is to attempt to unravel the component parts Of a given instructional situation by exploring the structure of relations among potentially important factors involved in the instructional process, and the conditions under which these relationships occur, in order to determine the relative importance of these factors to course criteria and Outcomes. The concept of 146 "importance" refers to the weight or usefulness of a given factor in this situation, defined by four characteristics: amount, location, time, and criterion. The above considerations necessitated including in the analytic model not only potentially important entering-course factors but also those which over time play a role. In choosing those fac- tors to be included from the total potential set of factors in the instructional Situation, it was necessary to consider not only recom- mendations from previous findings and theory but also the "model" of the instructional Situation. The review of the literature also suggested that the core of any given instructional Situation is the interactive relationship between student and teacher. Therefore, factors in the analytic model were confined to those from the dis- cussion section which reflect potentially important Student charac- teristics, and potentially important instructor characteristics relative to their aSSumed roles in this particular instructional Situation. Factors concerning students were chosen from the areas of performance, attitudes, personality, and personal characteristics. Prior knowledge and ability were reflected by a number Of factors: past academic performance in instructional situations with an underlying institutional Similarity to that of the instruc- tional situation under study, represented by Student gradepoint average; past performance in situations ostensibly indicating poten- tiality of general intellectual ability, not only verbal but also conceptual ability, measured by OQT-Scores; past performance on content 147 with a possibly underlying Similarity to content which is a part of the instructional process in this study, and measured by the pretest Score; performance on content covered early in the course as related to later performance in the instructional situation, and measured by score on the midterm. Various specific aSpects of attitudes were reflected by a number of factors included in the Study. One set of attitudes reflected qualitative judgments Of students about the instructional process itself; this set of attitudes was represented by three factors -- reason for enrolling for the instruction, Student pre- and post-course attitude and motivation toward the instructional Situation, and student attitude toward the discussion experience Specifically. A second set of attitudes pertained to the aSpirations and expectations students hold for themselves in terms of their per- formance. A third set of attitudes pertained to academic "set" or preference for factual or conceptual learning. Specific personality factors were explored with which, in interaction, Students and teacher could logically be expected to come to grips in the instructional process. The first set of factors reflected anxiety related to academic situations, both in general test-taking Situations and those specifically connected with test- taking in this instructional Situation. The second set of factors was chosen on the basis of potential relevance to performance Speci- fically in the discussion section. In a "discussion" section which really does not qualify as a small group, reticence or recalcitrance to make oneself known and heard was considered a factor with potential importance, as was a high degree of anxiety, generated by demands for personal commitment beyond that which the individual feels able to 148 make. Additional factors included were the personal factors of sex, age, major field of interest, credits earned academically, and current academic course load. Instructor characteristics were chosen on the basis of the assumptions of the teaching assistant's role in the course, and associated suggestions from the literature. The various instructor roles stressed the importance of the instructor as a model of predis- positions toward a ~critical-thinking, inquiry approach to issues raised through the course content and its application. Results of studies cited in chapter three Suggested: the importance of such thought processes in the development of cognitive functioning in children; the cruciality of Similar processes during development taking place in late adolescence and young adulthood; the importance of such prediSpositionS in Students who themselves are eventually to assume the role Of teacher in the instructional process (Such as the particular students in this Study); Suggestions pertaining to certain prediSpositionS characteristic of those teachers who offered the greatest sense of support to Students attempting to engage in an "inquiry" approach to learning. This led to the inclusion in this Study of factors on the following continuums: extent of conservatism and uncertainty in judgment in ambiguous contexts; the extent of con- servatism in beliefs, extent of proneness to control events, and to settle for the status quo; extent of rigidity in interaction with students; extent of need for social acceptance and defensiveness; ex- tent of reticence in interaction with students, and degree of anxiety 149 created by demands for personal commitment beyond that which the individual feels able to make; course load; teaching experience. An Open-ended questionnaire of instructor perceptions and judgments of the course was also included. For both Student characteristics and those Of instructors, rationales were presented in chapter three for the choices not only of each factor but also for each measure used as a proxy for a Specified factor. This Study was concerned with the relative importance of "predictor" factors in terms of the aSSumptions underlying the Specific instructional Situation in this Study, and focused directly on what, in fact, was the relative importance of these factors to actual course Outcomes -- the outcomes themselves defined in terms of assessed performance. That is, the term "outcome" referred to those factors considered to be tangible representations of the theoretical assumptions of the course "model". The outcome variables involved two forms: performance variables and reward variables. Course performance factors were represented by examination scores. Reward variables were represented by: grades assigned student scores On the examinations; grades assigned by instructors; final grade assigned in the course. (The term "reward" was applied to the grades because they represented the value placed on student perfor- mance by a highly competitive System.) Subjects included five hundred and thirty-two students enrolled in the instructional Situation, and fifteen graduate teaching- assistant instructors. The course was conducted over a ten-week 150 period, five days per week, in fifty-minute sessions. Students met aS a group for lecture three sessions per week, and in a given discussion section of approximately twenty-five fellow students two sessions per week. The structure of the research design approximated the instructional process Over time, by extracting data at different points in time as the instructional process progressed. Extractions were made on two categories of factors: outcome and predictor. Both types were extracted at the following three points: entering-course, mid-course, end-of-course. Table 2 (p. 124) summarizes the predictor factors concerning which data were derived. AS for Outcome data extracted -- the first extraction included pretest-Score data, the second extraction included midterm exam Scores and grades; the third extraction included final exam Scores and grades, discussion section grade, and final course grade. The data analysis paralleled the research design, involving analyses represented by the three extraction points in the design. In chapter one the discussion Of the purpose of the Study indicated that the analysis questions revolved around two major categories: instructional process and instructional outcomes. Interest centered on what happens to the explanation of outcomes when data are added from the time perSpective of the research model. An extension of this was interest in the explanation Of results as they occurred sequentially in the course. Table 3 (p. 128) indicates the attempt to maximize explanation by looking at changes in structure as well as changes over time. 151 The first-stage analysis (labelled "1") involved two separate analyses of two outcome factors and one set of predictor factors -- that is, analysis of entering-course characteristics first to the pretest, then to the final course grade. Entering-course predictor factors involved in the analysis of the two outcomes included: sex, age, major, GPA, OQT-scores, extraversion, anxiety, social acceptance, general test anxiety, pretest anxiety, learning set, reason for enrolling, pre-attitude toward the instructional Situation, and course-Specific motivation at pretest time. The second-stage analysis involved three separate analyses and one set of predictor factors. Separate analyses were made of the following three Outcomes: midterm Score, midterm grade, and final course grade. The predictor factors involved in the second- Stage analysis (labelled II) were: credits earned, current load, midterm anxiety, pretest Score, and the following instructor factors (extraversion, anxiety, defensiveness, social acceptance, rigidity, "conservatism"), combined with the predictors used in the Stage-one analysis. The third-Stage analysis involved six separate analyses and one set of predictor factors. The six outcomes,each part of a sepa- rate analysis,included: final exam-recall section; final exam- applied section; final exam-total; final exam grade; discussion sec- tion grade; final course grade. The set of predictor factors used in the analyses included: final exam anxiety, post-attitude, atti- tude toward discussion section, course-Specific motivation after mid- term, Success expecation after midterm, and accuracy of this judgment, 152 midterm score, instructor risk, instructor course load, instructor teaching experience, combined with the set of predictor factors in the stage-two analysis. In summary, the research strategy was to try to approximate the theoretical course model through the design and analysis models -- that is, through the data gathered, the timing Of the data gathering, and through the combinations of data used in the analysis stage. A total of eleven analyses of nine course Outcomes were made on the basis of three sets of predictor factors. Under such a strategy, assumptions of absence of interactions, nonclassifactory data, nonrandomness, and so forth, were not tenable, necessitating an analysis strategy the aSSumptionS of which met the requirements of the model used. Therefore, the strategy adOpted was an anlySiS process which accounted for variance in a Specfied Outcome factor by Optimal Splits of a Specified set of predictor factors. Parallel- ing the study design, the analysis strategy also, in the beginning, regarded subjects as a singde group. This initial group of subjects was then divided, through a series of binary Splits, into a mutually exclusive set of Subgroups, the Subgroups chosen so that their means accOunted for more of the total sum of Squares than the means Of any other subgroups. The considerations underlying the concept of "the instruc- tional process" and those underlying the concept of "importance" as discussed at the beginning of this summary Suggested a number Of expectations about the structure of relations among the factors at 153 any given moment and over time, in the particular instructional situation used in this Study, to be revealed by the analyses. Briefly Summarized, they are: 1. Coalescence will occur arOund certain classes Of predictor variables in their description of a given outcome variable, the coalescence differing according to conditions Surrounding the various Outcome variables. The results will Show differences among the outcome variables (assessed Student performance) in the type of predictor vari- ables accounting for variance and in the extent of the expla- nation of variance. Measured factors may together represent a weighting of their underlying theoretical constructs in the form of an interaction. Certain instructional process variables Operate differentially as facilitators or barriers to Student performance in meeting course criteria and outcomes. The results will Show isolated exceptions, normally relegated to measurement or Statistical error, which represent either important exceptions to general prediction, or real instruc- tional problems. TO some degree the data will provide an estimate of congru- ence between course goals and conduct Of the course. In chapter four the data are summarized and displayed to illustrate the analysis Strategy used in interpreting the data, and the data are discussed in terms of the concept of "importance". CHAPTER IV DATA DISPLAY.AND DISCUSSION OF "IMPORTANCE" This chapter has two purposes. First, the data is Summar- ized and diSplayed to illustrate the analysis strategy to be used in interpreting the data derived from the analysis of each set of predictor factors to an Outcome factor. Second, the presentation is a discussion of the relative importance of the predictor factors to the course outcomes, on the basis of the four characteristics of importance which together provided the basis for the evaluation Of the predictor factors -- that is, amount, location, time, and criterion, as delineated in chapter one. 1. DATA DISPLAY FORM Complete diSplay of the data is found in the appendices. Appendix B ppovides, for all analyses, a summary of the variables, including number, mean, standard deviation, skewness and kurtosis; Appendix C provides an intercorrelation matrix of the variables. For each Outcome factor used, the following are provided: Appendix D Each tree provides a visual representation of the groups formed on the basis of predictor variables. When two groups are formed as the result Of a Split, the group with the higher mean appears above the other. A group is represented by a box which lists: .a) the predictor classes of the predictor variable used in that particular Split b) group mean and standard deviation on the Outcome 154 155 c) the per cent of sample included in the group d) for a final group, the group Size, and an asterisk designating it as final. Figure 1, p. 157 is an example Of the form of data in Appendix D. Box number two (2) summarizes the fact that students appear in this group on the basis of grade point averages of 1.1 to 2.7, the mean final course grade is 3.376 with standard deviation .730, 72.1 per cent of the Students in the study appear in the group, and it is not a final group. Appendix E The Statistic (BSS/TSS)i is examined for each predictor over each group created during the partitioning process, showing the prOportion of variation in each group explainable for each predictor. The group numbers in the column headings correSpond to the group numbers in the trees of the figures in Appendix D. Indicators mark the predictor used in 3 Split, the next best choice, the Split-fail attempts, and the final groups. Table‘L p» 158 is an example of the tables in Appendix E. The prOportion of variation in group three (3) explainable for each predictor appears in the second column. The arrow from group (3) to groups (12) and (13) indicates that predictor (50-50-1) was used in the reduction Of variation. That is, differentiation among Students with higher grade point averages represented by group (3) in the tree and table (Figure l, p. 157 and Table 7 , p. 158) 156 FIGURE 1 Explanation of Final Course Grade by Entering Course Variables 157 0(2)) 50-50-1: ll $4.500 «15400 1,-3.31-20 (15) Phjor: Code: I S b 7 “7) Pa. Inf .d: 555 50-50-1: M, 2, ‘ ' 2.5 - 1.1 ' Y-4.320 .a-,595 27.5 .. MI.”- samba-m: (3) 2.3:. La (19) HI)“: $4.297 204:: 1.3.4.5 Id-,652 24,375 w-,szs 27.97. ,sz-u *(13) 5040-1 _1_.z — — — — ’ — — -— Y-1.000 -.ooo mar-z 02142: 1 - u. H.031 no.4“ a «:7 (6) “.1“: '(37) insuring Set: 5 x 1,5 .5 - In “3.84%; sci-.661 n-l .21 "' (’1) I : “I 3151“:- {xi-$3 "My 3-3 «29 Id- 573 u-5.JZ '06) «mm; s.:: '(17) 31:25 I - 15 .. ; 1.1.3“ sd-.771 (5) 3.:4uu-Jn n-z.u-u ._.3 - 2.7 ‘ . ”3'33; » (as) arr-Q: “'39'77' — _ _ — $3.625 311-.565 n-9.07y-45 (17') 21'“): $3,539 «1am , c -21.11 1‘) 3‘13”; \ m (10) A-E 'l'eIt Anxiety: ' 12 - “15522 “"6" «18) cur-o: n- . (2) (7) :31? 3 7 a 74-722 ““1” an; T-3.445 1:11.705 n-.6'L-J _.1 - 2.7 Y-3.376 . «h. 30 . ('11) A-E Tel: Mainly: _ ,._ _ r72.u ._. .. — \ u _ 63 , 3.1 - .2 1.3.565 .a-.s77 "“JI-ZB .(‘) —- GPA: or: a; _.... - 3.325 d-.700 _ .a-.69A :7 a, ' .97.) m, (39) 13mm: 5o:- 1162.41 ' l, - .0 _ “H.929 “.303 26.00? “1 .ooo u-2_5p1‘ n-.ZZ- (25) Bylinck Ixttawrt: 1(38) firing“ S“: 1" ' $3.077 “p.474 n-1.37.-39 I—hazs "9.556 «V. 51 s) lbjor: £0da:2,3,4,5_6,7 r-2.919 sci-.668 u-Is.z1. % 1 ’1 (24) ”Mock lunvert: ' g ‘- l4 23.757 Id-.7l.0 ,u .(32) oar-v: - 32 $2.700 pd-.678 n-7JW FIGURE 1 158 TABLE 7 AID I: Proportion of Variation in Final Course Grade by Group within Branch Explainable for Each Entering—Course Variable Predictor Group Number 1 3 12 13* 18 19* 22* 23* Sex .004 .022 .019 .016 .019 .016 .011 Age .002 .003 .003 .007 .129 .002 <::39 Major .043 .081 .109 .011 .022 .028 .063 CPA .254 .080 .084 .068 . 198 .083 CQT-V .087 .065 .056 .038 .079 .021 .204 CQT-Q .070 .102 .157 .045 .063 CQT-T .091 .086 .064 .121 .026 .210 Eysenck Ext. .025 .031 .029 NA .016 .129 .040 .306 Eysenck Neu. .003 .008 .018 .021 .170 .016 .211 Eysenck Lie .012 .020 .026 .018 .222 .014 .167 A-H Test Anx. .055 .018 .017 .023 .120 .037 .209 Text Anx.-1 .001 .001 .001 .000 Cons. .003 .013 Learning Set .041 .048 .054 .052 .152 .O§% .063 Reason Enrol .004 .010 .009 .009 .015 .009 .141 Pre-attitude .005 .015 .008 .018 .072 .049 .062 50-50-1 .085 J® .093 .021 Cons. N 532 149 147 2 123 24 103 20 TSS1/TSST 1.000 .176 .145 .000 .111 .019 .092 .009 NEAN’ 3.633 4.297 4.320 1.000 4.407 3.875 4.330 4.800 Proportion of variation in that group explainable for each predictor fl-Split made on this variable. -Next best BSS/TSS. * IFinal group. ISplit attempted but not made. NA -Split not attempted. (BSS/TSS)1 159 TABLE 7 (cont'd.) Predictor Group Number 1 2 5 6 30 31* 36* 37* Sex .004 .015 .008 .006 .012 .000 Cons. Age .002 .001 .009 .023 .015 .067 " Major .043 .031 .053 .011 .009 .033 " GPA .2fl73 .032 .044 .059 .046 " CQT-V .087 .028 .030 .107 .036 .293 " 001-0 .070 .039 .240 .034 .090 " CQT-T .027 .033 .064 .082 " Eysenck Ext. .025 .008 .007 .025 .037 " NA Eysenck Neu. .003 .009 .009 .036 .043 .051 " Eysenck Lie .012 .027 .020 .048 ® .060 " A-H Test Anx. .055 .035 .046 .065 .042 " Test Anx.-1 .001 .002 .006 .023 .015 Cons. " Learning Set .041 .029 .028 .104 .241 .074 "Q Reason Enrol .004 .002 .005 .015 .037 .040 " Pre-attitude .005 .006 .011 .013 .026 .006 " 50-50-1 (Asp.-Exp.).085 .009 .007 .036 .036 .053 " N 532 383 211 65 28 37 23 4 TSSilTSST 1.000 .570 .297 .079 .036 .025 .025 .002 MEAN 3.633 3.376 3.555 3.800 3.429 4.081 3.304 4.250 PrOportion of variation in each group explainable for each predictor A -Split made on this variable. -=Next best BSS/TSS . * =Final group. aSplit attempted but not made. NA =Split not attempted. (BSS/TSS)1 160 enmme\mmmv soooeeose some see oaesseoflexo .ooumEOOuw no: uwanm <2 .ovoa uo: use mougouum uflamm é .Qsoum Hodge ..mm.a\mmm ease 382 @ .Oaamfiua> menu so some egaam I arl\. msouw umau a“ cowuwfiuw> mo cowuuoeoum mme.m -e.~ som.e Nem.m mmm.m see.m exm.m mmo.m eoe.~ osm.m mse.m mmm.m oem.m mme.m zen: meo. mmo. wee. «so. «so. ewe. ema. see. «no. ema. «om. mom. can. ooo.~ Hmme\amme me n Ha Hm es no NHH NH oow sea sea Ham mwm «mm 2 see. m .. Hmo. ooo. ea. : oao. mmo. moo. ooo. owe. A.exm-.ooao H-om-om mac. ennv_ ooo. Ame. ago. : moo. mac. Hao. eoo. moo. oesoeoos-oso eao. mmo. mac. moo. Noe. : ooo. moo. moo. moo. soc. seasonsm soosom oeo. use. emo. «no. mac. : mac. sac. mac. euo. Hso. oom wesssooa .eeoo .eeoo oHo. .osoo moo. r moo. moo. o . Noo. Hoo. Huxooaxe<_oxoe oso. mao. one. H . moo. . ea . H. AWWWV nae. mmo. eooaxse.oeoe m-< So. «no. So. WW one. s. gem. So. So. So. So. one xoeoexm mo . Hso. so. awe. emo. r owe. ego. eoo. moo. moo. eneoeooosoz xoeooxm Annmw ez e2 mac. nee. sac. «Ho. ez : see. mac. Boo. moo. m .. roseso>onoxm xosonxm e8. a 8?. So. was. So. .. m8. «8. So. So. p.98 3F lmB So. 3?. «81. a .. $0. a one. g 2o. 0-98 “No. one. emo. sea muo. .9: «no. one. one. mmo. “we. >-eoo eoo. ewe. . ooo. mm . r m . Nmo. «mm. m» . mu. emu 62H. 93. So. .. AWW ago am «so. some: see. mac. one. «#0. Boo. : oHo. ooo eoo. goo. woo. om< moo. mac. mac. amo. Hmo. .osoo Hue. mao. moo. mac. soo. sow saw emu sea on raw ow NH sou sHH 0H m m N H nonasz mnouu aOuoavoam A.e.ocoov .m mam mace so some umamm I.firl|\ noouw umnu OH comuomum> mo coauaonoum mz 6 k. ooo.m meo.m mom.m omo.m moo.m ooe.m mma.m eoe.m oao.m omm.m oss.m Hso.m mmm.m omm.m mmo.m zemz ooo. moo. moo. omo. moo. Hmo. mmo. moo. mom. Hmo. moo. one. mmm. omm. ooo.H mmmm\emmm H mm mm ea m oe os me om mm es mma mam mmm «mm 2 soo. so . mmo. smo. emo. Mao. moo. mao. mmo. moo. moo. mmo. H-om-om smo. smo. mmo. emo. mmo. omo. moo. ooo. omo. moo. moo. oesoeooo-onm emo. moo. moo. Hao. moo. Hoo. moo. ooo. eoo. moo. moo. soo. moose soonom olnpwpmrnmwww. omo. .Hom. smo. mmo. moH. emo. mmo. mmo. mmo. moo. oom measures mcoo mac. wmwoo .meU .meU .mcou coo. .mcoo coo. ooo. moo. Hoo. Hu.x:< umow mom. moo. mom. moo. mmo. Hmo. mmo. moo. omo. oHo. mmo. mmo. .xr< oxom m-< moo. mmo. . emo. oo . emo. omo. moo. oao. omo. mmo. «Ho. sea xoeooxm ooa. oma. WWW So. @ mmo. mmo. So. moo. moo. moo. moo. .soz sosonmm mz mam. omo. m2 m2 mmo. meow meow meo. omo. mm. omo. mmo. moo. m .. .oxm eoeonxm oom. moo. ooo. moo. moo omo. mos. mo. omo. omo. e . m-moo So. o3. m3. So. emo. e .. m . R . omo. 3. ® o3. odoo m . 3o. omo. sod. a. % % So. So omo. mmo. mo. :8 @W as . armoi . . S. ooo. mmo. m .. mmorlemm. «so m2. moo. so: moo. «3... Son. mmw o8. @148. moo. some: emo. Hoo. moo. moo. Hmo. mmo. mmo mmo m .. omo. moo. moo. ems mmo. Go. Go. So. So. mmo. So. mmo. & So. So. eoo. xom xmm xmm xmm tom xmm xmm mm em mm ea so m s m m Hoeasz moouo acuomcoum 30.083 5 momma 162 is made on the basis of their aSpirations and expec- tations for the course (50-50-1), and the students are now represented by groups (12) and (13). Because of the number of groups represented in any given tree, to aid the reader the groups are placed together in the (BSS/TSS)1 table in the natural groupings repre- sented in the tree. For example, in Table 7 , (pp. 158-161) each of the four pages represents one of the four major branches in the tree of Figure 1, @u 157). The groups appearing on page 158 Of the table are the same groups appearing in the main branch in the upper part of the tree in Figure 1. All other figures and tables appear within the text of the discussion. A number of analyses were made involving outcome variables (dependent variables) and different sets of predictor variables. In the figures and tables, any analysis involving entering-course predictor variables is referred to as AID 1; involving mid-course predictor variables, AID II; involving end-Of-course predictor variables, AID III. For example, analysis of the final course grade involving entering-course characteristics is labelled AID I - Final Course Grade; analysis of the final course grade involving mid-course characteristics, AID II - Final Course Grade; analysis of the final course grade involving end-Of- course characteristics, AID III - Final Course Grade. 163 II. FINAL COURSE GRADE BASED ON ENTERING-COURSE CHARACTERISTICS DISCUSSED AS AN ILLUSTRATION REPRESENTATIVE OF ALL OUTCOME VARIABLES In this chapter the data results are summarized, diSplayed, and discussed to illustrate the analysis Strategy to be used in inter- preting the data. The data from one analysis, final course grade based on entering-course characteristics, is presented and discussed in detail: first, to Show a series of characteristic patterns useful in examining the data of each of the outcome variables; second, to illustrate the analysis strategy in interpretation. The figures and tables identify the data of "final course grade based on entering- course characteristics" as "AID I - Final Course Grade". Data of each of the other outcome variables is discussed in summary presenta- tions, with appropriate references to complete data and documentation location, following the discussion Of AID I - Final Course Grade. Characteristic Data Patterns The tree in Figure l, (p. 15), represents for the final course grade a pattern of groups formed using the student entering-course characteristics as the basis for discriminations. This tree illustrates a general pattern of two configurations common to all the trees. For the first few groups, each group produced by a Split is further sub- divided. For example, groups (2) and (3) branch from group (1), groups (4) and (5) from.group (2), groups (6) and (7) from group (5). A second configuration occurs when groups are Split from the main branch and not Split again, constituting final groups. For example, in the lower branch of Figure 1, groups (8) and (9) are formed from group (4), but group (9) is Split from the main branch which continues via the 164 series of groups formed from group eight. The combination of the two configurations results in a patterning of the boxes into several branches. For example, there are four main branches (separated by dotted lines) in Figure 1, each involving a different combination of Student characteristics useful in discussion of student final course grade. These four branches appear in Table7, (pp.158-16l), the (BSS/TSS)i table for AID I - Final Course Grade. The first page of the table contains data pertaining to the upper tree branch composed of groups (3), (12), (13), (18), (19), (22), and (23); the remaining pages of the table each pertain to one of the remaining main branches in Figure 1. This table expands the discussion of the characteristics of the predictor vari- bles by showing, in addition to the characteristic utilized in any given Split in the tree, the relative power of the other predictors to reduce unexplained variation. For example, in branch three of the tree, "major" was almost as effective as "CQT-Q" in the Split of group (10) into groups (16) and (17), as Shown in Table7 , (p.158). ‘Learn- ing set is used in several Splits in the tree, but also would have been used to Split several large final groups had the Split reducibility criterion been lowered, and this is evident throughout the table pages. Each main branch contains a number of final groups which are not Split further. Either the number of students is too Small to warrant the Split of the group, the proportion of variation in the group compared to the variation in the total sample is too Small, or no predictor in the analysis can reduce the unexplained variation in the group the required amount. The theoretical and statistical procedures resulting in formation of final groups and application of this procedure 165 in this study are discussed in Chapter III. Groups twenty-eight and twenty-nine at the far right of Figure l are representative of final groups in all trees. Group twenty-eight is too small to attempt a Split. Although group twenty-nine has Sufficient internal variation to warrant an attempt to Split (.043 with TSSI=15.250 and TSST=357.39O, compared with the Split eligibility criterion of .0001), group twenty- nine cannot be Split further (.005 with BSS=1.797 and TSST=357.39O, compared with the Split reducibility criterion of .006). Predictor variable "major" comes closest, but does not reduce the unexplained variation enough for the Split actually to take place. Additional variables appear needed in the analysis. Discussion of the Data for AID I - Final Course Grade The final course grade is a weighted composite of twenty per cent midterm grade, forty per cent final exam grade, and forty per cent instructor grade, with a range of one to five representing grades F to A. The Specific components of the final course grade and the rationakafor a range of one to five rather than zero to four are discussed in Chapter III. Only entering-course variables are used in this analysis of final course grade. These are.discussed in Chapter III and Summarized in Table4, (p. 129). The mean final course grade of the original total number of students in the study is 3.633, with standard deviation .821. Nine of the fifteen entering-course vari- ables were used to form twenty final groups. The nine variables accounted for fifty-one per cent of the total sum Of squares. (Table 8,pJ66) -' f 166 TABLE 8: PROPORTION OF VARIATION ”EXPLAINED” IN COURSE OUTCOME VARIABLES BY ENTERING-COURSE, MID-COURSE, AND END-OF-COURSE VARIABLES V r' 1- . . Squared Beta Coefficients From a lable A‘D Reduction in TSS(I) /TSS(T) Deletion Regression Analysis AID I,” m All!) II *2: hi Pd HAHID 11111 H H Del. Regldl L”Del, Reg. 11 H1 ’11 €DeéHRegfij III H hd ’U H- N3 >43 H- CDH- mH- OH- x :J H‘ "U H- :43 x: P“ me- me- Or-u 5: :1 H- :93 E28 $3 “a“: $35 88 88 38 SE 93?; $252 $8 535:? SE at: 98 85.? 86’ SB 2:3 53:”. 535’. 38 8." m8 o8 ‘3" 3" :3" *‘H mi 8.2 a“ SE: ”A” 0.8 mg ‘3" 3*“ a?" "“ off. a: ‘3’" (DO) (DO OH HH mO|LTJ|l1TJ ml?! HH (DO (DO (DID ('00 0'1 HH (DO III-1'1 ltd (0111 Hi—I (DO 030 .. E 9.8 88 5° “E ”E PS E 8 E .. as 28 .3 TE ”S PE 8 8 8 (D (‘D E E L m H (D m =1 1:1 :1 m H Sex .011 .008 002 005 007 .003 003 .002 .001 Age .009 006 .008 .010 .018 .008 004 003 .005 .007 .002 .003 .002 .001 Major .086 .042 .049 .045 .051 .036 .029 .028 .023 .014 .002 .001 .002 .003 .003 .004 .005 004 .002 Credits ** --- --- .023 .009 ~-- --- .006 .007 .009 .004 .001 .004 .003 Current Load --- --- .010 --- --- .001 .001 .001 .001 GPA .018 .306 .208 .194 .295 .040 .201 .222 .230 .077 .088 .007 .227 .123 .144 .213 .027 .059 .054 .040 106 .081 CQT-V .111 .008 .032 .026 .009 .007 .021 .018 .012 .001 .008 .002 .002 CQT-Q .053 .061 .041 .025 .017 .024 .010 .046 .018 .029 .004 .003 .022 .004 .011 .010 .013 .003 .009 CQT-T .022 -.016 .013 .018 .010 .037 .084 .173 .030 .005 .004 .004 .002 .002 .006 .015 .009 .012 .009 Extraversion .006 .012 .007 .007 .018 .004 .019 .010 .030 .003 .001 .001 .002 Neuroticism .023 .007 .036 .024 .018 .022 .006 .029 .001 .003 .006 .005 .001 EYsenck Lie .009 .015 .022 .008 .018 .014 .015 ,012 .007 .002 .005 .001 .009 .005 .006 .005 .002 A-H Test Anxiety .096 .021 .031 .018 .036 .053 .007 .002 .004 .005 .008 .002 Test Anxiety-l .011 .002 .003 .006 .006 .003 .005 .002 .005 Test Anxiety-2 --- -_- .006 ~-- -~- .001 .003 .008 .002 .005 .003 .003 .004 Test Anxiety-3 ..-.. ..-- --.. -..- --- .008 .008 -_- ..-_ --- --— --- .003 .007 .005 .001 .001 .002 Accuracy --- --- --- --- --- .010 .020 --- --- -_- --— --— .003 .001 .001 .001 Success --- --- --- --- --- .007 --— --- --- --- --- .001 Learning Set .059 .042 .013 .009 .024 .026 .021 .041 .009 .011 .007 .007 .006 .005 .001 .009 .005 .005 .002 .001 Reason Enrolled .049 .007 .006 .001 .001 .001 .008 .001 Pre-attitude -.015 .009 .019 .012 .020 .009 .009 .002 .006 .002 .001 .001 .001 .001 .003 .002 Post-attitude -.._ --.. -.... ....- -.... .007 -—- --- --- --- --- .001 .001 .002 .002 DiScussion Attitude --- --- --- -—- --- .015 .006 .046 --- --- --- --- --- .001 .001 .007 50-50-1cAsp.-Exp.-1) .009 .041 .010 .015 .031 .031 .052 .069 .024 .006 .001 .002 .004 .002 .003 .001 .001 .006 50-50-2(Asp,-Exp,-2) --- --- --- --- --- .021 .030 .007 —-- --- --- --- --- .012 .007 .012 .014 .009 .025 Pretest --- --- .076 .079 .019 .020 .014 .008 .012 --- --- .019 .029 .022 .007 .001 .004 .009 .001 .004 Midterm -_- --- --- --- --- .268 .049 .130 .139 .141 .539 --— --- --- --- --- .089 .022 .071 .069 .053 .172 InSt. Extraversion --— --- .024 .014 --- --~ .001 .006 .005 .004 .005 Inst. Neuroticism --- --- .026 .015 .008 --- --- .002 .002 .026 .001 .014 .007 .008 .065 .025 Inst. Eysenck Lie --- --- .014 .015 --- --- .004 .003 .002 .002 .030 .008 Inst. Marlowe-Crowne --- --- .008 .007 .006 .040 ~-- --- .002 .003 .031 .013 .002 .041 .025 InSt. Sanford-Gough --- --- .013 .009 .013 .021 --- --~ .008 .003 .012 .001 .001 .001 .003 Inst. California-F --- --- .007 .028 .009 .006 .013 ~-- --- .001 .004 .005 .009 .009 .005 .036 .004 Inst, Risk ___ -_.. ..-.. -..- -..... .027 --- --- --- --- --- .016 .007 .005 .003 InSt. Course Load -o— -—- --- --- --- .010 .007 .029 --- ”'- "" "" "" :002 .002 -001 -003 ~005 IDSt. TEaching Exp, --- --.. --- --- ~-- .010 --- --- --- --- --- .001 .011 .001 .001 Total proportion of 2 variation explained(R ).537 .517 .545 .539 .594 .540 .640 .782 .577 .605 .746 *.318 ~.354 .308 .287 .399 .356 .370 1.336 .450 -298 .589 *Beta coefficients do not add. This is an adjusted R2. ** --- Variable not included in this stage of analysis. I1 1!. lll-.||ll. 167 The analysis shows the powerful effects of past performance variables, particularly grade point average and to a lesser extent the quantitative portion of the College Qualification Test (CQT-Q) in determining the final course grade. The total group is divided first into three groups (3, 4, and 5, Figure l, p. 157) on the basis of grade point average. These major Splits on GPA are not unexpected. Numerous studies have indicated that grade point average is the single best predictor of potential Student grade in a college course. The tree indicates that on the basis of grade point average, students with a GPA in the range of B-minus to A-plus have a mean on the dependent variable of 4.297, those in the C to C-plus range have a final course grade mean of 3.555, and those in the range below C-minus have a mean of 3.157. While not all high final course grades are among the high GPA'S, the high GPA'S as a group (with one exception discussed below) do have high mean final course grades, and GPA does maintain strata in that with isolated exceptions the high mean final course grades among middle GPA groups do not reach the highest means of the high GPA groups, nor do the high means among 10w GPA groups reach the highest means of the middle GPA'S. However, among groups without the advantage of high grade point average there are entering-course advantages which cumulate into final course grades higher than grades of those who are at a disadvantage on the characteristic, and in some cases are competitive with those characterized by high GPA. These advantages appear associated with Skills in conceptual learning. The lower branches in the tree each present Splits on the variable CQT-Q. Middle and low GPA groups with 168 higher OQT-Q scores have higher mean final exam grades than their reSpective counterparts with lower CQT-Q scores, with one exception. (See groups 31, 30; 8, 9; l6, 17; 28, 29.) In the case of the reversal, those among the low'OQT-Q who achieved a high mean final course grade also were characterized by high CQT-T's. The import- ance of OQT-Q is illustrated not only in the Splits in Figure l but in Table7 where it is second choice not only in a number of Splits throughout the lower branches but in the upper branch of high GPA'S also. In addition, among low COT-Q's, those indicating a preference for conceptual learning have higher mean final course grades than those who do not share this preference, and, in fact, the mean for the former is competitive with that of their higher CQT-Q counterparts. (See groups 36, 37; 38, 39.) Why Should these variables be more highly correlated with the dependent variable than any other one in the analysis for these particular groups? The CQT-Q attempts to tap the ability to analyze, to break set, as Opposed to the vocabulary- analogy orientation of the OQT-V. The learning set instrument attempts to tap preference for conceptual rather than factual learning. The second half of the final exam was composed Of items involving problem- solving rather than straight recall Skill. In addition, it is possible that this preference and ability was applied in discussion of material in discussion section, and noted by instructors as part of their evaluation. There appear to be for high, middle, and low GPA'S different sets of barriers to high final course grade. In the middle branch of the tree the most important subsequent Split is one involving test 169 anxiety. The group of Students in the middle GPA range who are not elementary or social science majors is Split by test anxiety. (See groups 10 and 11.) It is possible that the Split has isolated a group of students with middle GPA who are handicapped not only by not being elementary or social Science majors, but by high test anxiety. Perhaps elementary majors have already taken some required courses in child develOpment or child psychology, and the social science majors have background in sociology and psychology courses. The importance of test anxiety extends beyond the isolation of the group on this variable. First, the group handicapped by high test anxiety has a mean depressed a letter grade below that of those not handicapped by high test anxiety, a mean nearly as low as low scores in the course. In addition, Table 7 indicates that test anxiety was almost as useful as was major in the Split of the entire group of middle GPA scores (group 5 into 6 and 7). In many Splits in any given tree, one factor is Often clearly Superior to any other for a given group. However, in this particular case, while test anxiety was not used to Split the entire group of middle grade point students (group 5 into 6 and 7), its relationship to the dependent variable was almost as strong as that of major, and in fact its relationship to final course grade was not reduced enough by the first Split to prevent its being used in the immediate subsequent Split of group 7 (Table7 , p.160 ). It is possible that in the middle grade point range test anxiety has a more pervasive effect than is indicated in the tree, and that its effect as a depressent of final exam grade is considerable. In the lower branch of the tree, the most important subsequent Split involves the personality variable extraversion. It is an 170 interesting sequence for several reasons. First, it involves 16.2 per cent of the original total group in the study (groups 24 and 25). Secondly, it involves about two-thirds of those at a disadvantage on GPA and CQT—Q. Third, Table 7 indicates that for the group who also had low GPA but had the advantage of high CQT-Q, a large group isolated early in the branch (group 9), extraversion was the choice to split the group, had the reducibility criterion been met. Why should a higher degree of extraversion be an advantage particularly among students of low GPA? Forty per cent of the final exam grade is instructor grade, and perhaps this extraversion is reflected in participation in discus- sion, in the contacts with the instructor. And indeed the group with low GPA and low conceptual skills coupled with lower extraversion appears as low-man-on-the-totem-pole in terms of course grade distri- bution (group 32). In the upper branch of the tree containing groups character- ized by high GPA the most important subsequent Split involves course- specific motivation reflected in the level of aSpiration and expecta- tion of course performance (50-50-1). It is of interest for several reasons. First, two of the three Splits of the upper branch involve this variable. In addition, Table 7 shows that this characteristic was the second choice to make the Split of group two, the only other split in the branch (Figure 1). Having important weight in every split in the upper branch of the tree, course-Specific motivation does not appear in the branches nor Show any Strength as second choice among those groups characterized by middle and low grade point averages. For these groups, other factors such as test anxiety and extraversion bear more weight in determining course 171 performance. It is among those groups characterized by high GPA that course-Specific motivation appears to have importance. A second reason why this variable is of interest is that it is used to isolate extreme cases of high Specific-course motivation. In the case of groups 23 and 22, group 23 isolates twenty students for whom course-Specific motivation is very high, the level of aSpiration for course performance set at a minimum of "A". The mean on the final course grade for this group is predicted at 4.800. An earlier Split in the same branch involves isolation of an extreme case of high specific-course motiva- tion also. The Split of group three into twelve and thirteen occurs early in the branch, involves 27.9 per cent of the sample, and isolates one individual with course-Specific motivation of A-plus and a predicted final course grade of F. When a small group is isolated so early in the branch, one interpretation is that this is due to chance. However, the groups appear in the data of final course grade based on mid-course variables, and that of the final exam application section and final exam total score, and also in the data of the final exam recall section and final course grade based on end-of-course variables, associated with lower midterm scores. All other instances of the use of course- Specific motivation follow the pattern of groups 22 and 23 -- higher course-Specific motivation associated with higher score or grade. A third reason why the variable (50-50-1) is of interest is that it illustrates the importance of identification of interactions if meaningful questions are to be asked of the data. In the tree of Figure l, course-Specific motivation is asymmetrically distributed in the tree. It is also apparent in the (BSS/TSS)1 table for Figure 1 172 (Table 7 , pp. 158-161) that the variable gains in explanatory power in this branch while this does not occur in the other branches. On the other hand, examination of the correlation matrix indicates a correlation between entering level of course-specific motivation as to course grade, and actual course grade, of .260. (Appendix C) In FigfirejL (p. 173) observation of the plotted points of aSpira- tion-expectation on course grade for the entire study group indicates that there is a regression. However, the regression rather sharply changes its slape at approximately the mean values of both indices, suggesting an interaction in addition. The evidence supporting regression plus interaction is further exemplified by examining the plotted points for the three levels of GPA occurring in the three initial Splits of the data. Again, as shown in Figure 1, (p. 173) the points for high GPA involving more than ten cases follow a positive accelerating path; the points for middle GPA follow a hori- zontal path; the points for low GPA follow a negatively accelerating path. Table 9 , (p. 154) shows the means, standard deviations and number for each of the plotted points by category. It appears that course-Specific motivation interacts with GPA in a positively accelera- ting manner. In summarizing the discussion of final course grade based on entering-course characteristics it can be said that grade point aver- age exhibits a powerful effect. It appears that higher conceptual ability and preference for conceptual learning are advantages associa- ted with higher final course grades among all GPA levels. There are, for high, middle, and low GPA'S, different sets of barriers to high 173 FIGURE 2 MEAN EXPECTED COURSE GRADE PLOTTED FOR VARIOUS GPA GROUPS COURSE ASPIRATION-EXPECTATION FOR COURSE GRADE BY ENTERING- coflumuommxmlcoflumuflmm< mwmuw mmusou mo Hm>mq mcflumucm “cowsum Ti Ha $5 3 Am: a TB n+8 8V m S o T8 m wwwa OH GME.“ OHOE m.~ m.m 6.... Tm ~.m m.m Tm aims 9mm 5mm 8 m.mo magma woofia A4 _N.¢ as» its mé oé «he may 174 mmmum>m ucHoa mpmuw 30H a msouu mmmum>m ucHoa mvmuw mmcmu mvaHS m asouw owmum>m ucHom opwum stm m ozone H ooo. ooo.H H ooo. ooo.H < NH m NoH. ooo.m m ooq. oom.m ow ooq. oow.¢ om mom. mmm.¢ n< HH 0H moo. Nom.m o qu. mmm.m oq com. mom.¢ we NON. mwo.¢ +m oH ma was. mmm.m mm owe. qom.m mm mum. NmH.¢ ooH mam. mwm.m m a mm mm“. «om.m ma moo. wwH.m mm Cum. Hmo.¢ 00H own. oom.m um m we moo. moo.m cm nmm. NN~.m NH «No. mmm.¢ NHH 0mm. cow.m +0 u m OHw. mmm.m mH one. mom.m H ooo. ooo.¢ mm Ono. on.m o o m ooq. oo¢.m oH ooo. oom.~ N oom. cow.¢ NH mmm. omH.m :0 m z b w 2 Lu w z .b w z :0 w 36.5 3500 H 2.3 I l l I now SOHumemm< m msouw a nsouu m macaw Hmuoa amusooumcHamucm maéo Magoo $sz mom zoflfiommxméofl§flm< mmgoudzgmyzm am .838 mmmsou .2sz amaumenm z In: s u:- unn A.mxo .am> .moum Hmuoavsuouoaz «Ho. moo. «Ho. omo. mHo. mac. one. -In us- * A.mxu .um> .noum Hmuoavummumum moamauowuom omuaoo mmo. mac. mam. mam. mam. HoH. Hmm. mow. SANA amm. emmw emcaaammS soaumapm> mo SOHuaomoum Hmuoa omo. mas. awe. “mo. oHo. mac. mao. sao.- NNo. H-8oo mac. sac. oHo. «No. “Ho. mNo. Hao. Hoe. mmo. o-eou Koo. moo. smo. «no. woo. HHS. >-eou wmo. Hao. 0mm. NNN. HON. oao. mam. SSH. mom. com. mac. emu mommEHowuom ummm r e m m m e e . . 0 .d a a a . d r S s t a xT X2 x1 8 ma mo C t C6 Ge 1.... E. E. E. C8 If. IC 8 SE .0 Ud 8G 6 e e .0 6G es 1d er 1a ra n 1r 1r .1... la t t 83 to ma 3 am 3 mm mm mm Mm E... .3 we .1 n x is is iS .1 Mx Mx F P F I E F F F F E E HHH oH< HH oH< H pHa AavmmH\Avame as coauoaoom QH< oasmaua> mamfimfiw mozgmommmm HmMHHOU Nm mmHm mo ZOHHMOHOMm oH mama. 220 of entering knowledge and abilities (as assessed by past performance) accounted for a far greater prOportion of variance than did combina- tions of attitude, personality, or personal characteristics. (Tables 11, 12, and 13, pp. 221, 222, 223) However, within that combination, the structure of relations appears to be different. In the case of the pretest, entering-course ability as assessed by the verbal score on the college qualification test was the single meaSure accounting for the largest proportion of variance. The discussion of the pretest'in chapter three indicated the questions did involve vocabulary in educational psychology possibly unfamiliar to students. This verbal factor was coupled with major field of study, in accounting for variance, suggesting the importance of an additional entering ability in the form of previous instruction on similar content. By contrast, the OQT-V score acc0unted for less than one per cent of the prOportion of variance in final course grade, compared to eleven per cent in the pretest; however, another past performance indice, gradepoint average, accounted for thirty per cent of the proportion of variance in final course grade, compared with two per cent for pretest. A comparison of the variables in Tables 10, ll, 12, and 13 indicates that attitudes, personality characteristics, and personal characteristics accounted for twice the proportion of vari- ation in the pretest as compared with final course grade. (In the case of past performance indices discussed above, the weighting was on the side of final course grade, although in both pretest and final course grade past performance accounted for the greatest prOportion of variance.) In addition, for the pretest, the prOpor- 221 .mHmszco mo swoon mHsu SH ooosHocH uos OHanuw> :1: « Hmo. mNo. use. wmo. mac. one. mmo. Hmo. coo. «so. nae. emaamaaxm soauaaua> Ho SOHuuomowa kuoa So. a..- E... .53 ucmsuso moo. mmo. ...i u-.. a. 3.25.5 «Ho. 20. 30. So. one. Hmo. moo. moo. «so. owo. scum: moo. 30. So. woo. A5o. moo. mma. moo. :0. new a m M q o a 1m o o m So t s t a x c x 2 x l 8 m d m 0 C 8 e C e c e l r ES E . E . C e r G r c e e r .d u.o m.G e e .a e ens .l.d t o l m H... m .1 m 11. 1 r 1.. r 1 a t m t M a e c me me Fm mm mm mm mm mm mm A as F. .1 an w.%u mung mung m” Mung MHMP F. HHH DH< HH QH< H QH< Havmme\Hvame as coauusemm QH< manuaum> mamfimfiw HSHOmmmnH HzmeHHm Hem mmHmmHm<5 HZOUHDO MmMDOo ZH :QMZHmmem: ZOHH mo onHmomomm HH Ema. 222 .mHmmHoco mo swoon chu SH ooooHocH uoc oHnoHuo> an: % mmo. NOH. nNO. COO. who. OHO. mmo. moo. moo. mmo. «OH. ooanmeo SOHOSHHO> mo OOHOHOOOHO Hmuoa moo. nun uuu nun an: nun mmooosm ONO. OHO. In: an: nun nun In: moousoo< woo. woo. --- --- --- --- -u- m-aumaxc< same OOO. nun nan s Nazuona< umOH Hao. H-auoaxs< same Koo. mmo. emo. wHo. Hmo. Huo. sac. Asmaxc< same m-< mac. «Ho. mac. moo. «No. mac. soc. mag assuage muo. coo. «No. wHo. emo. omo. poo. mmc. amaoauousoz omo. oHo. SHo. aoo. wHo. moo. noo. «Ho. coo. aoamum>muuxm U1 e m1 1m d e e . . 0 .d a. a . .d r S t s t a x c xnz x.1 s m a m o no S e Ce ce 1r ES E. E. Ce rr re 6 er .6 ud 8G 8 e .0 8G 63 1d t0 1. .2. mm a.“ 1:. A... 1m a. a. .2 my. Mnrm “Mm Fa nt MC nc MG .18 .18 1G P. i. n x .1 o ales area 41 ”M x Mnx w. umr I % Fhl W. F F E m... HHH OH< HH aH< H OH< HavmmH\Hvame as soauosemm oH< manuawm> mmumaaaas SeHaaZOmamm azmnaem we . mmumaeaas excuapo mmmsoo z” gmmznaammm: zoawanmas mo zoaamomomm NH Ema. 223 .mHthwcm mo omuum mHsu OH ooosHocH uoc oHnOHuw> In- a ONO. moo. moo. mno. Hmo. Hmo. mHo. OHO. Hoo. OOO. ooaHuH o OOHucHuu> mo SOHuuomoum HouOH moo. Ono. HNO. In: nun nun nun nu: «non-Om ONO. moo. Nmo. Hno. Hmo. mHO. OHO. HOO. moo. Huomuoo OOHumuooannSOHuwwHOm< ooo. qoo. qu. uno. moo. Hoo. omo. mmo. NNO. moo. moo. oochH xo :oHumHum> mo OOHuuomouo HmuoH oso. ooo. mHo. -u- nu- --- ..u n-- moSOHuu< conmsumHo Boo. nu- -u- -u- u.. --- a monuHuumuumom ooo. moo. ONO. NHO. mHo. ooo. mHo.u ooauHuumu0pm HOO. ado. ooHHoucm acmmom moo. qu. HNO. omo. emo. moo. MHO. moo. omo. uom wchumOH maeSBHuu< r e m m m e e . . 0 d a. a a . d r S s t a xc x2 x1 5 ma mo C t C6 C6 1.1 ES E. E. C6 Ir IC 8 se :0 Ud 3G 8 e .0 8G es 11:0 er .18 re n 11 1r 1.... la t t 83 t0 ar tr im aa a0 a0 ar dm dm nr ec BC SG Fa Ht nC DC RC .18 .13 .1G IS .1 n x .1 o .1 S .1 S .1 M x M x F P m. a. as F T F \Cr F in E HHH OH< HH OH< H OH< Emmifivmme 5 53263. a: 032:; mmHm mzooHDo MmMDoo ZH :QmZHonNm: ZOHH mMHm MQDHHHH< HZWQDHw Mm mH mHm ulna mHO. mmH. Nmo. Ono. omo. RNO. MNO. NNO. Nmo. In: nun ooaHmmeo :oHuwHuw> mo SOHuuomoum HouOH OHO. In: In: nun us: In: .mxm waHsowoy HeuosuumaH mNo. moo. OHO. II... III III III III Homo...” mmubou HouosuumaH “NO. II- 'II "I lull ll' V‘WHM HOUUDHUWGH mHO. oOO.. OOO. wNO. moo. In: In: manOHOMHHmO uOOOOHumSH HNO. mHO. moo. mHO. In: nun swaoolouomamm HouosuumnH oqo. oOO. BOO. moo. an: nun onSouonosoHumz wouoouumaH mHo. OHO. In: In: OHH xoSOmhm neuosuumaH woo. mHO. oNO. in: nu- SOHOHuonsoz wouosuumnH «HO. ONO. In: In: % OOHmuo>mHuxm HOOOOHumOH . r .m. m m m . .m. m . s o a a . anz a l s m a m o s t C e t e l r x c x . x _ C e r r r c C e S e d c.d m G E S E e E e .d e G pas .o e r 1m m m m .1 11 1 M l M m H mm m M m l m M... w as me Fm mm is my. 1G .Mx m... me as I F ml. F F F E E F HHH QH< HH QH< H QH< AemeH\AHomma as aoauosemm QH< mannawm> mMHm MOHUDMHmZH Wm mMHm MZOOHDO HmMSOU 2H :QmZHonmxm: ZOHH mo ZOHHmomomm OH mHmoH 226 Tables 10 through 14 indicate that there are differences in the structure of relations among the variables in accounting for variation at the secondary level. In the analysis based on entering-course variables, conceptual skills or orientation to conceptual thinking Show such secondary importance, particularly among those ',. in the mid-to-low GPA range. In the analysis based on mid-course factors, major field of interest and degree of general test anxiety (particularly among the middle range of GPA'S) appear associated with performance. In the analysis on end-of-course factors, degree of extraversion, focused among low GPA'S, has been added as having secondary importance. However, across the range of these analyses the tables indicate the overwhelming extent of the contribution made by factors associated with performance in past instructional situations, and the valuing of such performance. That is, coalescence in all analyses occurs around GPA and aSpiration-expectation for success in the course; in the final analysis, midterm performance is added to this combina- tion. (The importance of midterm performance can also be indicated by examining those factors importantein midterm exam performance; gradepoint average and performance on the pretest together accounted for the greatest prOportion of variance.) The importance of this com- bination of factors to final course grade can be seen in the groups in the trees of each analysis. Students high on this combination have overall mean course grades at the tOp of the course curve. This does not infer that others do not obtain grades competitive with those of these particular students. However, the trees indicate that far fewer do, and that it requires a series of advantages on other factors to overcome this deficit. 227 2. The results will show differences among the outcome variables (assessed student performance) in the type of predictor variables accounting for variance and in the extent of the explanation of variance. This expectation is an extension of the previous expecta- tion which was concerned with differences in explanation attributable to differences in contextual conditions. In the case of this expecta- tion, concern is primarily with explanatory differences between outcomes in terms of the inherent tasks underlying the outcomes. The concept of "task" was inherent in the discussion of the analysis of pretest and the analyses of course grade, in terms of role the outcome plays in the course -- for example, placebo exam or "run-through" experience, in the case of the pretest. However, in terms of the concept of "instructional process" develOped in the previous chapters, both pretest and final course grade are on the periphery of this process, although inextricably tied to it by the nature of their roles in the course. The outcomes of interest in this discussion are midterm score, the various sections of the final exam, and the discussion section grade, each representing a tangible post in the progression of the instructional process. 1. What accounts for the largest amounts of variance across the range of the different outcome variables? The largest amounts of variance across the range of examination performance are accounted for by performance variables. Table 10 (p. 219) indicates that for both midterm Score and final-1, the recall section of the final exam, performance variables account for approximately the same prOportion of variance -- thirty-eight per cent. In addition, Table 8, (p. 166) indicates that for both these outcomes about the same prOportion of variation is accounted 228 for in total -- fifty-four per cent. In chapter three, the similar- ity in content between these two tests was indicated. However, the structure of relations among the variables differs. In the case of midterm score, entering gradepoint average accounts for twenty-one per cent, and pretest performance an additional eight per cent of the total thirty-seven per cent contribution of performance factors. By contrast, GPA accounts for only four per cent, pretest for two per cent in the case of the recall section of the final exam. However, Table 10 indicates that twenty-seven per cent of the variation in final exam-1 is accounted for by performance on the midterm examination. To the extent that midterm performance does appear to have a component of GPA and pretest performance as a basis, this is reflected in final-1. In contrast, the applied section of the final examination, the variation in which is accounted for by slightly less an amount of a performance-variable combination (thirty-five per cent), does not reflect the importance attached to midterm performance or pretest performance reflected in the recall section of the final. Like the midterm, the greatest prOportion of variation is accounted for by an identical amount in the GPA -- twenty per cent. Au addi- tional ten per cent is accounted for by the OQT-total score. Further, unlike the midterm, and also final-l, seven per cent of the variance in the applied section is accounted for by a combination of attitudes as reflected in level of aSpiration, pre-attitude and discussion section attitude, and conceptual orientation as reflected in the learning set. The prOportion of variation in the instructor grade * 229 was accounted for by approximately the same prOportion in three areas. Tables 10, 12, and 13 indicate that two similar combinations of variables, student personality factors and instructor character- istics, accounted for 14 and 13 per cent of the proportion respectively, while an additional fourteen per cent was accounted for by midterm performance. ii. Are there differences between the levels of the different outcome variables in: a. type of predictors accounting for variance? b. the extent of prediction by both number of Splits and amount of variance explained? The trees in Appendix D for midterm score, final exam- recall, final exam-applied, and discussion section reflect the differences in type of predictor accounting for variance, and the extent of prediction by Splits and amount explained. Midterm exams and the recall section of the final exam bear a resemblance not only in amount of variance accounted for but in number of groups in similar branches, (Appendix D3 and Appendix D6) In addition, while Tables 10 through 14 indicated the differences in the struc- ture of relations among the variables in the two outcomes, and this is reflected in the branches of the trees, important similarities paricularly in initial predictors do appear. For example, in both cases, the performance of students with high GPA'S (who as a group have high mean scores) is almost totally "explained" solely by factors associated with "doing well" in school; by contrast, those with high midterm scores but not necessarily strong GPA'S are characterized by anxiety toward the final exam, and extent of ins- 230 tructor need for social acceptance -- higher final-l scores associated with lower scores on these two dimensions. By contrast, the trees for the applied section of the exam and that of instruc- tor grade appear quite complex. However, by comparison with each other they appear Similar not only in total amount of variation accounted for but in the extent of prediction by both number of Splits and amount of variance explained in the various branches. (See BSS tables, Appendices E7 and E10, and the trees, Appendices D7 and D10). iii. Are there differences between outcome variables similar in analysis conditions, but different in underlying task, Specifically: . a. between the final exam score-l (recall) and the final exam score-2 (applied)? b. between the final exam grade and the instruc- tor grade? The major differences in amount of variation accounted for, the coalescence of factors which are a part of that "account- ing", and these variations as reflected in the Subgroups in the trees has been discussed for both the recall and applied sections. However, on the applied section of the test the relationship among the variables appears much more complex. Whereas final-recall performance could be "explained" by high midterm performance, the roles of GPA and CQT'S, the variables appearing to have the most relative importance to determination of performance on the applied section, are more complex. Those Students who as a group had higher mean scores on the applied section were characterized by both high GPA and high OQT. Those students at the mean for all scores on the applied section were characterized by either high GPA or high OQT, but different characteristics distinguished both 231 groups. For those with lower GPA but high OQT-T's an orientation toward conceptual learning was an advantage which made their Scores competitive with those among the high GPA-high CQT group. Among Students of the reverse persuasion, (higher GPA but low OQT-Q), the factor characterizing Scores competitive with the high's in the above groups was level of aSpiration, higher scores associated with higher level. In addition, among these high scores, those with a less than favorable attitude toward the discussion section experience performed considerably better than those with favorable attitudes. By contrast with this group of high GPA'S, a group of low GPA'S expressed similar reaction on the recall section. That is, Students who had scores on final-l competitive with the highest in the course but who weaecat a disadvantage on other performance indices (C to D GPA, lower midterm scores) but had scored well on the pretest and were oriented toward conceptual thinking had less positive attitudes toward their discussion section experience. While the applied section of the exam involves a complex relation among factors in its explanation, its complexity is reduced when considered as a structural component of final exam grade which is the grade assigned the two sections of the test in toto. An examination of Tables 10 through 14 and the tree‘ for final exam grade indicate the importance of each of the major characteristics important in the "explanation" of the two test sections. Its comparison with the tree for instructor grade suggests the importance not only of the underlying task but the criteria involved in assess- 232 ment of that tasks in terms of the structure of relations which develOps among the variables which are a part of the instructional process. 3. Measured factors may together represent a weighting of their underlying theoretical constructs in the form of an interaction. The reality of the instructional setting is such that interactions should be expected. Given the number of students involved in the instructional situation under study, the highly heterogeneous nature of these students, the large number of instructional personnel, and the extensive number of objectives to which the instructional Situation was directed, the likelihood of interactions is probable. In addition, because of the lack of homo- geneity and independence involved, it would be expected that of the interactions which emerge there would be variations of three dimen- sions: (l) interactions involving levels of student characteristics to course outcomes and interactions involving levels of instructor characteristics to course outcomes; (2) the interactions involving positive instructional effects (results greater than expected) and negative instructional effects (results less than expected); and (3) the interactions involving different degrees of interactiveness -- that is, interactions might emerge independently of multicollinearity. 1. Are there examples of interactions between instructor characteristics and course outcomes? In Figure 3 (p. 233) observation of the plotted points indicates the presence of a strong interaction between levéis of instructor course load and student performance. It would appear that levels of instructor course load result in differences in student-achieved course grade. The plotted points indicate there is 233 FIGURE 3 MEAN EXPECTED INSTRUCTOR GRADE PLOTTED BY INSTRUCTOR COURSE LOAD LEVEL ACROSS STUDENT GPA/MIDTERM GROUPS 4.6 4.5 p O 3.9 (A) q) INSTRUCTOR GRADE w :1 a» Q‘ 3.} I_. LOW Midterm. Low GPA * Split Occurred LOW HIGH Instructor Instructor Course Load Course Load (9 hrs. or less) (10 hrs. or more) 234 TABLE 15 MEAN EXPECTED INSTRUCTOR GRADE BY LEVEL OF INSTRUCTOR COURSE LOAD ACROSS STUDENT GPA/MIDTERM SCORE GROUPING LOW-NORMAL HIGH INSTRUCTOR INSTRUCTOR Stuaent Sroups COURSE LOAD COURSE LOAD By GPA/Midterm Performance Y N Y N Total 3.725 302 3.883 230 Group 4 3.302 53 3.289 38 Group 5 3.618 110 3.735 83 Group 6 3.745 51 3.769 39 Group 7 4.093 86 4.508 63 Group 4: Low Midterm Score, Low GPA Group 5: Low Midterm Score, High GPA Group 6: High Midterm Score, Low GPA Group 7: High Midterm Score, High GPA High Instructor Course Load: Carrying More Than 9 Hours 235 a slight relationship between all students and level of instructor course load. However, there is a much sharper (interactive) rela- tionship between instructor course load and high performing students. Table 15 (p. 234) Shows the means, standard deviations, and number for each of the plotted points by category. In the discussion of question one of this chapter, it was pointed out that of the seventy-four per cent of the prOportion of variance accounted for in final course grade by end-of-course vari- ables, fifty-three per cent of the total proportion was attributable to score on the midterm. Consideration of not only the amount of variation accounted for in final course grade (in comparison to the analyses on the other sets of variables) but also of the extent of the contribution of the midterm to that explanation raises a number of possibilites. It will be recalled that in the discussion of the analysis strategy in chapter three and the discussion of results in chapter four the points were made that a variable may be used more than once in the analysis and that the Split eligibility and reduci- bility criterion were relaxed to create as many possible combinations as possible. This latter point creates the risk of small groups being Split off by chance rather than by real differences. However, the tree for final course grade on end-of-course variables (Appendix D11) indicates that in fact the Splits on midterm performance involved large groups of students across all ranges. 0n the other hand, it will be recalled from the discussion of instrumentation in chapter three that midterm performance as assessed by the examination counted as a twenty per cent component in determination of final course grade. In addition, attention to Table 10, (p. 219) indicates the importance 236 of midterm performance in performance on the recall section of the final exam. Further, Table 10 also indicates the importance of gradepoint average to performance on the midterm itself. This consi- deration would seem relevant to analysis of final course grade variance in terms of the fact that Table 8 (p. 166) indicates GPA alone contributed only nine per cent to the "explanation" of final course grade on end-of-course variables, when compared with the thirty per cent of the proportion accounted for by this factor in the other analyses of final course grade. To some extent this factor may be present in this weighting of midterm Score in the analysis of final course grade on end-of-course variables. The cruciality of midterm performance is made clear by an examination of the tree for midterm score.(Appendix D3) The tree indicates that while not all high scores were among the high GPA'S, the groups characterized by high GPA did have, inggneral, mean midterm exam scores at the tap or within a point or two of the tap Scores; any additional factors differentiating students in this group -- that is, factors which depressed scores of some by comparison to others with the same crange of GPA -- did not depress the scores of these students sufficiently to make them noncompetitive. That is, the students' midterm perfor- mance counted twenty per cent of the final course grade, and for those in the high GPA range no factor depressed their twenty per cent to the point where it was not an advantage as a component in determining final course grade. Figure 3 (P. 233) indicates that advantages in instructor grade accrue to those Students with instructors with heavier course loads, and is a differential advantage to those 237 students characterized by high midterm, high GPA. It is possible that this differential is merited by performance and by knowing "the name of the game", as reflected in GPA, in terms of grasping what is necessary in the discussion section to insure SucceSS.‘ However, the interaction raises the possibility that instructors with heavy course loads used the midterm score as a preorganizer in making discriminations among the students' performance in dis- cussion section. The possibility of this factor's importance must be judged in the light of the fact that final course grade was a composite of twenty per cent midterm, forty per cent final exam performance, and forty per cent discussion section grade, based in the end on rank with all other Students in the course. ii. Are there examples of interactions between student characteristics and course outcomes? In the discussion of final course grade based on entering- course characteristics in chapter four, attention was drawn to an interaction involving course-Specific motivation with GPA. Figure 2 (p. 173) suggested that the plotted points of aSpiration-expectation on course grade for the entire Study group indicated that there was a regression, However, the regression rather sharply changed its SIOpe at approximately the mean values of both indices, suggesting an interaction in addition. The evidence supporting regression plus interaction was further exemplified by examining the plotted points for the three levels of GPA occurring in the three initial Splits of the data. The points for high GPA involving more than ten cases follow a positively accelerating path; the points for middle GPA 238 FIGURE 4 MEAN EXPECTED FINAL COURSE GRADE PLOTTED FOR VARIOUS GPA, MIDTERM PERFORMANCE LEVEL GROUPS ON END-OF-COURSE ANALYSIS BY ENTERING-COURSE ASPIRATION-EXPECTATION FOR COURSE GRADE onemeommxmuonaamHmma moemo ammooo mo am>mq AImv H+Ov HOV AIOV Almv fl A+mv fl Amy m If m h o L L! F I HMPOB ‘\.EuouoH2 amoz m>OQoo< uoo< .«mo HH< "m OSOHO ouoom EHOOOHZ coo: o>oo< " High GPA. High Midterm 3.5. 3.0< ‘ * Split Occurred Low HIGH Instructor Instructor Eysenck Lie Eysenck Lie (0 - 2) (3 - 7) 259 TABLE 20 MEAN EXPECTED INSTRUCTOR GRADE BY LOW'AND HIGH LEVEL OF INSTRUCTOR EYSENCK LIE SCORE ACROSS GPA GROUPINGS LOW INSTRUCTOR Student Groups EYSENCK LIE By GPA/Midterm HIGH INSTRUCTOR EYSENCK LIE Performance §: N T. N Total 3.746 346 3.881 185 Group 7 4.286 91 4.241 58 Group 6 3.585 65 3.800 25 Group 5 3.636 121 3.718 71 Group 4 3.131 61 3.633 30 Group 7: High GPA, High Midterm Group 6: Low GPA, High Midterm Group 5: High GPA, Low Midterm Group 4: Low GPA. Low Midterm Low Instructor Eysenck Lie: Score of O - 2 on Socially Desirable Response Scale 260 appears to be Slightly less of an advantage among students of high GPA but low midterm, and to be a slight depressant among students of high GPA and high midterm. By contrast, this instructor predis- position is a Slight advantage among those of low GPA and high mid- term, and appears to be a distinct advantage among students of low GPA and low midterm. Table 20 (p. 259) shows the means, Standard deviations, and number for each of the plotted points by category. Further evidence of the role of the instructor in determina- tion of grade is suggested by student reaction to the discussion section experience in comparison to instructor grade. In the above discussion, reference was made to students of average to high GPA but low midterm scores who received higher instructor grades from instructors characterized by more rigid behavior than did comparable students with instructors not so characterized. Among the former, a less favorable attitude toward the discussion section experience was associated with lower mean instructor grade. A Similar differenti- ation on instructor grade appears among students of high GPA and high midterm, a lower mean instructor grade associated with less favor- able attitude, although fewer students are involved prOportionally. By contrast, an examination of discussion section attitude as asso- ciated with test performance among certain levels of students indi- cates the reverse effect. In the previous discussion of the applied portion of the final exam it was indicated that as a group, the high scores were among those having-both high GPA and high CQT'S. But among those of higher GPA but lower CQT'S, those with higher level of aSpiration were put at an advantage, and among this latter group those with less than favorable attitudes toward the discussion section 261 experience had even higher mean scores, competitive with the tOp scores on the test. Further, in the previous discussion of the recall section of the final exam it was indicated that among Students of low GPA and low midterm scores but characterized by high pretest scores and an orientation toward conceptual learning, scores on the recall section of the final were competitive with tOp scores among all students, among those also expressing a less favorable atti- tude toward the discussion section experience. 5. The results will Show isolated exceptions, normally relegated to meaSurement or statistical error. which ggpresent either important exceptions to general predic- tion or real instructional problems. This expectation is an extension of the concept of charac- teristics as barriers/facilitators, discussed above. In the discussion in chapter four the role played by chance fluctuation in the creation of small isolated groups, due to relaxed eligibility and reducibility criteria,was indicated, along with illustrations such as those in the tree for midterm grade, where one or two individuals are Split from a large number of Students, particularly late in the tree, and such groups appear nowhere else in any analysis. (Appendix D4, groups 54 and 55, groups 50 and 51, 52 and 53). Such cases were differentiated from those created by the nature of the outcome variable -- Specifi- cally the case of the instructor grade where the criteria for assess- ment of grade varied across instructor as did grading level. (Appendix D10) In addition, there also appear in the trees over the timSSpan of the course small groups isolated on the same variables For example, although the trees indicate that over the timeSpan of the instructional process general test anxiety Operates among the middle range of GPA as a depressant to performance on tests, and the 262 interaction in Figure 5 indicates that for Students in toto higher general test anxiety is associated with depressed grade, there is among those with high GPA (who as a group are characterized by a low to normal range of general test anxiety), a group of sixteen for whom very high general test anxiety is facilitative, resulting in high performance in the course. An additional example is provided by the Student extraversion continuum. Figure 6 (p. 244) indicates that students characterized by low midterm Scores and average to low extraversion have low course grades. The pretest indicates that among this group of students those characterized by lower COT-Q's still have pretest scores above the mean. An examination of the data of the applied section of the final exam indicates that their mean Score is higher than that of the more extraverted Students with low GPA. However, over the timeSpan of the course this group's performance is increasingly depressed, and it appears that the problem represented in the interaction of Figure 6 is even greater for these Students in terms of overcoming barriers to performance and of using their enter- ing-course abilities under these instructional conditions. 6. To some daggee the data will provide an estigat§_of congruence between course goals and conduct of the course. The questions posed in the study and the design to provide answers to these Questions were not conceived as an evaluative com- ponent. However, to the extent that it was necessary to consider the course "model" in the study design, an estimate of congruence between course goals and conduct of the course can be inferred, based on a reexamination of the considerations in the previous questions and a further examination of associated data. The goals and assumptions 263 were discussed in chapter two, and Summarized in the discussion of this expectation in chapter three. i. was educational Opportunity a part of instruction so that entering characteristics of students would not be the main determiner of degree of instructional impact (performance outcomes)? The goal of provision for educational Opportunity was based on the assumption that Opportunities would be provided for students to overcome barriers created by entering characteristics. The extent of the impact of entering-course characteristics, discussed in the previous sections of this chapter, can be Summarized by the example in Figure 9 (p. 264). Pretest performance operated as a consistent "facilitator" or barrier (depending upon student location in the plotted points), with a slight indication that the factor may have operated selectively across GPA. Table 21 (p. 265) indicates by category the means, standard deviations, and number for each of the plotted points. As representations of the course over time, the data in the trees indicate the extent to which GPA maintained Strata across all outcomes. In addition, the data indicate the extent not only of student heterogeneity but the extent to which characteristics Operated differentially among student levels, having their major effect as facilitators in the upper ranges of GPA, and as barriers in many cases among lower GPA levels. One of the beliefs underlying the above assumption was that common exams would not create an atmOSphere based mainly on compe- tition for the high exam score. However, exam content was recall- oriented, based on Specialized information drawn from disaparate lectures and readings. In addition, it appeared that in reality the 264 omausooo uHHmm an am msoaw m msoaw v msouw Ho.eum.mv AN.NIm.NV AN.NIH.HO amo.mon «me wanna: «so zoo :HNIOV .m.m9 amoumum 30H W Ammummo ammumum roam .e.e Hm>mq MUZNZmOmmmm EmmBMmm wm Hm>mq 4 x . I r . . 4 (x r‘ n o r . — : 1 f . . . f L. .. / . . v n «A . > I \I I. . y. . \¢ » l \ ‘1 . . .\ r . I , . II. P . _ . _ I- y - I .l . . t . l 1.. \ 5 O I u a v a \ \, \ I‘u I V - I. 14 ill. 1 . . . . . A 74v . . r. 1 . ) . - , . .. x . . r I r, . \ .l . I . f. ». .. . \J ‘ i . . .. . . 7 I .5 I y. y . . . . r . . . ‘1. I . \. . (I ~ i p a \ C . o. p I I I . V P. I I \ \ .l _ F u . , J . . x \. ~ , i . 1, Ir I 1. .9 . u IL A I D l I t t O a ‘ O O n. ,1 . - l: . . . t I o a n t A I I I 9 A; . x \1 « . . . . II r s . A i . a r . n p \ x . . I; . APPENDIX C INTERCORRELATION MATRIX OF STUDY VARIABLES Pretest (1) Midterm Sc. (2) Midterm Gd. (3) Final-1 (4) Final-2 (5) Final-Total (6) Final Ex-Gd (7) Inst. Gd. (8) Final Cs-Gd (9) Credits (10) Current Ld. (11) GPA (12) CQT-V (13) CQT-Q (l4) OQT-T (15) Extravert (16) Neuroticism (17) Eye. Lie (18) A-H Anx. (19) Test Anx.-l (20) Test Anx.-2 (21) Test Anx.-3 (22) Accuracy (23) Success (24) Learn. Set (25) Reason En. (26) Pre-att. (27) Post-att. (28) Disc. Att. (29) 50-50-1 (30) 50-50-2 (31) Inst. Ext. (32) Inst. Neu. (33) Inst. M-C (34) Inst. S-G (35) Inst. Cal-F (36) Inst. Risk (37) Inst. Load (38) Inst. Teach (39) Inst. ED200 (40) 315 TABLE Cl INTERCORRELATION MATRIX OF STUDY VARIABLES (1) 1.000 .290 .293 .282 .248 .297 .326 .151 .308 .094 .011 .247 .308 .145 .286 -.110 -.017 -.121 -.216 .067 -.137 -.110 .005 .053 .216 -.063 -.O2l .019 .003 .207 .283 -.018 .097 .044 -.006 -.061 .002 -.031 -.061 -.089 (2) 1.000 .918 .503 .412 .526 .532 .382 .672 .065 .049 .458 .302 .201 .292 -.096 .050 -.010 -.259 .029 -.112 -.176 .110 .135 .230 .037 .041 .096 -.064 .255 .601 -.056 .029 -.035 .037 -.073 -.053 .042 -.136 -.091 (3) 1.000 .475 .348 .495 .502 .372 .671 .078 .093 .470 .243 .194 .268 -.081 .063 -.070 -.245 .031 -.090 -.162 .084 -.O45 .205 -.010 .001 .132 -.089 .257 .680 -.035 .008 -.006 -.006 -.076 -.053 .041 -.124 -.077 (4) 1.000 .511 .864 .824 .319 .730 .142 .030 .392 .261 .250 .302 -.061 -.017 -.076 -.244 .024 -.l68 -.119 .010 .083 .212 .018 .019 .043 -.075 .217 .414 -.034 .047 .036 .056 -.044 -.116 .020 -.098 -.142 (5) 1.000 .858 .806 .282 .655 .053 .053 .452 .359 .339 .420 -.l3l .038 .126 .256 .042 .116 .078 .051 .057 .253 -.024 -.017 .051 -.043 .235 .354 -.007 .043 .053 .001 -.035 -.O78 .059 -.085 -.120 (6) 1.000 .932 .343 .790 .106 .057 .484 .348 .332 .405 -.116 .004 -.104 -.286 -.009 -.159 -.116 .050 .089 .258 -.011 .001 .050 -.056 .257 .439 -.016 .041 .054 .034 -.046 -.119 .051 -.098 -.156 (7) 1.000 .307 .781 .101 .054 .474 .341 .335 .402 -.O92 .028 -.123 -.281 -.014 -.161 -.138 .049 .097 .269 -.004 .009 .066 -.085 .271 .452 -.025 .064 .086 .048 -.032 -.129 .025 -.069 -.l63 (8) 1.000 .642 .111 -.004 .371 .101 .134 .113 .037 .025 -.073 -.091 .005 .060 .088 .012 .038 .038 .063 .019 .045 .002 .163 .310 .090 .141 .068 .004 .004 .031 .129 -.074 -.147 Final Cs-Gd Credits Current Ld. GPA OQT-V OQT-Q OQT-T Extravert Neuroticism EyS. Lie A-H Anx. Test Anx.-l Test Anx.-2 Test Anx.-3 Accuracy Success Learn. Set Reason En. Pre-att. Post-att. Disc. Att. 50-50-1 50-50-2 Inst. Ext. Inst. Neu. Inst. M-C Inst. S-G Inst. Cal-F Inst. Risk Inst. Load Inst. Teach Inst. ED200 (9 ) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40) 1 (9) .000 .141 .065 .547 .329 .302 .362 .108 .009 .095 .254 .014 -.142 .143 .035 .071 .226 .007 .038 .062 .085 .260 .526 .023 .023 .044 .063 .042 .107 .071 -.127 -.159 TABLE C1: (10) (11) 1.000 -.O62 1.000 .052 .158 .080 .120 .171 .101 .125 .138 .009 -.043 .034 -.003 .008 -.043 .078 -.123 .018 .018 .010 .027 .064 -.039 .009 -.021 .029 -.O72 .067 .023 .020 -.032 -.153 -.058 -.068 -.017 .044 -.064 .121 .008 .106 .033 .070 -.088 .014 .061 .035 -.024 .029 .052 .027 .019 .021 .008 .016 -.012 .044 -.O2l .103 .005 316 (continued) (12) (13) 1.000 .365 .000 .334 .326 .424 .793 -.209 .199 .035 .034 -.026 .075 -.390 .394 .050 .035 -.126 .179 -.178 .167 .071 .035 .089 .088 .208 .257 -.019 .043 -.101 .125 -.086 .042 -.163 .118 .472 .266 .442 .246 .044 .082 -.026 .059 .094 .003 .038 .053 .035 .038 -.119 .033 -.003 .029 .002 .109 -.O9l .029 (14) 1.000 .709 -.077 .017 .002 -.239 .054 -.138 -.109 -.059 .004 .126 -.038 -.078 .024 .010 .247 .212 .052 .006 .079 .023 .025 -.048 .019 -.055 -.l40 H (15) .000 .185 .044 .068 .415 .006 .208 .217 .030 .049 .262 .061 .143 .031 .080 .290 .268 .039 .048 .029 .015 .028 .021 .005 .094 .089 (16) .000 -.138 .155 .076 .031 .042 .049 .019 .042 .084 .026 .034 .066 .012 .045 .050 .079 .069 .006 .051 .040 .057 .047 .068 .013 Neuroticism (17) EyS. Lie (18) A-H.Anx. (19) Test Anx.-1 (20) Test Anx.-2 (21) Test Anx.-3 (22) Accuracy (23) Success (24) Learn. Set (25) Reason En. (26) Pre-att. (27) Post-att. (28) Disc. Att. (29) 50-50-1 (30) 50-50-2 (31) Inst. Ext. (32) Inst. Neu. (33) Inst. M-C (34) Inst. 39G (35) Inst. Cal-F (36) Inst. Risk (37) Inst. Load (38) Inst. Teach (39) Inst. ED200 (40) (17) 1.000 -.221 .238 .045 .222 .193 .068 .084 .066 .154 .086 .008 .034 .069 .085 .005 .004 .032 .007 .024 .038 .047 .013 .060 TABLE Cl: (18) (19) .000 -.014 .000 .079 -.024 .058 .314 .058 .322 .034 .062 .075 .137 .076 .207 .017 .052 .052 .043 .002 .059 .046 .074 .005 .392 .023 .353 .017 .016 .017 .019 .019 .034 .101 .017 .070 .023 .081 .025 .019 .035 .014 .029 .027 .029 317 (continued) (20) (21) 1.000 .045 .000 .045 .537 .059 .034 .007 .062 .025 .144 .016 .019 .054 .048 -.015 .079 -.123 .023 .009 .187 .053 .167 .013 .012 -.048 .082 .067 .036 .039 .003 .027 .010 -.019 .045 .020 .061 .059 .016 .059 ..008 (22) .000 .030 .186 .115 .056 .069 .053 .030 .213 .206 .014 .004 .033 .088 .057 .048 .029 .048 .026 (23) .000 .383 .022 .069 .131 .099 .105 .003 .071 .026 .059 .006 .054 .029 .050 .028 .023 .006 (24) 1.000 .036 .008 .113 .024 .098 .147 .176 .004 .013 -.007 .055 .025 -.058 .012 -.019 -.013 Learn. Set Reason En. Pre-att. Post-att. Disc. Att. 50-50-1 50-50-2 Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Inst. Ext. Neu. M-C S-G Cal-F Risk Load Teach ED200 Neu. M-C S-G Cal-F Risk Load Teach ED200 (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40) (33) (34) (35) (36) (37) (33) (39) (40) (25) 1.000 .015 .024 .050 -.084 .185 .189 -.O72 .123 -.031 .083 .031 -.076 -.103 -.017 -.059 (33) 1.000 -.253 .493 .055 .093 -.332 -.129 -.129 TABLE Cl: (26) (27) 1.000 .394 1.000 .174 .391 .091 .121 .074 -.018 .067 .074 -.001 .004 .055 .028 -.025 -.058 -.038 .035 -.056 .011 .010 -.029 -.085 -.052 -.029 -.022 -.021 -.049 (34) (35) 1.000 -.271 1.000 -.012 .528 -.388 -.411 -.143 -.004 .396 .016 -.299 -.O96 318 (continued) (28) (29) 1.000 .161 1.000 -.O69 -.103 .072 -.063 .020 -.038 .096 .085 .074 -.O78 -.037 -.022 -.040 -2108 .046 -.037 -.089 -.016 -.037 -.112 -.O72 -.055 (36) (37) 1.000 -.179 1.000 -.074 -.004 .453 -.227 .453 -.227 (30) 1.000 .465 .078 -.006 .050 -.032 .018 -.032 -.012 -.006 -.058 (38) 1.000 -.415 -.278 (31) (32) 1.000 .035 1.000 .023 -.632 .054 .432 -.023 -.565 .024 .037 -.049 .151 -.005 :068 -.039 .166 -.003 .166 (39) (40) 1.000 .095 1 000 APPENDIX D FIGURES D1-D11: FORMATION OF STUDENT SUBGROUPS 0N SETS 0F PREDICTOR VARIABLES OVER EACH OUTCOME 319 FIGURE D1 AID I: Explanation of Pre-Test Score by Entering Course Variables ‘ I M OQT-V: 2 4 1'24. 172 Id-J.081 Ian-17.31 r2.11-11 All. lubjccll 111 and, 742.205 404.365 r1001 (5) tat-V: :18. 750 -1.186 94.51 (20) A4 1:0: Aral-Ky: Y:25; 033 156-1.588 (27) A-l Tau Anxiety: 27 47 743.194 4:14.937 .67. 6 2 _-l - 2.9 241.735 444.248 046.21 *(69) Learning 50:: turning Sat: . 1 747.400 «4.091 u-2.81.-15 “(06) Lannie; 5.x; 742.000 40-.000 n-.1..-1 ‘(67) learning 54:: 14 - 14 744.793 444.200 5.5149 (38) Sex: a $24,000 ad-2J49 n-O. :(39) 54x: gala Y-Zl .588 ad-Z .402 n-3 . 21-1 7 41) 001-2: 243.491 404.445 134-10.01 n-J.21 17 (50) lynx-ck 1.10: 745.417 “4.019 2.37.42 4m) Byunck 1.14: 742.927 404.321 1 (40) 002-2; 140. 444 441-1. 707 7.7.4 (8) lea-0n lazuli-d: l Y-u.144 u-3.524 04.72 (9) lesson lnnlladx $41.53: ad-3.139 1:46.27. 320 (45) burning Sat: 744.393 444.0,: 1,-5.3! ‘(44) turning 50:: _a_ - 12 (32) Major gods: 1,1,4,5,6,7 Y‘ZJ l4 sci-3.469 OJ) mjorga Y-19. 333 ld-l.364 .1-61 (11) (121-V: l4 — 38 Y-21.l)1 Od-J.079 n-JIJI (10) an." 3 - __ 13 140.721 044.050 n-XTJ‘l 4(13) OQT-0: J7 - 40 743.409 444.023 ”4.1242 (12) GIT-Q: 1 - 56 141.409 $3.052 .5561 16) Ill)“: Sande 2, “21.685 ..Id-J 48 N. 71 (1 .056' 84.526 ”NI!“ 004013.4347 1-20 '11.“ (58) hernia Sol: 13 740.500 444.500 .424 59) hauls; Sat: 1 - 344.077 904.960 lav-4.91 (63) QT-Tx 31 - 37 Y~Z5.636 “31,267 n-2.11F11 '(62) 002-T: ; - 20 142.933 ad-Z.886 n-2.SI-1$ n-IJ.31 ) A-l In; Amory: 15) 50:50-1: Y-ZI. 1025 od-Z. 745 2.11 ('0) A-l Tat: Anxiety: - 33 740 917 44,220 n-6 .57. (1‘) HIM node: 3,4 1'21. 759 Id-SJJZS .1 ‘(29) Major: soda: 3 7 F193;}; ad-J.156 (73) turning 30:: Y-12.900 Iii-l. 700 n-l. 91-10 [72) huntn. Sn:- 0- 2520.154 Id-1.195 M4. 91 '(64) A-l fur Anxiety: 743.667 444.145 n-l. I'D-O 14) 50-50-11 1 - 7 Y-ZOJ“ ad-JJoll r11.41 PG” Ad! In". Amery: 5 o 1.11.706 “ILIN .J.21)-l7 ‘6‘) GIT-Q: 1 .. 15 144.400 444.242 04.0245 '(55) WT-Q: 1 F20. 100 841-1. 7! B-l. 927-10 (0) ka-Aturnda: _2_0 - 32 !-20.098 Isl-2.986 n-7.71 K42) Pro-Attitude: g3 - 24 MJZO sci-8.880 mafia-1 (75) OQT-Va 28 N4. 500 Ill-5.000 n-.47.-2 (18) lyunek Neurotic: 2 - 11 740.410 «4.954 v8.31 FIGURE D1 (19) lyunck Inn-one: 1518.501 lei-4.454 H.“ 400) lyunck Hun-0:101 2 542.000 40-.000 “.224 4,,75 7,8 1-19. 989 Id-2.1OJ I'L 101-18 (”Hts-:1 3, Mama-000 ‘71 if“? 5'" ;- 21:045 '40-2. 720 "51‘“ 741.023 041-2.“! ""7 114.12 '01) lyuntk Neurotic: - 11 I016) turning 50:: *(53) Major $41,541 “1.2.345 1 god“ ”4.57924 «34) gar-v: $49,700 net-2.147 4-3 .0240 '05) CIT-V: 11 - Y-IZ.5D $7.27 n-,87.-4 ‘ (J7) Ian-on Enroll-d: 345.833 424.951 64.124 *(Jé) lesson Enrolled: -A N2. 692 “II. 424 I'D-12. 2H5 (60) Iyunck Extravert: 742.214 4:14.020 u-Z.67a-14 (61) lyunck Extravert: 14 - 19 749.930 «4.051 “LOX-16 321 FIGURE D2 AID I: Explanation of Final Course Grade by Entering Course Variables ) A11 aubjaczg in study (4) 322 (22) 50-50-1; g — 10 14.330 46-.564 1.19.4040) (19) my“; code ,3,0, 8 sci-,5” 24 _ : 2 Y-3.B75 n-4.51F (6) Major: *(37) harnin; Sat: soda: 1,5 16 - 18 Y-J.800 161-.651 Yd.250 “$433 1:42.27. nun-4 '(36) 141qu 5.1: g - 25 Y-3.304 sci-.621 . 1'13 (10) GIT-Q: _4 . 22 Y-3.746 Id-.689 (l7) cur-Q: 1141.97. 74.569 Del-.692 69-21.“ (10) A-l Tut Anxiatyx 1 - 7- .540 ld-.6‘I7 (7) 11:16:: ”’3'“ 0‘) OQT-0: '12:) 00M: c0da:234673 _- anus 74.445' 1:13.705 “3.083 Id-.640 Y-J.388 uni-.566 047.52 r2.3%-12 64.2%“ (11) A-B '1'“: Anxiuy: $4.909 sci-.596 “.174-12 6 (v) cur-o: (as) 321111“ 2 1 35 — 49 -; - Y-3.4l49 40-.574 zingng-sn u-9.27.-49 ' l4) hjaf: gods: 1,8 Y-J.324 ld-,700 n-7.01 '(34) GPA: ‘09) Burning Sat: 1 7 J 7'2-929.ld-.703 24.000 46-.ooo 2.6244 ”.227-1 (15) lynnck lam-Ivan: “(357 161-Arnie. Sat: . _ . 74.125 .6 .556 114.077 4.1._474 1’) 1.7.51 - 64.31.49 h soda :4 J“! 1 : ,3,4,5,8,7 919 Idl.668 (33) T-V: ; - 49 2-3 667 ul-.471 6562-3 (24) lyunck Int-van: ~ _2_ . 14 Y-Z.757 [Cl-.710 n-8.11 4(32) 002-v: 1 - 32 FZJOO ld-.678 r7.5‘b-40 FIGURE D2. 6(27) arr-2: 24 - :0 . 64 46-.771 64.1241 ' (3!) 021-32 74,625 nd-.564 .ow 26) 002-1': - 23 13.572 n-9.61 *(13) GIT-Q: 4 54.722 6.1-.193 p.624 323 FIGURE D3 AID II: Explanation of Mid-Exam Score by Mid—Course Variables 324 1' ) Pittelt: 22 - 3: as 3.21% '19.9: (J) GPA: 5.5 - A.o 1-33.534 . . ”27.31 '(anynmk lnnv‘rtt ~(51) 1 001.0, 1531. 79s 1.1-2.257 ' ‘5 21. 61- 1: 33, 550 n-3.sz-zo .05) "“3." at) 611-9: 1, a l . 76:. 000 1.1-1.531 1-01.730 ld-2.935 ~ 1: (H) 11.3“ I“'21-“ *(Z6)lyuml:k hut-w": 5°69: 56. . 78. 2 (22) 131.075 111-3. 230 117-24.000 sd-.ooo Prue“: (50) '12- 75'. R‘JI-l 1s - 21 who > :53; gig 3 ‘ 2 (25) 931-4. 771 ‘1-30,le “ - . “1:25:30 ‘3 2 128.175 nd-z.034 - ' "'1'" 113) cm 1.11-6 1.11 - 2.7 F31.073 «14,230 1.43.61 (11) oar—r: (37) 2041311: §-;0.967 111-3309 $30.30; 1.1-3.112 (5) 3.35.1.1. 1140,21 A-E Tut Au misty- “) I. - 1.5 All tubj-ctu Y-SDJH in nkudy ad-3.391 1.20.711 “'35-” «(36) 30:52-1: u-3.363 $23,750 044,763 n-1ooz nan-6 (5) Pure“: 31 - 33 1-3o.357 Id-3.6H *(34) Major: *(49) 11.2.2112? 5:2: -u..71 _ - n 1-31. 111 8111-2. 555 2-20.369 (7) u- 3. 41-1 13 n-3.ez-19 A-E Tel: . Aux luxury: 26 ' 53 '(40) A-H Tut Anxiety: Y-28.563 .. - -3.sa3 t-30.655 141-2.5m ‘ n .01 *(As) mums; Sat: n-SJZ-Zs . 1-27 033 ad-a. 373 7.24.142 Id-J.27c 1.31-7 (30) nyunck 1.1.: (1) Y-29.667 «14.404 GPA: n‘9.01 1.1 - 2.7 F29.630 ' 18 12K yz: ctor 1.1-3 934 (a) or. my) gar-v2 ‘ ) “' “ Mk “mm. (29) new": (1.1) 71.11 1.:11 Amuy: ‘ 7 - 9 _ - _ r7211 coda: 1.2.3. Y-JAJOO “_l. g.” 000 111-1 927 ”29‘109 .¢.3_597 Y-Za. 1589-91-19. 110 - ' ' . 1. 31 1 3. 6- ’17-29. 163 n '51“ n-l.J7.-7 n-12 1 ( )éyrlrk 12.: n- 1.1 747.1133 Id-3.Ao9 '16- 2~ n-3.oz-16 (b) 19 In-cruct or . 30:21:“ (16) for-3'; ) y" nck nun-n: (2°) 00'5"; '6") 3‘s“? 123.223 =5211.902 tel-6.000 7.331320 111-3 936 123.153 171-3 706 $25300 1.1-3.210 .d-k.)_30 was.“ ”15:11-12 37- .-.51.-a (55) Puzuutudu 25‘” 741.272 «H.430 n-2.11-11 (52)!“tructnr (32) 001-0: (‘2) A ' (“flu-tr . (9) or 1 . 12 5020162123 :1 3 073 2 ”fluff." “alum S:n§otd-ch5h. ‘ 1747.220 sci-3.239 v- 7. . - . ._ soda: 4 5 s _ 4.37. - _ 1:49.333 ad-2.809 F26.927 ' “ 9-‘1 “ 3:53;" .a 2.379 114.51 ld-JJSG . 9‘ 6(0) Ass: «59) Pajama“: 7.2.3: Loo .d-1. :5: 1127.692 1.1-1,997 n-.91F5 “Bin-tructor fi(53)!n::fru:§-o£° h: n- 2 .61- 3:“ Incl: 211::an 10 _ 13 "5 125.100 “1.2.300 :fi‘éfif‘m656 run-1o - FIGURE D3 325 FIGURE D4 AID II: Explanation of Midterm Exam Grade by Mid-Course Variables (3) (1) All luhjnctl L11 ntudy 13.502 “kl.” 114001 326 GPA: 1.1 - 2.7 FIJI! 11.72.12 "(13) 001.17; (23) Iutxuctm- Sanford—Cough; l} (2100 .4. 00° 2 - 13 Y- Z 2 ' Y-1.571 1.1-Jag 11.. 1'- 1.43.22 719) oar-v: '17-3. 51!. “1.31, 97. '(22) In::xuctor |"(37) Pra-Atncudn: go . 320114-6011“ Y-3.553 ld-Joz 2 “.51'15 1:13:22 MI.“6 '(31) A-l Tull 11x1. A 19-20 321-.701 n'l.lH * 54 Re. Ion Enrolled: 'i-s. ooo , M- 000 n-. (01-2 '(36) Pto-Acttcudc: ._ ' 9 114.070 ““759 ”.111-A3 “51) 021-1: 19 - 22 174.800 F.91-5 (60) CDT-T: Y-J.6:7 n-2.-311 sad-.699 (5°) (QT-01 nu) Ago; (13) 3 hum: 1'4: 666 .a-.573 23 — 32 -75 741.595 1.1-.630 n-lS.BZ '(43) A02: 6(6)) Preten- _ode: 14,510 _ 22 14.000 -d-.a1e v.1 31:11i Id-,SB\ 11-1,77.-9 “_3' 6'- 29) Le-rning s“: «52) Page”. 3 ‘ 3" e . 19 14,171 -d- 701 “3.2“, “a.” n-3.51-19 1 I _. 1.4. (12 ] Pun-1 16 - 22 14.079 394-. 3 3) 1:41.01 9 Major: , 59 1.2; coda: 2.3.0 ,S,6 ( )0) 1- -.3 051. .4— s11 17.-4.1053 “EA” n-a 32-u. M 9 _ (A8) burning Set: *(58) EDT-T: 24.529 011-.794 1-3 250 -d-.601 m3,” an}. 01-1 21) 50—50—1- ’(26) 1111:1032:- 15 - 30 %:i;5,_ 935 “”327 74.250 Id-.829 " [14.37.45 (7) Eysenck N—uxatic: ' ’1 «(27) Inauuctor. 113073 ud-.3ss 111-2: 045.02 32 - 53 $3,753 Id-.770 ”9.27.49 3%?“ i110) Pre—Antitude: $3.550 10'- 18 d- 932 y—..ooo 0d-.000 ‘ . 7 u-.52-3 (15) Prev-Attitude: 19 - 4 14.553 -d-.937 (6) lylenck "auntie: 14.31 1 . F1399 111-.917 - “'25." (29) Bylcnck neurone: - 12 14.25:. .d..779 n-12 52 '(28) Eysenck Numeric: 11-2 0297 Id- 720 1143 30) Instructor €111.10 n11.4- _ -9. v 3.612 d . (15) nyunck 1.1.: “:9 21 ' ' M3 0 . 1::6AE72. ““906 *(31) Instructor ' California-F: _1_0 - 11. 1o) 11., , $2.327 171-1.030 god“ 1.2.3.5,7 Y 3 275 ul- 9110 *(57)A=-B Tu: Anne“: -63 7:3. 0002 0n.1-.7o7 17) Byunck 1.1.: n-3.B $2.792 .a-Jss 114.51 "50141 res: Anxiety: 26 - 30 Y-lJSO nal-.319 (A) . Pun-t: n .8140 12 - 20 WJM 06-.933 n-25.9‘l. 33 1-3. 071 .11. 7°, .91.“ (35) Inn-1:. - ~20 1:2: 918 od- 806 .21 (45) 0214): 11-3. 023 1.1-.762 5.7.1 ”.1 50401156,! ‘1-2. 010 :d- 31.0 *(M) Prue-t: lb .. 15 74.221 N.,-[35 “'1 . 71-9 . NM) OQT-Q: 1 - I. 74.157 1.1-.531 11-1.17.-6 FIGURE D4 n-J.‘1P13 ‘(557 lawn Enroll-d: Y-J. 59‘ ld".701 M. 327 FIGURE D5 AID II: Explanation of Final Course Grade by Mid-Course Variables (I) (ii All subject- ln Itudv III-72.11. '(13) 50- (5) an: 1.3 - 2.7 [-3.555 I 111-39. '71 1 P‘Ierull' £2 . 1: Y-L,-~’oe 04-45312 n-19.8'. N291 paler: so»; 7,4. nal-.000 ':‘-A_000 u-l.9"¢-1\7 (J9) OQT-T: Pte :231': 16 Y-I: 976 212-.1106 328 0(3‘1‘ A-fl VI). Auxerv ~00) 71-11 1“. 1.111.: 4, ' m «1.11 m- 352 n-lb “~51 nun Lanna; 5.1: 114 . '1-43300 111-293 n-e.27.-33 .35) OQT-1’: $3.393 03-.552 M7,?! N48) tanning Sat: $3,200 m. .00 upJ‘L-S (13) burning 50:: '(35) Pro-Autumn 30 - 32 22- - 1-4 000 Id-.b82 “5.500 -3-.soo n-S.17. n-2.37.-1z '(22) mung SIR: ' 16) Emmi Lt- ?-3.eAJ .4-.s1o n-1o.51-56 $3.790 ind-.666 ”10.51 P20412120“: 12 - 29 .(6(2) '11" tructor C.: 10) HHUK- (17) Eyuuk 1.1. gods: 5,6,7 3 1 - a “”1“” v-3.1 '17-3 we .0- so: ' ' 17-3. 750 0.3- $36 n-ZS n-S.6'L 5.3 57.- I(k7) Inurucmr (6) Ad! 12;: Anxiuy: c112 far 1310 h . 6» Y-zl 3. 6197 Id-, 672 n-32.0 (11) Mun: l(51) (311.17; Y-3 133 04-.599 code: i.“ __ - '15 Y-J. 33! Id-.3se Y-u.509 Id-,500 $6.22 n-Jd-Z '(50) QTIV: 1 - 30 Y-3.256 III-.506 ’(32) KIjar: n-5.31.-31 50a: ,5 YI3. 658 blur-.662 (“5 A.B 12;: Luxury: n- 3- “’1 Y-3:17! “I, 120 .71 '(55) hunts; SII: -31 V»: 099 0.34- 1.90 . (9) oat-q: " 3- 5"" 55 - k9 Y'3.“9 lei-.571 7.4.9 (to) lyrack Neurotic: 12 3"}. 500 nd- 5511 (1‘)Xn0rnxtm- .11 '(56) hItnin' SIz: I act Int-vary 0 6 5'“ Y-ZJOO ud-.500 m. 357 ld-JSD (13) mm "‘-‘ '2 “'1 coda: 1 7-3.207 uni-.330 (5) (121-0: 5.61 (H) Eylenck Nauwnc: 1 - ~18 (15) 113-tructor $3.041 u-.703 043.21. ”:2“ amt-v.2: $2.991 04-.689 ”11.51 14.713 .3- 452 1.3m (26) Anluyslz l 1 $3.195 .3-.6u 1,-5.31 (25) lyul‘uk 1.1.1 “2. 937 “I 039 I”'15.:4 19) HIjor- code: 2,333,157, ' Y-ZJOS ldI.666 27) Mann-2: 3 I S . 31.-2.01.4 sa-.595 (25) Smack 1.10. 98.51 Y-ZJZS ld-JM n-lJI-B FIGURE D5 '(“l A-B Tut Mainly ’17-s.ooo 04-.000 I'D-.‘l-Z A-B To“ “Macy: - M r3324 m. 501 ’(51) All: god" 1,3 Y~3.600 “I 590 n-l3.7'D-ZS . (M) Eysenck :xzuvarz. $3.121 uh. 391 n-3.61-19 (61) brack Extnwn 2 2.33] Mb. 553 13:2. 31.12 (3 7}! yunck Naurouc: Y-Z. 55B ‘11-. 537 III-31w *(36)Eynnck Mutant: 1 1131.000 w-.0oo a-JZ-l 329 FIGURE D6 AID III: Explanation of Final Exam Score-l by Final Course Variables «an AnxieKJ-l: 1 - 1 Y- 32 22:. 54-1 995 n_-5 a7.- (M) Inuxuztot Hulouewcrm: 330 0 - 143.000 Id-2.560 I143) many-3: “114.11 1 2 . 5 (5) \'-32.273 Id-2.606 mdtorl: n-BJ‘L-U- 35- 39 32. 'd'z-M‘ '(35) lnltructo: 16. me). '(48) Maw-non 15M. .(39 COT-T: ) 37 7 so {:32 117 .a-z.57-. n—IaJ‘L-ll (31) man-mag SIC: go - 3 (3) :3??? 'd-z‘m Diuuuian nut-n: ' m1 udn: 29 - 39 _ - 23 3:33.23: (38) (m r y-so.157 Id-ZJJII ' . - ' .5 . 1 ”71.01 . 35 “'9 7‘ ’ 330.653 “4.215 ”7.11!” (66) 1h jot: goth: 1,2,6 (6) 1-30.133 n.2,”; manna. n-2.81-15 g9 - 3h 2749.565 .0-7.161 1.1-3.51.1 (30) mining Set: a $6.91 — ' $20,400 .a-3.6aA n-5.61 (6) cm- I(20) xyuun 7.1.: '(28) 021.17: 1 z - 9 .. —' ' 1-32 000 ua-z 320 2-31 3000 Id-Z. 369 Y'29.022 16.3.395 ' ' ' _ (25) Euzmcmr ”.91 {19) Enfoit: 1.1-1.11-7 (22) #:rntng Se:.n-1.91-1c‘u£“m_h » 71-10309' 9.3-3.1“ v- 23 036 “1.2 996 5- ‘ (17) Imnuctor «23.57. E. 21. 5 222.532 Id-Z.853 _ autumn—r7 . 745,303 u-3,263 (21) :yuuk 1m (17) GTE-1': I03) Major: -23.91. 1 - 7 g - 342.1,; 53 (I) 1-2s.7zo ud-J.1oa 1-28.6M lei-2.9102 5 29 human , y— 23 494' .d-z. 6A1 A11 :ubizc3l (a) “.5"; *(13) pun-x; «22.27. 9(23) Lgunxng 5m n-19.9'L ( ) California-F: n. .11. 57. 77 II II _. _ y 3:2; 500 «4.425 1922.500 -d-1.500 y-zo.ooo ud-.ooo £26.53. “M." 322.33; 5.91 PM" °‘ 1‘" rum: 1 '1 (16) Inltxuctaz C-m'uf nLl-h 2-25.ooo sen-3.30 ”2.17.41 '(15) 002—7: 740.653 nd-2.959 ”.11 was) lijun soda: 1,7,0 71-27 059 541-2 920 13) 1110252.: rain-17 . ’(37) OQT-Q: 1“ ' Y-Jz, 250 .a-z. 272 .131-6 «00) thjar: gods: 1,2 (10) “28'5“ “-2'8” (50) Dilcunxon 50.52;“ (2h) Midterm ”3.0146 Antwan: - _ n - . Y-16.736 y-z7.o19 90-3372 _. .a-s.a9a (u) “21...“ 11.10.22 (36) 09M: 3:22:11: run-2,424 n-17. - . - 23 ' (1) r-zs.333 .0-3.926 “IS-5°" "ha-3°“ maun: $13.01 ' M 61 (‘1) H050: 15 - za coda: 3,1. ,,5 6 0 III}. 676 d ':2 23 33110 *(25) mm“. .Ia ' - 9 "5" “mm“ 17'“ 20 - . Altituda: . - 1a - 1-23 867 um Mo — . . Y-Zb.375 ad-z.7za (12) mal.“ *(53) OQT-3; ””7“” max-2:. 15 - _ - 7(11) 7.13.330 “1.3.392 v-zl..a95 nd-2.770 50-50-11 nfi.7z I 61-19 1-10. 000 “'- -°°° *(sz) cor-Q: nI. 217-1 _ ' 140.057 lei-2.74! ' 301.11-6 331 FIGURE D7 AID III: Explanation of Final Exam Score-2 by Final Course Variables 6:60) Age: ZES code: - 149 8.3 (19)Zysenck Neurotic' "7.97.42 23 7mm 1‘48. 920 sd-Z, 601 (21) Midterm: 644.11 1320.635 0.1-2.007 ' 07. \ '(28)Eysenck Neurottcx fi(l,[\ 4283’ l - 5 1 MI . Y-ZS,869 nd-2,923 747. 755“ M. 2 629 will) Inutr1‘clorzn-9.47.n.6 2, -3_ Calif!) rnil- F: l 146.113 «14.957 1141 61 *(59) OQT-Q: - 50 mm“ . (9) 140.750 «14.112 . T-Tr (20) Midterm: / n- 131-1. *(68) Major“ 20 - 50 y - 31 ’ Y 27-55 146.62) 6114.917 E3: 020 211-1.000 664.172 -5 57. 55 c 1 ~28.6' < ) 90 42:5 Y-25,960 6d-1.Z71 n-h,71 *(hs) 5040-21 (12') [nutructor- 1:30i500 I'd-2'1” Culifornh- P- .0. 50.50... 145,311 «44,156 ”56),“ m g - 1 “'7-01 gode' 1,3,5 ,6 Wat-m... Y-26.81 PM 361. Zed-2:206 0d~3.h71 (2,7,; 5040.2, 111-17 (66‘, Au mum-1n 49.51 6 - 10 145.212 ad-2,796 / v40 16.7 1d- -1 7 2 n-6.z7. 5(57) mama Inn-mm WLm-m, 5M.- m‘a‘m—‘mx' f7- 43. 000 «14.1.66 n-2.17.-150)Nsruns{on Atrirudu; . 0 147.080 Bd-2,S90 (3) . "4.77. GPA: . 5— ’ 6 - 6 0 . . $326.71; (31) :re:f.nltudm (21) “may.“ ‘\67) Am-m:y~.: :33'3? . $25,760 Lad-3.050 146.035 Id-Z,726 1746.105 ed-..EM ' - (27) 50—50-2; n-9.47. n-3,B7. l-).61-I9 5 - 11 . _. -‘i...__.. (9) 1745.005 0d-3.316 (QT-T; 11.17. .(53) Anxiety-11 (smirmuniou Attllud .1.- , _ 3 1- 44. 9933 ‘ 7;”.500 “.3 500 1621.95 .64.»: .11— '34“ ~(3o) Exe-Actltudez emu-2 “"-”~'“ n-ll.3'l. . 1,) 141.165 [Cl-1.3.56 50-50-1~ *(26) 50—50-2: -1.31-7 Y-6.000 146.000 Id-,000 000 n-.21-1 n-.27.-1 "38”,.“an murotkcx l - 15 145.651 «14.611 (33) 1111166261: -13 - 34 _ 37 11 .01-69 (23) 143.713 604.755 ”Mm“! ”13.21 1139) Eysenck Neurotic: 1) Set: 22 .10 - 35 140,000 ed-.OOO All lubjccu 37.25747 ”'27.“ In Itudy 164.911 _. n-1A.31 145.615 . _ 44.960 (32) "$553- n-1001 (5) 142.000 «64.517 ‘ (121-T: “.137-5 (A9) Learning Sun 11(63) Pretest: g6 - 50 2.0 145.096 Y-ZS. 0002 011-2 ,251 y-zs_ 750 ”4,395 Id-3.262 n-lo.11-2n..3.'1 645.61 ( ) a 22 (19) lnltructor . .. . turning run Land: Set! 7 - 0 1 9 — - 146.367 ld-1.677 . Y-10.833 , ':“ _ ’ A Learnt San * 62 ' Id-fzJu " 9-21 .. ,( 8) 2 . ”“8 ( ) Pang. ”Lu-6 Y-23,A07 Id-2.628 Y-ZZ. 8163ldu1 zoo n-5,11 “.1. 31- (51.) 1153:: god 143 0975: .64. .22 127) CM: 2.0 - 2.5 (11) . . 143.650 644.613 (2) Midterm (30) Eygericsk Extravert. "_7‘57‘ “1’“ - Y-23.326 «14.359 1.2 - 2.5 143.357 n-8.u 143.207 06-} 061 «14.656 4111 0-60. ' *(ks) cm, 7 l- a . (18) lnatructor Y-.19 mogul-1.4601 Caurn: n-,62-J *(65) Instructor Hutu (h) ' mT-T. (35) figscngl; Extravert: 1 - 2 _ _" , 142.520 \1 141.000 .111-2.950 d-3.263 n-3.87. 3.67. *(65) Instructor . 1161.: "' I, 16 7A 11 +1601 54.7611 11.: - so 17 4 7' ~ , *16 3611.111 3 ~ g ‘- ( ) oy—eir I: 71:.07'ub 1-2 '31,. “1.243).! 143.300 «14.651 1.17.41 (15) Cur:ent Load: n-b.57.=24 A 17:22. 063 60:3. 157 I (2") :3:- 2 ‘(SD Evaenck Lie: 1 <19 1425 925 d- 2 951 ’ ' 7 “(:3 . Midterm: (17) Eysenck 1.1:: a 149,057 “"1687 fl ‘ Erie“;- ls - 29 _ .. “""7’6 1:43 05» m4 711' “21335 141.659 211-3.168 r.-3 13 - Id'3.19h 646.01 ’ 22.07. (36) Eysenck Elli-V221: 141.490 44.612 n-9,21 I'(lb) Current Lad; _oda 1 148.750 mil-1.785 11-1.51-8 Preten: 13 Y-ZO. 581‘ Id-I. 0134 n- -S.8 (37) Eysentk Extravert- y; - 2 FIGURE D7 333 FIGURE D8 AID III: Explanation of Final Exam Total Score by Final Course Variables 1) All Iubjecu 1n ltudy (3) GPA: 1.5 ., b 0 Y-58.9§9 Id-6,281 n-27.BZ *(11) 50- 504: Y-A6,000 Id'.000 n .Z'L-l (5) Midterm: 39 ~ 7 Y-53.569 5d-5.589 mail». (2) GPA: 1.2 - .7 Y-51.B39 d-6.J73 n-7l.21 (1‘) Hidnm: 15 - Y-GG. sd-6.518 25.87 L‘QT-T - Q 3.3),“: sd-J n77 n-Z 6’- (19) Hidtarw: ~93 {-60 316 Id-A 650 n- 9 21 334 '(41) so-so—z: g . 12 JA .d-LW» .al ‘(AO) 50-50—1: ._ - B Y-So,625 :d-3.160 ”LEI-8 Midterm: ‘2-5 7. 563 ld-S 105 n-6.07 , 6.) Midterm: 29 - Y-52,9$2 sci-5.1403 n-38 91 18) Midtevm: A - 3A Y-s7 .301 Id-AJ“ 15.61. '(Zsflnsuuctor Risk: 6 71-61375 ud-a 01a n-l sz-s (27) Instructor Nah: 21 - 5] 156.600 nd-AJJJ ”14,17 ‘ (1.6) Inauuuor link: 15 - 3D Y-eom: “14.262 n-2. 31.-12 (5‘) Major: soda: 8 Y-60,800 ld-4.167 '.97.-S (47)Instruc:or Rink: Y:55. 600 Id-S. 00!- "(59 Major cade:1,35 Y- 3 867 ud-J 92.7 n-2 at- 15 (11) "81°? code:l,2,3,b, 5,6 Y:5S. 099 ld-IA 6477 *(23) Mum-:8 ceod : 3:51; 278 sd-S 665 (3) GPA: 5.1 - 2.3 Y'51.6ZD ld-5.362 n-ZO.3Z (l 7) Midterm: ‘ 2 721) OQT- 26 - Y-SJO 875 .d-a 986 (20) OQT-'1': .1. . 25 Y-50.671 .d-s,230 n-IAJZ 7*(38) Ey:enck Lie: Y:5621767 Id-J. 568 (13) OQT-2: é - Y-50.252 .d-5_937 “15.57. OQT-T: 2 - 6 YMSJBA :d-6.787 n-7.lZ 6) Kid? Y:43 3M ld-6,108 .61 (39) Ey::nck Lie: Y-5072901Id-AJM n-5. 874-3 (18) Eysenck neurotic: Y-za: 800 ld-i. 099 “.9 *(29) Eyunck Neurotic: V-zmyoo .a-_ooo «.21-1 Ins :ru ylenck omurouc i ll- 7-146,69O ud-5,1o9 n-6.87. *(Mflnatruccor Cgluc tnh-F: $49.05!) «4.330 n-3 . az-zo Instructor Bylanc k Neurotic: 15 Y-Z7.500 tel-7,500 n-.kZ-Z *(65)Inlttuclor Cnllfornu-x: 7-a3.7so sd-um n-J.°7.'16 FIGURE I)8 (3Z)Ins [run '01 Risk: A6 V-SS 520 Id-A. 1055 n- 16 75 *(numtmcmr Risk: 149,331 sd-3.670 n-lJZ-b (“3) H15" ”(49) Hum-,2 Wdc: 5,8 Y-SO. 7506:1791 L09 171-6.072 (25) I\\ltn:c:or COI eLo-d: cod 2 H: Y-52. 361 sd-S. 138 ‘26) Instructor C : - (37) Frau-L: 30 Y:50. 9S73 ld-Q. 639 .'.\‘/.-13 ‘ (36) Pratt“: Y'Ag. 963 Ill-6 761 n-5.17.-Z7 ' (t3) lyscnck Extravert la - z “55'“? ““1"“ ‘ (51)Inltru:tot “'91- Ha: 1a hue-Crowns: 1 7- 6.33 d-l. .5 3 I 886 *(53) Pant-Autumn: n LII-6 go - 32 Y-53.276 uni-3350' L'yunck Extravert: . n-2.lZ-ll z - 17 Y:;lék52 “'43:” 50)Inltructor n ‘ 1 {law -Cto\.1ua- 7-so.2so sd-b.720 “14.71 Poll-Attitude: ?-L7.926 Id-AJSD n-Z,6Z-lla 335 FIGURE D9 AID III: Explanation of Final Exam Grade by Final Course Variables (25) (117-1': A.‘ . CDT-Q: ' (29) lnnruuor 35 - so ‘ ‘ Rink- Y-AJS) 1 - 1; lei-.581 14.31.: -¢-,a39 16.0' n-lO.J7.-55 «mm (2!) Xmuucmt llllx g - z y-J.ou Mum. n-3.01-16 (12) CIT-Q: _1_ . 2A 16.696 (31) [ya-nal: lunatic: .009 _ n-u.a1. *(20) Park-n: r-JJM u-.sss 12.933 «#531 (19) “I“?! “'5'” .9 ID ' 2 ' “ “L 7" 10) n- v-JJno “you yet: n-u,51. our : gothuJJ 5 :30 r-J.167 04-.7“ 7”.“ “_J“ nun.“ '(JJ) lyunclz Nuuzozlcz n-J_kz-la 13,176 «5567 (9) mum; Sat: «1.) Q14: .431.” n - 36 l _ 5 $2.51.: «man n-2.61-lb (6) Run“: (1) ' An I b t 1: 33¢; “'39.“ (8) §Alrn1q s“: Y-J.327 'f-ZJJ: .a-.9s7 ul-Lou “Lab” '01) gnu: -loo _ . 14.375 u-.599 (s) Lox-16 '(36) Major: Hldtezr soda: 3,8.S,5 26 - 37 v4.33: u-.ose $3.132 z-s nth-.866 4.0.07. (30) Put-It; Q 15 . u 753.560 .4559: “’7’ ”‘1"! nut, 1. ‘ ”LOX-16 Hunt-t 32 - Yin/.58 ud-Jzo . ”13.11 (26) com; 1 - s (16) (Inf-:3 13) (133!- 2 ’ “‘33.”, M-.668 $3.193 Id-.701 $3.351 Id-.660 "1.11‘“ 3.15.61 n-le! .7 . (17) cut-v: $1.255 u-.su (39) flat-II: 11.12.62 51 - 15 r-z.as7 r3314]. lyu'uk .In-Vt!!! LL - u r-z.ee7 u-Jzo “5.11 (33) Kidnrnz ls . 20 yuz.ooo ld-.816 “Lu-6 (0 urn-m y - 25 r-2.2u Id-913 "10.71 . ”Lab-u 16) lynnck Int-van: A 13 9-1. 925 “new n-S. 11 $1,250 «an: “2.1m: FIGURE E )9 335 FIGURE D9 AID III: Explanation of Final Exam Grade by Final Course Variables (25) (117-1': 2.7 . CDT-Q: ' (29) Innrurtor 35 - so ‘ ‘ Rink- Y-AJS) 1 - 1; lei-.381 74.31.: -¢-,a39 16.0' n-lO.J7.-55 «mm (2!) Xmuucmr llllx g - 2 7-3.ou ud-.63I. n-3.01-16 (12) CIT-Q: _1_ . 22 16.696 (31) [ya-nal: lunatic: .009 _ n-u.a1. *(20) Park-n: r-JJM u-.sss 12.933 «#531 (19) “I“?! “'5'” .9 ID ' 2 ' “ “L 7" 10) n- v-JJno “you I": n-u,57. our : gothuJJ 5 :30 r-J.167 u-.75¢ 7”.“ “_J“ nun.“ '(JJ) lyunclz Neurotic: n-J_kz-la 13.176 1.1-.567 (9) mum; Sat: «1.) Q14: .431.” n - 36 l _ 5 $2.51.: u-.au n-2.61-lb (6) Run“: (1) ' An I b r 1: 33¢; “'39.“ (8) §Alru1q s“: Y-J.327 'f-ZJJ: .a-.9s7 ul-Lou “Lab” '01) gnu: -loo _ . 74.375 u-.599 (s) Lox-16 '(36) Major: Hldtezr soda: 3,8.S,5 26 - 37 74.333 u-.ose $3.132 z-s nth-.866 4.0.07. (30) Put-It; Q 15 . u 753.560 .4559: “’7’ ”‘1"! nut, 1. ‘ ”LOX-16 Hunt-t 32 - Yin/.58 ud-Jzo . ”13.11 (26) com; 1 - s (16) (Inf-:3 13) (133!- 2 ’ “‘33.”, M-.668 $3.293 Id-.701 $3.351 Id-.660 "1.11‘“ 3.15.61 n-le! .7 . (17) cur-v: I-Lzsa u-.su (39) mam-2 11.12.62 51 - 15 r-z.as7 FIJI-11 lyu'uk [mavens LL - u r-z.ee7 u-Jzo “5.11 (33) Kidnrnz ls . 20 yuz.ooo ld-.816 “Lu-6 (0 Mann. y - 25 r-2.2u Id-913 "10.71 . ”Lab-15 16) lynnck Int-van: A 13 In. 925 “new n-S. 11 $1,250 «an: “2.3m: FIGURE E )9 337 FIGURE D10 AID III: Explanation of Instructor Grade by Final Course Variables "(68) Eyocnck Lie: - 6 $4.593 ud-.I.91 39)!)hcuuton Attitude: n-UJ'L-SQ .11 - 13 (13) 174.567 .d-.529 lnltructor “'11.“- Con ’(69) Eernck L121 7 1 y-3.ooo «15009 1 (38)”1ncusaion Attitude n-.Z'/.-1 *(37) Pro-Attitude: I 7 _ 1 (72m: _ 7.1. 197 313-3.521 1 L6 _ “.0 n-13. J'L- 71 . Ydo.269 . Id-,631 . 13.07. (33) gyzelluék Neurotic! Y-anék 513-.559 1'1) u-13.71 Instructor Course . Id-.579 %°";,5;§" *(36) rye-Attitude. Id-.603 l3 - "' Y-3,ooo -d-.ooo ”16.22 0) ~(32) Eylenck Neurotirx n-Uo'L-l Hutu-1 _ _ 4 £1 - 39 1-3.571 Iii-.1995 Y4.070 «(25) 0114;, “.1 31,7 Id-. 6” _ u-u. 77. 14.250 1.1-.433 *(18) n-.8'I.-A Eylenck Neurotic: %.2.;31 (“wig-Mg Neurons; ad-.799 , _2: __ - . .- a 5 ”2.67313 (28) Succzu (mmqu‘ :_i.;1:jnsd ..7 ( LEA: 17.-7. ooo 6.35612 9 Sunford-cough: ‘ ‘ n-J, 0"- .. . .f 1'7 ' 1'5 H”) :0 ”I: v-3.7a3 tad-.550 1 3.756 _1 - "_S 52 Id .672 Y-b,250 311-.933 - 4.6.91. ( n-.87.-A 04-.565 157-ant" (bhfllyaenrk Newark: Neuron ~ 1 - Q 6 - 20 (29) Succesn-Z: (“7) 50'50'25 . 1-3.zae 1.3-.sz 1-3.e75 ; . 1o (Shun-mum: ”Lu“ Id-.6\3 y. 3. 575 "1.,535 {'3.527 Id-.567 s. nford~ ’:COufih 13.57. ".11 1 n-10 31 (176) (m4): - -3 :33 sd- 13‘ 1-3,o911-d-.2aan-,32-15 n-21. 7.- Attitude (21}Zl1gt1uctor - 22 _ 23 Snford-cgugh- (24) A-H Ten Anxiety yd. 1929 “”723 (R6) Madam-37 _ - 13 - ' n-2,s~L-1A —_ . 1-3.97o Id-.701 Y"“°°° “"651 Y 52.002 Id .000 N52) E - ntk Ur ”12.27. n-12.0‘l n-. In Oy-J c . 1-3 927 ud-,SIJ (9) 30)D£:cuu£ou Attitude: n-7 7-1 Hidlgrm 3-1 1-3. 890 ud-.5na 7-3. 903 9(25) A-fi Tut Anxiety! n-9. A: (A7) Amman-3: 1(3)) Eylcntk He: 1.1-13.21 77-33933 ud-.3s3 1-3 m m-mz n-?.D7. n-l J7. *(20)Inl:xuctor er-d -Gough1 (1) 13.000 Id-.633 All wbjccu n-.9'L-5 in Indy ‘ 1'(Sl)!nltruttor "3'7” cnuo rnu- r- od-.713 7 _ 1 "m V4129 d-.613 “(70)!nnruclor (23) A-H tent Anxiety: / "1‘61“: Harlow-Cram: 5) 39 - 1 _ 13 W” ($0)Inlnuclor v-moo .a-Joo 1" ‘ "7 c.1110.» «.1. r- n-l.97.-l0 Y-J.668 “' 7“ (16) Instructor 1-3. 739 lei-.529 *(7l)1nln:ructor '36- 6‘ nck Neurotic r) (ink-Crowne: _ 17 Y-3.803 3.. 1-3 6621;d- A99 “.12'51 ' 701 (22) A-fl-Tut Amuyx *(56) ASH T935! Anxiaty: "_2' 714.552 ud-.67h $3.929 -d-. 707. (H) Cradle-x n-5.5'L n-2.67.- k gods: 3.5 ,5 1-3. 659 Id-.699 .11 *(57) Ad! Tun Amitty: -37 17 1 t c 1-3. 200 «15100 ( ) n- ruc Mzruflmm «53) Credits: ”1 37,-13 (5) ZE ' y15 $3.571 ud-,499 thterlt Y-3.260 Id-.SJI ”Lu“, 19 - 29 n-b.77. 1-3.537 uh. 713 “.23 u “69) grunt-2 ‘-2 873 Mb 331 *(59)Inuructor . Y ' ' Hill-rug gun}. n-IJW is - 31 ”3:. _ 1-3.931 17.3.500‘ “550° Id-JW Credits ”3'01-“ “.53“! $032119}?! 4 s34 ' - ' '1 *(i8)1nltru:lot “'6-01 Con 21,0111: 5 : 1,2,3 1-1 5751-4- 600 n- 30 *(50) Acculz'lcy-Zx (15) 1r-I..3339 uni-.667 Inatructar “.1 *(63) ha Tut Anxiety: ‘yudtk .5} - 5° Y-5.000 Id-,000 n-J'L-Z (“Mummy-2, '(62) A-II Ten Anxiety: 3 - y. - 19 y-3.333 a3-.712 Y-3.158 ld-JEB n-J.97. n-3,6I-19 ' “(61) Le-rning Sat: 30 15.000 '11-.000 73‘ n-17.|7. 33$):ynenck Bxu'avert: null-1 ' l - . 6) 3 14.379 Id-.6ll ‘ Ila-tructor 11.5.51. , M60) Leninlng Set: gay-I’m . _ . Lac; Y-JJZI Id-.§JB Z n-5.3'L-28 Y-JJJI 2116;; fi(67)lnntructor H" lam: -Ctom1 $3.043 .d-Jns (3k)Ey:e11|:k xxtnvertl “4.11.22 Id-. 567 ‘11.:37 i(66)1nluuctor n . Harlow-0mm: 5 Y-1.269 ld-J95 "Lil-7 FIGURE D10 888 339 FIGURE D11 AID III: Explanation of Final Course Grade by Final Course Variables I!) Hutu-m Eu!) A311; 33 - 2 . ,3 $9,010 u-.313 wuss mum a1. 1|-2.11-11 11'11. (26) 505394! 2 - YELZBG “In“? .01 _J ”('38) A531 ‘1 soda: 2 Y-3.900 ld-JJG 134.9%“, (7) _2_.6 - 4.0 “.389 131-.551 ~00) Pro-Attitude: ”25.11 33 - 27 H.611 Ida“! 13- .614-19 10) Midterm '01) Pro-Attitude: 31 - 39 £5 - 31 . NJ!” £45457 (3) pawn-21 niacin: 32 - 39 1d,!“ tug) GPA: u-.us 1. - ’.9 INSJ‘L Y-JJN “'.362 . 19:-1.11838 my) Midterm gs - 37 (5) 14.300 ld-.‘$8 GPAI ”3319-10 L9 — 1.5 ‘[-3.753 “5556 rlex ‘0') Midterm .0!) lyunck lhmotin 32 - 35 y - 9 1-3.599 u-.3zo 1-3329 u- .Aos FILM! n-IJPIO (11) cm 1- - ~ . FJJZO Ida-317 134.”. NM) Iyunck hmotiu . - 13 ?-3.972 04-.)71 16) Major! n—6.Bi-Jb god-11 ”5,6,7 1-3.e7i «mm 13.15,”. (I) All lubjccn (”Mn-tructor in Rudy .63 ld-JZO (20) (1391 2 2 mm $3.3M .4- .670 n-S.5'I. Hutu-1 3.9 ' 31 114,373-12 Y-JJ‘S a .621 2,61 ”(17) Credit" a 1 . .3 Y-3.291 ‘d-JIS r. " 11 'l-JJE? tel-.55! 13.01 9(37) Ad! Tut Anxiety: 63 \1 3 74.000 -4-.om (16) 171-nun: ’ .3...th codex 1 1 7.2.9“. iii-.5253 / l “w.“ L 125-6.31 J (2) a 0—1 - _ y. Midterm 0” ”'5 I am) 90113;: Ania!“ £333; $1.000 314-.000 3.1,“; “.An . .3 . «c.1133 “ 5‘“ up“: 34 In”. 71 (inky-luck lxtrlvu’t: 33 . u 1.3.000 M-.311 “Lug“ fl) 4 u: is - 25 1.6“ “"632 mu.“ “Human Extravert: 33.193 ld-,350 n-l.3'L-7 (31) inane: have": - l ‘51.)“ $.60!) III-6.21. 0(3))lyunck nun-rt: 1'2 . In “'.W “Mme FIGURE D11 OWE APPENDIX E TABLES El-Ell: PROPORTION OF VARIATION IN EACH OUTCOME BY SUBGROUP WITHIN BRANCH EXPLAINABLE FOR EACH VARIABLE BY PREDICTOR SET 341 .vmuaamuuw uo: uHHmmu «z .mvwa uo: “on vmumamuum ufiaama .9 .msouw awcwmu .mmH\mmm ammo uxozu Amunv .wanmwuw> mafia co mama uaaamu ir\\ HAmmmkmmmv nauowvmnn 3000 you manmcamfiaxw macaw umsu cw coflumauw> mo cofiuuomoum mom.¢~ ooo.~m oo~.m~ oom.- oom.- ooo.o~ ooo.m~ mmo.m~ “mo.m~ mofi.nm ~n~.o~ mo~.- zauuxm aboo. moo. ooo. Hmo. ooo. ooo. Ammmv H-Hoo So. .0 sh ooN ., 3o. 2o. So. So. So. oéoo So. I 1‘ ooo. ooo. So. So. 8o... o8. >48 @ 3o. oNo. ooo. @ So. So. «mo ooo. moo. omo. muo. quw n HNH. muo. Homo: omo. ooo. ooo. Hoo. ooo ooo. moo. mm< moo. n .omomo ooo. ooo. ooo. ooo. ooo. xmm Amo *oo *Nm «on ¢om on no om mm ANN m H Hwnasz adouw Heuuwcmum manmfium> mmusoolwcwumuam nomm pom anNGHMHme suamum GHSuHB asouu kn muoom ummumum ca coaumwum> mo coauuoaoum "H QH< Hm mAm mwzu so opus uHHam n filll\ macaw umso ca coaumfium> mo coaouomonm omo.mN moo.NN NNo.oN oom.oN mom.qN ooo.HN mmo.oH oHN.mN ooHumN mmN.HN Noo.HN mNN.HN moN.NN z..H.oo Noo. Noo. moo. on. «No. oNo. ooo. omo o o. moo. N o. N H. oo HHo. oHo. oNo. “ohm: ooo. oNo. H ulumdanluooo. moo. HHo. Hoo. moo. moo Hoo. NHo. oHo. ooo. Hoo. ooo. ooo. ooo. xom *mo *No om Now no «oq *mm Nm o o m N H monasz,moouw Hauoapmum A.o.ucooo Hm on mafia co owns uaamm n.4ll|\ HAmmH\mmmv nauoavoum Loam now ofinmafioamxo asouw uosu :N coauoauo> mo aowuuomoum ooH.oN ooo.oN ooo.NN HNH.oN ooo.HN nmN.HN Noo.HN NNN.HN moN.NN zoo: moo. NNo. omo. oNN. ooo. mHo. moN. oNN. ooo.H HmmN\Hmme oH mH mN no ooN Hmm oHo ooo Nmm 2 fi ooo. Bo. ooo. Ho. Ho. oNo. Adahons Tomom - . NHo. NHo. NHo. ooo. Noo. ooo. oosHHooo-oHH oHo. Hoo. it. No No. NHo. NNo. ooHHoEm sonoom oNo omo HHo. HHo. NNo. Nmo. oom wsHsnsoH .onsoo Hoo. & o . NHo. HHo. ooo. H-335“... some NHo. omo. NHo. & Ho. Ho. ooo. 3602;. “moo H74 ooo. o .. mHo. NHo. oHo. oNo. NHo. oHoom oHH Nosooem mmH. oHo. ooo. ooo. Noo. oHo. soHoHoosooz Nosooem oz oz HHH. o.o. NHo. mHo. N . o .. m . ooHouo>voxm NosonNN HHo. ooo. ooo. Hoo. fimoou @ av H48 4. m ooo. oNo. moo. ooo. oHo. oHo. oNo. 99% NH. oN . N o. Noo. NHR oer. no. >4 mHN.on1I.ooo. on. No. No ooo. ado oo . oH. NHo. mo HHo. oHo. mNo. ooflsz oo oHo. moo. moo. HHo. Noo. moo. ow< oNo. ooo. Hoo. Hoo. ooo. ooo. ooo. . xom «mm 83m 0H 0H m o m N a ... Nonesz awouw youofipoum A.o.usooo Hm mHm wasp so opus oHHmmunl\ HAmmH\mmmv ocuowpoua zoom now manmcaosaxo macaw ooze a“ cofiuaHuo> mo coauuomoum Hoo.oN ooo.om omm.mm oHN.om mHm.om HmN.Nm oNo.om ooo.Nm ommqom oHN.om zsuuxm Noeonem ooo. @wu oNo. ooo. 8o. EH. ooo. H348 oNNr ooH» . .1nHN. ooo. o-Hoo HoH. Noo. ooo. oNo. >-Hoo ooN. ooo. NNo. Noo.n “NH . «so ooo. Noo. oNo. oHo. oooH ooouuoo ooN. ooH. om . HNo. mHo. ooHovo . m giLoN. @ ooo. NHo. No.52 NNH. ooo. ooo. Noo. Noo. 6N4 Noo. Nmo. oHo. ooon. HHo. 86m tum « . mm m H Monasz macaw Hauoapoum oHanHm> omuooolwfiz zoom Now manmafimamxm gunman afinufia anouu hp ououm mem Enouwfiz GH cowumfium> mo coauuomoum NHH QH< mm m4m mo comuuomoum mom.e omm.m ooo.o omm.m sea.s omm.m ooo.o soo.e soo.m oem.s mom.s oeoqs msm.s meo.m zrmz moo. Hoo. moo. moo. omo. «mo. ago. ooo. omo. ooo. moo. meo. ems. ooo.H Homo\emmo on o m oo mm mm o me so om so me ooH mmm 2 mos. moo. moo. Nmo. moo. omo. oHo. eoo. roam ”nouooeuoem ems. Hoo. moo. omo. Hoo. emo. moo. moo. o-uassoooaoo “souoosuocm Ame. NHH. moo. ooo. omo. ooo. moo. moo. .two-ouooaum "souoouuocH moo. oHo. ooo. omo. omo. smo. oHo. moo. .ueo-osoauuz "nouoouuucm moo. one. oHo. mmo. moo. moo. «so. ooo. .ooz xoconem "nouoouuocm oeo. omo. ooo. mmo. soo. smo. ooo. ooo. .oxm xosooem "souooeuosm f M om. omo. omo. ooo. So. omo.. m2. 4% oooooeo 111mmo. omo. . omo. omo. ooo. moo. A.oxm-.euoeoxo xosooam {£162 2o. omo. mmo. ooo. ooo. mmo. onooo mo omo. omo. moo. ooo. ooo. moo. emo. o-ooo ooo. e . omo. meo. omo. omo. moo. omo. >-eoo mmo. @ omo. omo. :o. omo. HmoCE. «so ooo. omo. ooo. omo. moo. moo. moo. Boo. oooa ososuoo one see. omo. moo. emo. sod. omo. moo. oueoouo so 3:. ooo. ooH... u omo. oow. @ 2o. Sam: mmo. Hoo. . meow Hoo. ooH omo. ooo. o s ooo. moo eoo moo oHo. omo. moo. ooo. xom rmo rmo som 4mm so me «me roe om som mo mo m H Honesz macaw Hauofiwmum oanmmum> monsoolcmz comm now oanmdmoaaxm cocoon amsufis macaw kn oomuo Eudora: am .um> mo coauuomoum "HH QH< «m mqm¢H 362 TABLE E4 (cont'd.) Predictor Group Number 1 2 5 7 20* 21 26* 27* Sex .004 .008 .008 .004 .014 .003 .010 Age .006 .006 .015 .030 .037 .030 .076 Major .012 .019 .018 .054 .120 .100 Credits .015 .002 .004 .002 001 .009 .010 Current Load .007 004 .004 .012 017 .051 .020 CPA .195 w .030 .031 026 .0 1 .081 CQT-V .058 . 9 .008 .031 063 .037 CQT-Q .037 .019 .018 .042 062 .135 .101 CQT-T .053 .007 .012 .023 .004 .109 .015 Eysenck Extraversion .011 .007 .010 .012 017 .164 Eysenck Neuroticism .007 .015 .056"?031 024 .023 .o55 Eysenck Lie .014 .013 .004 .031 .026 .068 .020 A-H Test Anxiety .047 .030 .042 NA .082 .067 Test Anxiety-1 .003 .001 . 02 .012 ..17 .000 .020 Test Anxiety-2 .010 .002 .003 .025 022 .009 .054 Learning Set .037 .023 .017 .031 .032 .059 .11Nl Reason Enrolled .007 .006 .009 .040 .026 .064 .049 Pre-attitude .006 .009 .011 033 .034 .098 .041 50-50-1 6Asp.-Exp.) .065 .016 .030 .118 ' '?022 .030 .038 ' Pretest we .056“?003 .010 020 .069 .063 Instructor: Eys. Ext. .004 .010 .003 .051 050 .084 .059 Instructor: Eys. Neu. .008 .012 .010 .041 055 .084 .059 Instructor: Marlowe-Cne. .003 .008 .006 .034 .051 .084 .039 Instructor: Sanford-Ggh. .005 .004 .008 015 .024 .084 .059 Instructor: California-F .013 .014 .011 027 .047 059 .024 Instructor: Risk .007 .004 .006 .039 .085"1059”‘9090 N 532 384 238 80 2 77 28 49 TSSi/TSST 1.000 .649 .382 .116 .004 .097 .036 .054 MEAN 3.662 3.385 3.550 3.875 2.000 3.935 4.250 3.755 Preportion of variation in each group explainable for each predictor I"3I=Split made on this variable. =Next best BSS/T88. * IuFinal group. =Split attempted but not made. NA =Split not attempted. (BSS/T88)i 363 TABLE E4 (cont'd.) Predictor Grouprumber 1 2 5 6 8 14* 15 18* Sex .004 .004 .008 .004 .002 .003 Age .006 .006 .015 .017 .016 .014 Major .012 .019 .018 .055"".‘024 .009 Credits .015 .002 .004 .005 .022 .028 Current Load .007 .004 .004 .002 .012 .010 GPA .195 .030 .031 .024 .019 CQT-V .058 . 09 .008 .012 .055 .064 CQT-Q .037 .019 .018 .028 .066 .038 CQT-T .053 .007 .012 .009 .019 .021 Eysenck Extraversion .011 .007 .010 .024 .043 .041 Eysenck Neuroticism .007 .015 .059“?020 .058 Eysenck Lie .014 .013 .004 .019 . 61 .056 A-H Test Anxiety .047 .030 1631' .029 .027 NA .014 NA Test Anxiety-1 .003 .001 .002 .003 .002 .003 Test Anxiety-2 .010 .002 .003 .003 .013 .020 Learning Set .037 .023 .017 - .067 (3;? Reason Enrolled .007 .006 .009 .004 .011 7 Pre-attitude .006 .009 .011 .020 .083 E ‘.033 50-50-1 (Asp.-Exp.) .065 .016 .030 .012 .033 .041 Pretest .056‘7003 .002 .032 .035 Instructor: Eys. Ext. .004 .010 .003 .028 .029 .021 Instructor: Eys. Neu. .008 .012 .010 .010 .014 .016 Instructor: Marlowe-Cue. .003 .008 .006 .005 .017 .014 Instructor: Sanford-Ggh. .005 .004 .008 .017 .027 .024 Instructor: California-F .013 .014 .011 .010 .027 .031 Instructor: Risk .007 .004 .006 .008 .027 .024 N 532 384 238 153 79 3 76 2 TSSi/TSST 1.000 .649 .382 .238 .135 .000 .123 .000 MEAN 3.662 3.385 3.550 3.399 3.608 5.000 3.553 5.000 PrOportion of variation in each group explainable for each predictor F’s-=Sp1it made on this variable. (BSS/TSSM @- Next best BSS/TSS. =Final group. a I=Split attempted but not made. NA =Split not attempted. 364 TABLE E4 (cont'd.) TABLE E4 (cont'd.) 365 Predictor Group Number 1 2 5 6 8 15 19 22* 23 24* 25 32* 33 4o 41* 52 53* 54* 55* Sex .004 .008 .008 .004 .002 .003 .005 .000 .001 .017 .053 .015 .061 Age .006 .006 .015 .017 .016 .014 .013 .019 .010 .016 .041 .083 .002 Major .012 .019 .018 .055“ ’.024 .009 .011 .032 .069 .047 .095 .157 .015 Credits .015 .002 .004 .005 .022 .028 .032 .032 .038 .052 .098 .119 .111 Current Load .007 .004 .004 .002 .012 .010 .018 .025 .021 .009 .026 .024 .023 GPA .195"? .030 .031 .024 .019 .018 .026 .027 .093 '983 =. .098 .080 CQT-V .058 . 09 .008 .012 .055 .064‘ ‘.018 .025 .052 .123 .19 .024 .038 CQT—Q .037 .019 .018 .028 .066 .038 .064 .o 7 .016 .026 .027 .024 .032 CQT-T .053 .007 .012 .009 .019 .021 .015 .020 .060 .055 .081 .068 .058 Extraversion .011 .007 .010 .024 .043 .041 .039 .052 .070 .060 .082 Neuroticism .007 .015 .059 ‘.020 .058 .067 .068 .081 . 00 .105 .074 1 Eysenck Lie .014 .013 .004 .019 .061 .056 .059 .051 an» I085 .107 .144 .1531!’ A-H Test Anxiety .047 .030 .029 .027 .014 .016 NA .039 NA .10 NA .055 . 14 NA .107 NA NA .103 Test Anxiety-1 .003 .001 .uz .003 .002 .003 .004 .003 .002 .005 .002 .006 .011 Test Anxiety-2 .010 .002 .003 .003 .013 .020 .028 .034 .034 .044 .026 .039 .023 Learning Set .037 .023 .017 .067 - .054 .080 .055 .075 .093 .130 Reason Enrolled .007 .006 .009 . 4 .011 .017 .008 .012 .007 .016 .191"""‘" '5 '51. 38 Pre-attitude .006 .009 .011 .020 .083 "'0. 33 .023 .012 .045 .036 .040 .077 .023 50-50-1 (Asp.-Exp.-1) .065 .016 .030 .012 .033 .041 .039 .047 .070 .071 .054 .040 .018 Pretest .056 .003 .002 .032 .035 .045 .038 .019 .054 .090 .061 inst. Extraversion .004 .010 .003 .028 .029 .021 .032 .012 .021 .019 .046 .24 .039 Inst. Neuroticism .008 .012 .010 .010 .014 .016 .013 .019 .045 .027 .040 .049 .032 Inst. Marlowe-Crowne .003 .008 .006 .005 .017 .014 .008 .010 .026 .006 .061 .021 .018 Inst. Sanford-Gough .005 .004 .008 .017 .027 .024 .071 . 10 .029 .011 .090 .018 .050 Inst. California-F .013 .014 .011 .010 .027 .031 .035 .031 .022 .077 .093 .130 Inst. Risk .007 .004 .006 .008 .027 .024 .0 9 .020 .029 .027 .026 .024 .023 N 532 384 238 153 79 76 74 4 70 9 61 6 55 37 18 34 3 2 32 TSSi/TSST 1.000 .649 .382 .238 .135 .123 .115 .006 .102 .009 .084 .006 .070 .045 .016 .036 .000 .000 .029 MEAN 3.662 3.385 3.550 3.399 3.608 3.553 3.514 2.500 2.571 2.889 3.672 4.500 3.582 3.784 3.167 3.676 5.000 5.000 3.594 Preportion of variation in that group explainable for each predictor (BSS/TSS)i "‘*=Split made on this variable. =Next best BSS/TSS. * =Final group. Q’=Split attempted but not made. NA=Sp1it not attempted. 366 .opoe uoc usn wounsouuo umammuo. .mmH\mmm umop uxoZmemHv .vouasouuo uoa umfimmnmz .aoouw Hoammur .oanmmuo> omsu :0 mode umaamlgrl\ . “Amme\mmmv ocuomooua sumo new manwamwfimxo macaw umsu cm somuwmum> mo acmuuomoum mmm.m ono.m smm.m oma.m osmdm oom.m omm.m mmm.m moo.m zoo: mmo. ooo. mno. noo. ooo. mmm. mmm. moo. ooo.H Home\ammo em me no A on mmm mmm smm «mm 2 2 moo. 3o. moo. ooo. ooo. Bo. seam £86285 moa. mno. omo. mmo. oHo. HHo. sao. moo. orseanooemoo ”nouosuuoam moo. moo. omo. mmo. moo. moo. ooo. moo. :msoorouooeom «nouosuoaam m .. mno. ooo. emo. moo. ooo. moo. moo. oasoeorosoanuz "nouosnuosm mnmv meo. omo. omo. omo. oHo. moo. moo. .aoz outdone "nouoauuasm m H. mao. omo. . mHo. mmo. moo. oHo. s... .uxm xonoomm "nouusuunem omo. omo. o8. omo. moo. moo... omo. % ooououo omo. mao. omo. smo. «mo. omo. omo. moo. A.oxm-.oamo arom-om sao. eao. ooo. omo. omo. HHo. ooo. ooo. oosueuusrono moo. _awqo_ mao. ooo. soo. ooo. ooo. moo. ooomonso condom ooo. omo. So. So. a :o. mmo. emo. uom meanness moo. mno. moo. smo. moo. moo. moo. oHo. mrnuoexss onus ooo. ooo. ooo. ooo. moo. m... Hoo. moo. arxuonxss ueoo moo. ooo. «8. <2 mmo. So. a omo. 3o. bores some min sea. mmo. ooo. oao. ooo. omo. mHo. sao. 6mm xoaoomm mso. oeo. omom m omo. omom omo. moo. moo. sneonuoesoz xenonmo .omma. meo. emo Hoo. omo. omo. moo. HHo. sonouoaseoxm xoaooxm saa. emo. mmo. ooo. moo. mmo. soo. mmo. . H-Hoo moo. mmo. nmo. moo. mmo. mao. oHo. emo. o-eoo son. omo. m . moo. mao. moo. so . mmo. >1Hoo omo. 3o. % @ Hmo. omo. gluon moo omo. mmo. Noo. moo. moo. soo. soo. moo. coon usoenso nee. omo. ooo. mao. moo. soo. moo. mao. someono moo. 46mm. omo. omor mmo. mmo. ooo. «no. Home: omo. Hmo. Hmo. mmo. smo. mao. ooo. ooo. 6mm moo. moo. HNH. ooo. ooo. moo. moo. ooo. xom rum rem .mu «mm a o m N H Honssz macaw acuomvonm A.w.ucoov «m mqm mmzu so some umammu.nrl\ “Amme\mmmv acuomwoum sumo Ham manwammamxo macaw umsu cm acmummuw> mo domuuoaoum oom.s noo.m oem.m omo.m Hem.m «Harm mms.m mem.m omo.m mmm.m moodm, zsmz Hoo. seo. emo. omo. omo. mmo. mmo. mme. mum. moo. ooo.H Emmo\emmo m mm mm on n as on ow me- awn «mm z mmo. mmo. omen. omo. mmo. moo. ooo. moo. xmmm1MumcH mmo. mom. omo. 3. ca mmo. 3o. m8. 538828 Beam mmo. m . doom. omo. mmo. ooo. ooo. moo. .smorouooaem “uoaH ama. oso. moo. oHo. NHo. moo. moo. .osorusoansz “66am omo. mom. ooo. moo. omo. omo. mmo. .soz xosoexm "seem ooH. mmo. oso. mmo. omo. mmo. omo. .uxm ausonmm ouasm RH. oso. mmo. mmo. So. So... mmo. onouono mHo. mma. omo. mmo. ooo. HHo. oao. atom-om ooo. doe. mmo. omo. oHo. mHo. ooo. ooo. ooouaouorono m . smo. omo. omo. mmo. ooo. ooo. moo. Honam condom Ammmv mmo. mso. mmo. mmo. omo. mmo. mmo. 66m measures .oeoa. ooH. oso. omo. omo. ooo. moo. omo. m-.xs< some ooo. ooo. ooo. ooo. moo. moo. Hoo. moo. H-.xs< some az az omo. ooH. a2 amo. ooo. eao. Hm.. omo. moo. .xs< once arm oma. m2. mmo. omo... Hz. a ma. So. 63 soaoamm omo. ooo. mmo. moo. omo. oHo. mao. Noo. .soz xosunxo mmo. ooo. moo. Hmo. oHo. «Ho. moo. HHo. .oxm soaooxm Noo. em. nmo. omo. mmo. omo. moo. mmo. H-9oo L44 3o. ooo. moo. mmo. mmo. 3o. mmo. oéoo oeo. mam. mmo. moo. ooo. mmo. moo. mmo. orooo omo. mmo. ® QU omo. mmo. ghee. moo oHo. ooo. oHo. omo. Hoo. moo. ooo. moo. osoa unannso moo. ems. oHo. omo. moo. moo. moo. mHo. oueoono 93% L2. mmo. oHoCoo. 3o. «S. Home: moo. mmo mmo. omo. moo. mmo. ooo. ooo. 6mm mmo. ooo. ooo. ooo. moo. Hoo. moo. ooo. xom ram soc and as ram om om oH s m m nonasz macho Houomvoum A.o.usooo so mam mo acmaacmcam ooo.H aao.m mmo.m moa.m mao.m mmm.m oam.m omoam mmmam moo.m zoo: ooo. mmo. oso. moo. omo. oao. omo. mmm. ooo. ooo.a amma\amma a me me o as o mm oma omm «mm 2 omo. mmo. duo». sao. moo. ooo. moo. anamnaouosauoaa mmo. mmo. omo. mao. mmo. sao. mao. arsaeaooaasouaooosaoosa moo. mmo. omo. ooo. ooo. ooo. moo. .rwo-oaoosomnaouosaooaa Hoo. moo. ooo. mmo. aao. moo. moo. .oao-csoanozuaouusaonsa omo. mmo. aso. mmo. omo. mao. moo. .soz aosoommuaouosaueaa mmo. ooo. moo. omo. omo. oao. so.. .uxm xosonmmuaouosaonea ooo. «8. ~81. u omo. moo... omo. @ ueouono omo. omo. omo. mmo. aao. oao. moo. A.exm-.eoao arom-om omo. moo. ooo. mmo. mao. ooo. ooo. coaoauumuoam moo. ooo. mao. mmo. ooo. ooo. moo. ooaaonso sooeum mmo. omo. mmo. mmo. omo. mmo. mmo. uom msaaauoa ooo. moo. moo. oao. ooo. moo. oao. mrmouaxa<.onoa ooo. ooo. ooo. ooo. moo. aoo. moo. a-muoaxs< 666a omo. @ SS. So. a o. omo. 3o. boasts coca :4 So. amo. mmo. So. @ So. So. 63 rotunma sz emoa. Sa. a2 a mz mmo. So. So. moo. aoaoaooasoz xuaonmm omo. ado. ado. sao. mao. moo. aao. soaouoaonoxm xenonmm mmo. ado. ooo. 66.. omo. moo. mmo. a-aoo mm . mar. n “2. @ mmo. So. mmo. oraoo . 1% omo. to. mmo. So. so . mmo. >-.._..oo omo. mmo. moo. sso. mmo. Ammwuruimoa. moo mmo. oao. omo. amo. moo. ooo. moo. mesa uncenso moo. moo. mmo. amo. moo. moo. mao. nuaouno ooo. moo. mmo. So... ooo. So. So. nomsz moo. moo. omo. omo. mao. ooo. ooo. oms oao. omo. aao. ooo. aoo. moo. ooo. rum ram som me 486 mm ssm aa s m a aonaaz macaw acucmooam A.w.uacov cm mam mo acauacmcam 00048 oc~.m ooo.m mow.m ooo.m w¢¢.q 00043 mm¢.¢ wc¢.¢ omm.m ooo.H oam.¢ mm~.¢ mmoqn zmmz «no. «00. ooo. mmo. ooo. mmo. ooo. who. «we. «so. ooo. mam. 0mm. ooo.H HmmH\ammH mm m m wm w ow OH mm mod as N mam mom «mm z 080. mmo. mmo. mmo. mmo. Hoo. omo. mmo. «no. xomm “acucaaaoaH woo. woo. omo. m .. wmo. moo. mmo. mmo. mmo. mnoaaacwaaco macucaaumaH moo. Noo. ammo. @ mmo. mmo. mmo. mmo. NNo. .swouoacmamm macucaauoaH woo. woo. mmo. omo. mmo. Hoo. moo. moo. ego. .caoucscmamz “acucaauoaH omo. mmo. omo. mmo. mmo. «do. oHo. mmo. omo. .acz xcacomm «acucaauoaH ooo. mmo. omo. omo. amo. mmo. moo. ooo. .oxo xoaoaoo ”nouusouesa mmo. uocacam omo. moo. omo. omo. om . ooo. mmo. - So. moo. omo. mmo. moo. . moan. n2. mmo. moonstone Tomrom Hoo. omo. mmo. omo. mmo. Hoo. mmo. moo. coaoaooouoao mmo. omo. oo.. moo. oo.. moo. ooo. oao. moo. ooaaoaso sonaoo omo.". m loom. @ moo. @ ooo. mmo. moo. aso. uom meanneoa ooo. ooo. mmo. mmo. moo. moo. oao. ooo. ooo. o-oooaxsm ueoa ooo. . . mao. ooo. aao. ooo. moo. moo. ooo. armuoaxsm coma doom. mz mz mmo. mz mo. moo. mmo. mz omo. mmo. mmo. muoaasm ueoa o-m ooo. moo mmo. mmo. moo. mmo. omo. mao. ooa xosoamo ooo. mmo. mmo. mmo. ooo. moo. moo. moo. soaoaooascz aosoemo. mmo. mmo. mmo. mmo. ooo. omo. omo. mmo. soaoaoasaoxo xosoomo moo. . mmo. mmo. omo411wwm. mmo. ooo. gaunt a-aoo omo. amos. mmoa moo. omo. moo. moa. omo. o-aoo omo. mmo. ooo. moo. mmo. omo. moo. mmo. >-aoo moo. omo. omo. mmo. omo. mmo. omom ”mm. moo moo. moo. ooo. moo. moo. aoo. ooo. moo. oooa oaoanso «ma. moo. moo. moo. moo. m... omo. mmo. nuaooao @ So. mmo.. NB. oS. ooo. mmo. 8?: ooo. mmo. om omo. ooo. moo. moo. moo. omm ooo. ooo. moo. moo. omo. amo. mmo. moo. moo mom «ms mom mm ram mom 4mm mm am om rma ma m a acnfiaz.macaw acucmocam canmmam> cmaacuupmz Scam acm mammaamamxm acamam Gmsums macaw an comaw cmaaco madam am .am> mo .mcam "HH QH< mm mam mo acaaacmcam ooo.m oom.m maa.m oom.m mo~.m aoo.m mmm.m oms.m aeo.m mam.m ommsm mmo.m zmoz moo. ooo. ooo. omo. omo. omo. moo. mmo. omo. mmo. omm. ooo.a amma\amma om m a mo om com a as moo moo mmm mmm z moo. moo. moo. m . mmo. omo. omo. omo. meao "sooossuoaa So. NS. 2a. RWU @ omo. So. So. orsaaaooaaso "souossuusa mom. mmo. ooo. omo. omo. mmo. omo. omo. .aoorouooesm «souusuunsa moo. oaa. ooo. moo. oao. moo. moo. ooo. .ocorosoaaez naouooauoaa moo. ooo. mmo. omo. amo. o .. aao. omo. .soz sonoamo usouosuuoaa So. So. So. omorr Loo. @ mmo. So. .uxo monsoon... 3363665 mmo. mmo. moo. omo. mmo. ooo. moo. mmo. uuouoso ooo. ooo. omo. ooo. mmo. ooo. mmo. mmo. m.oxo-.oeao aromuom omo. moo. ooo. moo. ooo. ooo. ooo. moo. oesoauus-oso mmo. mmo. mmo. moo. ooo. moo. moo. moo. ooaaoseo sousoo moo. mo. ooo. mmo. mmo. mmo. omo. amo. uum mamassoa mmo. omo. moo. mao. ooo. ooo. ooo. omo. mrmuuaxsm uooa .5 ooo. ooo. moo. ooo. ooo. ooo. moo. aoo. aumuoaxam area m ammo. mz mz moo. oS. 3o. mz mz moo. So. @ omo. muons... 88 am omo. omo. sea. ooo. moo. amo. mmo. moo. cam soaosmo a . mmo. om. moo. ooo. ooo. ooo. moo. anaoauosscz xenonoo Ammwv oao ooo. ooo. moo. moo. ooo. a o. soaouoassuxo xoaoomo ooo. mmo. omo. omo. omo. ooo. omo. Ammmu ataoo moo. moo. «no. So. 4 omov. to. mmo. omo. oaoo omo. mmo. ooo. mmo. moo omo. mmo. mmo. oraoo SSH ommn oSH omon amon mmo..n moo... mmo. moo omo m mo oao mao oao. moo. moo. coca usoaaoo Ra. @ So. So. So. So. mmo. euaooso ooo. o o. ooo.... mmo. moo. mmo. moo. mmo. some: moo. omo. ooo. oao. ooo. ooo. moo. moo. 6mm omo. mmo. moo. mmo. omo. moo. moo. moo. aom 4mm smm mas om ma ma sea so m m m a acnaaz macaw acucflooam A.o.ucooo mo momma 376 .voamEcuam uca uaamm u «z .ova occ uan wouchuao uammm I_9 .macaw madam u % mmimmo noon oxoz u® .cmncaamca 35 so come umamm ac aAwwH\mmmv acucaooam scam acm canaaacmmxo macaw umsa am acaummam> mo acmuacmcam amo.m mmm.m omm.m ooo.H oom.m ooa.m mma.m amo.m moo.m aom.m aoo.m mamsm oam.m mmosm zmoz ooo. ooo. mmo. ooo. moo. omo. moo. omo. ooo. omo. omo. mmo. omm. ooo.a aoma\omma mo ma mo a mo am m on om moo moo «ma mmm «mm 2 ooo. mmo. omo. ooo. moo. mm.. moo. omo. ooo. omo. memo “souosaunsa moo. ooo. moo. omo. omo. @ @ omo. So. So. orsaasooaaoo “souussunaa mmo. aao. mmo. ooo. aao. o. o o. mmo. amo. omo. .aoorouoossm "souossuaca ooo. moo. ooo. oao. moo. moo. ooo. moo. moo. ooo. .oso-osoassz “souossuosa omo. mmo. mmo. ooo. omo. omo. amo. a o. moo. omo. .soz xososoo snooossoesa mmo. mmo. to. ooo. omo. So... omo. @ So. So. .uxo soaosmo "nouossuesa ooo. mmo. moo. mmo. omo. ooo. mmo. ooo. moo. mmo. onouoso moo. moo. mmo. ooo. ooo. ooo. moo. moo. moo. mmo. A.oxo-.oaao aromrom mmo. ooo. ooo. mmo. mmo. moo. ooo. ooo. ooo. moo. oosoauusroso moo. moo. omo. ooo. ooo. moo. ooo. moo. moo. ooo. ooaaosso sousoo omo. mmo. omo. mmo. moo. moo. mmo. mmo. omo. ooo. uoo msaauoua mmo. omo... moon. “no. omo. So. So. ooo. ooo. omo. mrmuouam uuoa ooo. ooo. ooo. ooo. ooo. ooo. ooo. ooo. mo.. aoo. armuoaxam oooa mz mz moa. mz omo. ooo. mz moo. mmo. ooo. moo. So. Q omo. ouoaém ueoo. om moo. omo. mmo. moo. a. moo. moo. amo. mmo. mao. oaa xosoomo moo... I So. So. . lmlhmo om.. moo. ooo. ooo. ooo. soaoauousoz rusoaoo ooo. ooo. moo. nnmwv a . omo. moo. mmo. moo. soaosoassuxo soaoamo . o... omo. ooo. ooo. omo. ooo. omo. araoo ooz. ooo. AM! So. So. So. omo.. :o. mmo. odoo ooo. mmo. omo. moo. ooo. mmo. moo. omo. mmo. mmo. oraoo mmo. ooo. omo. moo. mmo. omo. omo. mmom mmomriomm. moo ooo. omo. mmo. omo. mmo. oao. moo. oao. moo. moo. coca usonuso mmo. omo. mmo. ooo. mmo. ooo. aao. omo. oao. mmo. ouaoouo ooo. moo. ooo. ooo. mmorrtmmo. ooo. moo. mmo. mmo. somsz moo. mmo. ooo. ooo. moo. ooo. ooo. oao. moo. moo. oom moo. moo. omo. moo. oao. mmo. omo. mmo. mao. ooo. xom xmo smo ram mom no mo 4mm om ma ma m o_ m a acnaaz macaw acacmpoam A.o.ueouo mo momma 377 TABLE E6 AID III: Proportion of Variation in Final Exam Score-l by Group within Branch Explainable for Each Final—Course Variable Predictor Group Number 1 3 5 34 35* 42* 43* Sex .002 .002 .000 .015 .007 .034 Age .002 .008 .008 .022 .082 .012 Major .027 .039 .053 .010 .091 .021 Credits .025 .023 .050 .054 .063 .097 Current Load .003 .002 .019 .040 .041 .063 GPA .063 (OH) .3304 .017 CQT-V . 3 . 64 .062 .071 .244 .046 CQT-Q .054 .040 .040 .050 .239 .092 CQT-T .065 .065 .087 .077 .044 Eysenck Extraversion .019 .024 .058 .034 .083 Eysenck Neuroticism .006 .003 .020 .021 .090 .048 Eysenck Lie .014 .008 .014 .045 .063 .036 A-H Test Anxiety .047 .050 .029 .036 .131 .102 Test Anxiety-1 .001 .000 .000 .000 .000 .002 Test Anxiety-2 .021 .011 .021 .033 125 .077 Test Anxiety-3 .016 .019 .050 .14I .021 . 02 Accuracy .003 .005 .015 .005 .026 .062 Success .011 .005 .038 .040 .039 .051 Learning Set .038 .016 .030 .029 NA .116 .1154 Reason Enrolled .001 .002 .013 .014 .028 .015 Pre-attitude .004 .006 .072 .062 .070 .058 Post-attitude .013 .010 .039 .018 .066 .024 Discussion Attitude .012 .014 .019 .014 .153 .047 50-50-1 (ASp.-Exp.-1) .054 .070 .101 .030 .153 .009 50-50-2’(Asp.-Exp.-2) .113 .071 .078 .069 .184 .063 Pretest .071 062 048 .062 .129 .029 Midterm .159 5125 £059 .056 .147 .037 Instructor Eysenck Ext. .008 .007 .0 1 .012 .084 .072 Instructor Eysenck Neu. .006 .005 ‘15? I021 .150 .045 Instructor Marlowe-Cue. .002 .004 . 6 .010 .026 .020 Instructor Sanford-Ggh. .005 .009 .046 .006 .052 .020 Instructor California-F .021 .030 .026 .007 .095 .060 Instructor Risk .021 .030 .034 .024 .054 .022 Instructor Course Load .023 .030 .021 .015 .041 .034 Instructor Teaching Exp. .008 .006 .013 .010 .117 .058 Instructor ED 200 .013 .009 .092 .015 .150 .027 N 532 382 90 75 15 31 44 TSS1/TSST 1.000 .556 .083 .055 .014 .014 .033 MEAN 29.254 30.272 32.567 33.080 30.00034.226 32.273 PrOpcrtion of variation in that group explainable for each predictor fl =Split made on this variable. -eNext best BSS/TSS. (BSS/TSS)1 Q, =Sp1it attempted but not made. NA =Split not attempted. 378 TABLE E6 (cont'd.) Predictor Group Number 1 3 4 7 38* 39* Sex .002 .002 .004 .003 .001 .049 Age .002 .008 .011 .015 .018 .075 Major .027 .039 .034 .042 .032 .165 Credits .025 .023 .015 .065 .038 .024 Current Load .003 .002 .003 .031 .082 .009 GPA @ .091 ”.021 .042 .3260' CQT-V .063 . 4 .038 .077 .074 .176 CQT-Q .054 .040 .036 8 .207 CQT-T .065 .065 ® . .040 .076 Eysenck Extraversion .019 .024 .024 .037 .078 .265 Eysenck Neuroticism .006 .003 .006 .066 .079 .161 Eysenck Lie .014 .008 .011 .049 .041 .098 A-H Test Anxiety .047 .050 .040 .095 .037 .176 Test Anxiety-1 .001 .000 .000 .003 .014 .000 Test Anxiety-2 .021 .011 .014 .018 .049 .053 Test Anxiety-3 .016 .019 .006 .019 .084 .062 Accuracy .003 .005 .008 .017 .025 .112 Success .011 .005 .004 .020 .049 .026 Learning Set .038 .016 .017 .034 .043 .093 Reason Enrolled .001 .002 .009 .094 .083 .136 Pre-attitude .004 .006 .013 .092 .1320 .102 Post-attitude .013 .010 .015 .072 .088 Discussion.Attitude .012 .014 .016 .059 .0 8 . 6 50-50-1 (Asp.-Exp.-1) .054 .070 .038 .061 .055 50-50-2 QASp.-Exp.-2) .113 .071 .020 .032 .I 0 .102 Pretest .071 .062 044 .065 .049 .106 Midterm .159 .125 .b28 .011 .038 .088 Instructor Eysenck Ext. .008 .007 .023 .022 .030 .060 Instructor Eysenck Neu. .006 .005 .011 .070 .084 .042 Instructor Marlowe-Crowne .002 .004 .011 .026 .017 .044 Instructor Sanford-Cough .005 .009 .024 .073 .053 .076 Instructor California-F .021 .030 .032 .018 .019 .044 Instructor Risk .021 .030 .032 .036 .069 .028 Instructor Course Load .023 .030 .038 .077 .068 .136 Instructor Teaching Exp. .008 .006 .007 .070 .084 .006 Instructor Education 200 .013 .009 .003 .044 .056 .044 N 532 382 292 61 38 23 TSSi/TSST 1.000 .556 .404 .071 .044 .017 MEAN 29.254 30.272 29.565 31.623 30.658 33.217 Proportion of variation in that group explainable for each predictor (BSS/TSS)1 v":|=Split made on this variable. -Next best BSS/TSS. Q; =Split attempted but not made. ' =Final group. NA ~Split not attempted 379 TABLE E6 (cont'd.) TABLE E6 (cont'd.) 380 Predictor Group Number 1 3 4 9 30 31 46* 47* 48* 49* Sex .002 .002 .004 .016 .007 .006 .038 .007 Age .002 .008 .011 .006 .007 .007 007 .005 Major .027 .039 .034 .026 .2 .070 .091 Credits .025 .023 .015 .019 2 .188 .023 .011 Current Load .003 .002 .003 .005 .008 .004 .007 .015 GPA .09 .044 .173 .043 .034 CQT-V . . 64 .038 . 16 .045 .078 .053 .038 CQT-Q .054 .040 .036 .007 .026 .191 .025 .038 CQT-T .065 .065 942$) .013 .031 .041 .053 .025 Extraversion .019 .024 .l 4 .020 .025 .135 .023 .034 Neuroticism .006 .003 .006 .009 .023 .100 .027 .035 Eysenck Lie .014 .008 .011 .007 .024 .046 .034 .031 A-H Test Anxiety .047 .050 .040 .015 .040 .021 .093 .072 Test Anxiety-1 .001 .000 .000 .003 .000 .000 .000 .000 Test Anxiety-2 .021 .011 .014 .013 .039 .073 .058 .023 Test Anxiety-3 .016 .019 .006 .005 .022 .039 .048 .092 Accuracy .003 .005 .008 .015 .056 .026 .078 .044 Success 011 .005 .040 .004 .012 035 .011 .018 Learning Set .038 .016 .017 .019 .11 .058 .020 NA NA NA .041 Reason Enrolled .001 .002 .009 .001 .008 .023 .028 .048 Pre-attitude .004 .006 .013 .014 .031 .137 .011 .014 Post-attitude .013 .010 .015 .023 .051 (§;Z> .017 .038 Discussion Attitude .012 .014 .016 .018 .046 . 5 .200 3' ”.017 50-50-1 (Asp.~Exp.-l) 054 .070 .038 .010 .033 .050 .095 .100 50-50-2 (Asp.-Exp.-2) .113 .071 .020 .010 .038 .074 .078 .067 Pretest .071 .062 .044 .036 . 13 .168 .0 8 Midterm .156 3'1. 25 "o. 28 .020 .023 .116 . 0 Instructor Extraversion .008 .007 .023 .016 .050 .061 .0 7 .032 Instructor Neuroticism .006 .005 .011 .006 .050 .061 .076 .059 Instructor Marlowe-Crowne .002 .004 .011 .008 .024 .047 .039 .036 Instructor Sanford-Gough .005 .009 .024 .015 .033 .008 .039 .042 Instructor California-F .021 .030 .032 .032 .023 .016 .025 .053 Instructor Risk .021 .030 .032 .032 .030 .019 .025 .005 Instructor Course Load .023 .030 .038 .032 .015 .044 .025 .036 Instructor Teaching Experience .008 .006 .007 .009 .022 .006 .025 .059 Instructor ED 200 .013 .009 .003 .002 .022 .016 .025 .011 N 532 382 292 228 90 30 59 15 15 7 51 TSSi/TSST 1.000 .556 .404 .293 .101 .046 .041 .012 .024 .001 .031 MEAN 29.254 30.272 29.565 29.022 29.822 28.400 30.627 30.133 26.667 33.571 30.157 Prcporticn of variation in that group explainable for each predictor (BSS/TSS)1 f“1§=Split made on this variable. Q=Next best BSS/T33. I=Final group. =Split attempted but not made. NA =Split not attempted. 381 TABLE E6 (cont'd.) 382 TABLE E6 (cont'd.) Predictor Group Number 1 3 4 6 8 16* 17 18* 19 20* 21 22 23* 26* 27 28 29* 32* 33* Sex .002 .002 .004 .016 .020 .025 .025 .032 .016 .007 .002 007 Age .002 .008 .011 .006 .008 .008 .010 .010 .010 .013 .013 .004 Major .027 .039 .034 .026 .052 .046 .044 .046 .045 4934 .094. s ”.047 Credits .025 .023 .015 .019 .012 .014 .019 .017 .014 .o 2 .027 .045 Current Load .003 .002 .003 .005 .015 .004 .002 .003 .003 .010 .023 .024 GPA (gép .091 75:29 .034 .026 .030 .024 .036 .047 .047 .0786 CQT-V . 63 . 4 .038 .016 .025 .035 .038 .059 .052 .029 .036 .048 CQT-Q .054 .040 .036 .007 .021 .040 .036 .031 .028 .031 .051 .071 CQT-T .065 .065 as) .013 .028 as» .049 @ .064 V ‘.038 .049 .071 Extraversion .019 .024 .0 4 .020 .041 .037 . 2 .036 .036 .052 .061 Neuroticism .006 .003 .006 .009 .022 .017 .019 .019 .019 .021 .024 .042 Eysenck Lie .014 .008 .011 .007 .011 .021 .056 5“ 2616 .011 .030 .023 .032 A-H Test Anxiety .047 .050 .040 .015 .034 .021 .011 .010 .015 .031 .016 .027 Test Anxiety-1 .001 .000 .000 .003 .003 .005 .006 .005 .007 .006 .010 .008 Test Anxiety-2 .021 .011 .014 .013 .029 .029 .013 .014 .020 .026 .023 .034 Test Anxiety-3 .016 .019 .006 .005 .005 .003 .003 .007 .002 .000 .003 .026 Accuracy .003 .005 .008 .015 .006 .003 .008 .010 .010 .008 .009 .018 Success .011 .005 .004 .004 .001 .002 .006 .015 030 .019 .024 .032 Learning Set .038 .016 .017 .019 .026 NA .036 NA .040 NA .06 .011 NA .010 .019 NA NA .052 Reason Enrolled .001 .002 .009 .001 .004 .006 .005 .008 .007 .004 .005 .024 Pro-attitude .004 .006 .013 .014 .026 .026 .025 .020 .023 .026 .031 .047 Post-attitude .013 .010 .015 .023 .014 .018 .023 .017 .020 .030 .038 .024 Discussion Attitude .012 .014 .016 .018 .017 .021 .018 .011 .009 .017 .010 .038 50-50-1 6Asp.-Exp.-1) .054 .070 .038 .010 .013 .017 .019 .020 .023 .036 .032 .038 50-50-2 (Asp,-Exp.-2) .113 .071 .020 .010 .002 .010 .003 .001 .001 .002 .007 .011 Pretest .071 .062 .044 .036 .045 .066 ‘r" "5022 .029 .028 .015 .010 .046 Midterm .159 "3525 3038 .020 .016 .013 .017 .028 .019 .036 .019 .019 Inst. Extraversion .008 .007 .023 .016 .070 .027 .015 .022 .032 .033 .029 .044 . Inst. Neuroticism .006 .005 .011 .006 .047 .039 .027 .037 .048 .026 .021 Inst. Marlowe-Crowne .002 .004 .011 .008 .016 .010 .008 .015 .0 2 .016 .035 .040 Inst. Sanford-Cough .005 .009 .024 .015 .033 .018 .020 .027 .031 .030 .035 .019 Inst. California-F .021 .030 .032 .032 .092 ‘ #047 .042 .038 .048 .069 3018 ' .034 Inst. Risk .021 .030 .032 .032 .039 .039 .045 .054 .055 .023 .032 Inst. Course Load .023 .030 .038 .032 .o90 .012 .010 .018 .038 .054 . Inst. Teaching Exp. .008 .006 .007 .009 .010 .012 .015 .012 .016 .024 .104 . 07 Inst. ED 200 .013 .009 .003 .002 .000 .011 .012 .014 .018 .036 .016 .012 N 532 382 292 228 138 11 127 2 125 7 118 116 1 10 106 93 13 16 77 TSSi/TSST 1.000 .556 .404 .293 .181 .013 .152 .001 .142 .006 .128 .116 .000 .006 .103 .085 .011 .017 .060 MEAN 29.254 30.272 29.565 29.022 28.500 25.000 28.803 22.500 28.904 32.000 28.720 28.836 20.000 31.300 28.604 28.892 26.538 30.813 28.494 Proportion of variation in that group explainable for each predictor (BSS7TSS)i “=Split made on this variable. -Next best BSS/TSS. av=Split attempted but not made. =Final group. NA =Sp1it not attempted. 383 TABLE E6 (cont'd.) 384 TABLE E6 (cont'd) . Predictor Group Number 1 2 10 12 13 14 15 24 25* 36 37* 40* 41 44* 45* 50* 51* 52* 53* Sex .002 .009 .006 .011 .003 .004 .021 .021 .034 .003 .032 .002 Age .002 .012 .007 .142 .004 .028 .008 .023 .048 0 2 0 3 .041 Major .027 .026 .025 .141 .032 .047 .15( .022 .177"-r-;?041 ”.022 " .085 Credits .025 .025 .030 .145 .029 .068 .045 .077 .086 .114 .057 .139 Current Load .003 .004 .005 .079 .017 .013 .017 .031 .023 .003 .017 .066 GPA .017 .030 .115 .096 .043 CQT-V . 3 .032 .031 .047 (21:; .2624 CQT-Q .054 .079 .059 .099 .049 .071 CQT-T .065 ® .128 .131 .187 Extraversion .019 .017 .022 .134 .2354 Neuroticism .006 .037 .041 .125 .047 .091 .0 2 .126 .070 .022 .083 . 1 Eysenck Lie .014 .044 .074 .033 w) .092 @ .091 .091 .159 .131 A-H Test Anxiety .047 .030 .030 .143 .27 .037 . 39 .061 .021 . 5 .039 .062 Test Anxiety-1 .001 .008 .008 .000 .008 .000 .013 .000 .000 .000 .010 .000 Test Anxiety—2 .021 .045 .042 .167 .034 .051 .028 .030 .040 .026 .134 .051 Test Anxiety-3 .016 .004 .009 .140 .006 .027 .052 .039 .062 .056 .046 .127 Accuracy .003 .002 .006 .005 .006 .025 .026 .029 .028 .036 .018 .139 Success .011 .009 .008 .067 .008 .021 .006 .026 .038 .008 .012 .006 Learning Set .038 .043 .043 .091 .068 .057 .113 NA .159 NA NA .124 .075 NA NA .098 NA NA Reason Enrolled .001 .006 .009 .198 .011 .013 .041 . 01 .011 .045 .186 .022 Pre-attitude .004 .029 .029 .034 .020 .062 .030 .086 .037 .037 .103 .031 Post-attitude .013 .052 .056 .147 .060 .097 .057 .133 .037 .10%___11£____§W.218 Discussion Attitude .012 .013 .033 .034 .042 .066 .040 .090 .147 .20 .069 .194 50-50-1 (Asp.-Exp.-1) .054 .Tlmm .040 .013 .010 .054 .034 .o o .074 .098 .094 50-50-2 (Asp.-Exp.-2) .113 .034 .035 .146 .049 .061 .027 .105 in» .115 .039 .219 Pretest .071 .040 .047 .090 .035 02% .047 . 100 .070 . 030 Midterm .159—”.692 .115 3'0. 37 .020 .11 .057 .070 .062 .060 .122 .127 Inst. Extraversion .008 .008 .006 .190 .010 .024 .128 .018 .057 .055 .199 .139 Inst. Neuroticism .006 .003 .004 .190 .008 .014 .042 .025 .057 .052 .054 .139 Inst. Marlowe-Crowne .002 .011 .011 .106 .017 .019 .052 .022 .026 .099 .058 105 Inst. Sanford-Gough .005 .011 .011 .066 .009 .033 .060 .036 .038 .148 .070 .159 Inst. California-F .021 .002 .002 .081 .005 .021 .081 .007 .053 .044 .120 .039 Inst. Risk .021 .007 .011 .106 .028 .011 .061 .029 .043 .070 .076 .058 Instructor Course Load .023 .010 .009 .099 .011 .018 .031 .011 .063 .133 .162 .123 Instructor Teaching Exp. .008 .002 .003 .063 .003 .010 .127 .008 .021 .052 .201 .043 Instructor ED 200 .013 .002 .003 .021 .005 .018 .077 .039 .042 .048 .105 .048 N 532 146 144 25 119 69 49 54 15 50 4 16 34 32 17 9 24 6 19 TSSi/TSST 1.000 .276 .245 .031 .187 .119 .048 .073 .033 .057 .002 .015 .032 .024 .016 .006 .020 .005 .016 MEAN 29.254 26.623 26.736 23.880 27.336 26.333 28.653 27.019 23.867 26.600 32.250 28.563 25.676 29.500 27.059 27.889 24.875 20.667 24.895 Proportion of variation in that group explainable for each predictor (BSS/TSS)i "_‘*=Split made on this variable. ”=Next best BSS/TSS. ’1: =Sp1it attempted but not made. ' =Final group. NA =Split not attempted. 385 TABLE E6 (cont'd.) Predictor Group Number 1 2 11* Sex .002 .009 Age .002 .012 Major .027 .026 Credits .025 .025 Current Load .003 .004 GPA .017 CQT-V . 63 .032 CQT-Q .054 .079 CQT-T .065 (139 Eysenck Extraversion .019 .0 7 Eysenck Neuroticism .006 .037 Eysenck Lie .014 .042 A-H Test Anxiety .047 .030 Test Anxiety-1 .001 .008 Test Anxiety-2 .021 .045 Test Anxiety-3 .016 .004 Accuracy £003 .002 Success .011 .009 Learning Set .038 .043 NA Reason Enrolled .001 .006 Pre-attitude .004 .029 Post-attitude .013 .052 Discussion Attitude .012 .013 50-50-1 (ASp.-Exp.-1) .054 .113 :1 50-50-2 QASp.-Exp.-2) .113 .034 Pretest .071 .040 Midterm .159 .092 Instructor Eysenck Ext. .008 .008 Instructor Eysenck Neu. .006 .003 Instructor Marlowe-Crowne .002 .011 Instructor Sanford-Gough .005 .011 Instructor California-F .021 .002 Instructor Risk .021 .007 Instructor Course Load .023 .010 Instructor Teaching Experience .008 .002 Instructor Education 200 .013 .002 N 532 146 1 TSSilTSST 1.000 .276 .000 MEAN 29.254 26.623 10.000 Prcporticn of variation in that group explainable for each predictor “=Sp1it made on this variable. (BSS/TSS)j_ =Next best BSS/TSS. $.=Sp11t attempted but not made. =Final group. NA =Sp1it not attempted. 386 TABLE E7 AID III: Proportion of Variation in Final Exam Score-2 by Group within Branch Explainable for Each Final-Course Variable 387 TABLE E7 (cont‘d.) Predictor Gr0up Number 1 3 6 9 13 20 21 28* 29 40* 41* 58 59* 68* 69* Sex .000 .010 .007 .003 .002 .000 .000 .002 .064 .017 .003 .012 Age .006 .024 .039 .039 .058 .007 .088 .157 5.000 "‘.003 .061 .093 Major .017 .021 .027 .053 .028 @335 .058 .055 .1724 .079 .428 "—""" ”1. 04 Credits .006 .012 .012 .020 .057 .055 .072 .024 .125 .071 .092 Current Load .006 .008 .009 .012 .006 .056 .023 .034 .022 .096 .021 .003 GPA .188 ’.‘071 .064 .067 .085 .052 .050 .137 .1852 .065 .110 CQT-V .101 .100 .070 .018 .025 .059 .027 .041 .054 150 071 .093 CQT-Q . .090 .093 .07 .053 .054 .321 030 .022 .072 .150 ‘!054 2’ .147 CQT-T Q9 @ .116—m .228 .044 .046 .070 .092 .108 .188 Extraversion .015 .008 .010 .022 .o ' 4 .123 .010 .041 .057 .116 .142 .093 Neuroticism .009 .012 .022 .040 .067 .175 .112 5 "o. 31 .022 .022 .069 .093 Eysenck Lie .016 .013 .018 .020 .041 .104 .090 .041 .022 .120 .216 .128 A-H Test Anxiety .055 .028 .028 .022 .058 .252 .029 .037 .062 .070 .053 .067 Test Anxiety-1 .002 .025 .030 .006 .010 .000 .020 .031 .120 .000 .000 .000 Test Anxiety-2 .011 .010 .012 .024 .033 .009 .063 .030 .036 .057 .076 .029 Test Anxiety-3 .006 .003 .005 .011 .008 .016 .014 .015 .008 .054 .047 .063 Accuracy .010 .004 .005 .008 .015 .052 .025 .035 .008 .029 .064 .158 Success .012 .007 .007 .010 .008 .029 .035 .038 .089 .116 .091 .101 Learning Set .049 .040 .030 .029 .051 .056 .087 NA .032 .049 .144 .177 NA NA .099 Reason Enrolled .006 .009 .008 .006 .008 .053 .006 .006 .076 .013 .004 .034 Pre-attitude .001 .015 .022 .016 .038 .071 .089 .059 .109 .132 .197 Post-attitude .006 .034 .030 .026 .053 .065 my .053 .030 .099 .040 .108 Discussion Attitude .004 .020 .013 .030 .070 .185 .037 .050 .129 .202 .2790 . 50-50-1 (Asp.-Exp.-1) .049 .144 "?’032 .057 .053 .085 .036 .029 .089 .047 .037 .067 1 50-50-2 (Asp.~Exp.-2) .089 .057 .053 .054 .027 .011 .003 .016 .101 .047 .071 .093 Pretest .049 .037 .034 .029 .054 .154 .028 .033 .099 .100 .108 Midterm .126 .070 .060 .071 .087“ 3.055 ”.013 .017 .036 .071 .128 Instructor Extraversion .003 .004 .003 .015 .005 .028 .005 .017 .058 .042 .065 .090 ' Instructor Neuroticism .010 .009 .005 .027 .015 .085 .011 .014 .092 .027 .138 .125 Instructor Marlowe-Crowne .006 .011 .007 .013 .025 .053 .014 .013 .034 .020 .102 , .083 Instructor Sanford-Gough .005 .007 .008 .014 .013 .129 .004 .022 .048 .018 .142 .156 Instructor California-F .005 .010 .007 033.010 .070 .029 .026 .051 .005 .174 .128 Instructor Risk .011 .013 .019 .097‘ ’.‘030 .128 .018 .015 .044 .027 .262 .208 Instructor Course Load .014 .035 .035 .048 .012 .101 .003 .026 .058 .063 .099 .128 Instructor Teaching Experience .003 .003 .003 .025 .011 .066 .007 .017 .058 .063 .078 .090 Instructor ED 200 .015 .010 .003 .013 .033 .139 .006 .012 .044 .020 .239 N 532 213 212 152 115 29 85 9 75 42 33 25 4 2 23 TSSi/TSST 1. 000 358 306 183 121.030 080.009 .061.024 .027 .015 .005 .000 .009 MEAN 24.615 26 714 26. 811 27. 552 28 113 26 621 28. 635 25.899 28. 920 29.833 27. 758 25.960 30. 750 31.000 25. 522 Proportion of variation in that group explainable for each predictor (BSS/TSS)1 “=Sp1it made on this variable. -=Next best BSS/TSS. 4' =Split attempted but not made. * =Final group. NA =Sp1it not attempted. 388 TABLE E7 (cont'd.) Predictor Group Number 1 3 6 9 12 44 45* 56* 57* Sex .000 .010 .007 .003 .110 .046 .072 Age .006 .024 .039 .039 .024 .026 r945 d Major .017 .021 .027 .053 .190 .322” .058 Credits .006 .012 .012 .020 .013 .013 .210 Curr. Ld. .006 .008 .009 .012 .022 .020 .018 CPA .188 ”.071 {E0 .064 .170 .143 .060 CQT-V .101 .100 .o 0 .018 .069 .064 .3070 CQT-Q .090 .093 .074 .073 .057 .131 .129 CQT-T £33 .116 my. (216) .041 .119 Extravert . 15 .008 .010 .0 2 .096 .118 .038 Neurotic .009 .012 .022 .040 .062 .050 .115 Eye. Lie .016 .013 .018 .020 .013 .156 . 3 A-H Anx. .055 .028 .028 .022 .096 Anx.-1 .002 .025 .030 .006 .000 .002 .000 Anx.-2 .011 .010 .012 .024 .051 .110 .134 Anx.-3 .006 .003 .005 .011 .088 .030 .082 Accuracy .010 .004 .005 .008 .019 .021 .039 Success .012 .007 .007 .010 .011 .010 .137 Learn. Set. .049 .040 .030 .029 .106 .171 NA .087 NA Reason En. .006 .009 .008 .006 .032. .036 .041 Pre-att. .001 .015 .022 .016 .080 .098 .091 Post-art. .006 .034 .030 .026 .060 .083 .056 Disc. Att. .004 .020 .013 .030 .080 .091 .171 50-50-1 .049 .14Z ”’0. 32 .057 203% .072 50-50-2 .089 .057 .053 .054 .26 ' .028 .034 Pretest .049 .037 .034 .029 .021 .073 .178 Midterm .126 .070 .060 .071 .127 .156 .134 Inst. Ext. .003 .004 .003 .015 .005 .004 .143 Inst. Neu. .010 .009 .005 .027 .003 .023 .138 Inst. M-C .006 .011 .007 .013 .005 .004 .143 Inst. S-G .005 .007 .008 .014 .003 .023 .138 Inst. Cal-F .005 .010 .007 .033 005 .023 .143 Inst. Risk .011 .013 .019 .097 . 05 .004 .143 Inst. Lcad .014 .035 .035 .048 .005 .005 .143 Inst. Teach .003 .003 .003 .025 .005 .010 .143 Inst. ED200 .015 .010 .003 .013 .003 .023 .059 N 532 213 212 152 37 33 4 22 11 TSSilTSST 1.000 .358 .306 .183 .045 .031 .002 .013 .008 MEAN 24.615 26.714 26.811 27.552 25.811 25.242 30.500 26.364 23.000 Prcporticn of variation in that group explainable for each predictor “=Split made on this variable. (BSS/TSS)1 =Next best BSS/TSS. Q =Split attempted but not made. * =Final group. NA =Sp1it not attempted. 389 TABLE E7 (cont'd.) 390 TABLE E7 (cont'd.) Predictor Group Number 1 3 6 7* 8 26* 27 30* 31 42 43* 50 51* 66* 67* Sex .000 .010 .007 .007 .004 .028 .017 .015 .003 Age .006 .024 .039 .023 .024 .037 .035 .116 .033 Major .017 .021 .027 .025 .044 .082 .101 .108 .178 Credits .006 .012 .012 .038 .059 .037 .012 .024 .067 Current Load .006 .008 .009 .033 .041 .050 .076 .030 .120 GPA .188 T071 <é§g> .062 .053 .024 .014 .242 CQT-V .101 .100 70 .062 .087 .064 .041 .100 .043 CQT-Q .090 .093 .074 .073 .053 .029 .028 .153 .055 CQT-T . 116 5‘102 .071 .038 .019 .049 .079 Extraversion . 15 . 08 .010 .042 .052 .017 .028 .108 .162 Neuroticism .009 .012 .022 .028 .042 .065 .098 .124 .120 Eysenck Lie .016 .013 .018 .060 .010 .005 .023 .056 .056 A-H Test Anxiety .055 .028 .028 .054 .052 .067 072 .212 .133 Test Anxiety-1 .002 .024 .030 .085 .102 .182 .000 .000 .000 Test Anxiety-2 .011 .010 .012 .068 .083 .076 @ .100 .188 Test Anxiety-3 .006 .003 .005 .039 .039 .088 97 .159 .071 Accuracy .010 .004 .005 .025 .027 .071 .076 .4 1 .2743: Success .012 .007 .007 .012 .015 .033 .049 .011 .140 Learning Set .049 .040 .030 NA .039 NA .027 NA .023 .045’ NA .062 .126 NA NA Reason Enrolled .006 .009 .008 .012 .003 .028 .025 .080 .040 Pre-attitude .001.015 .022 11!).144 . ”023 .025 .035 .192 Post-attitude .006 .034 .030 .0 3 .0 4 .076 .043 1 .096 Discussion Attitude .004 .020 013 .086 q? .151"—""" $138 "9075 50-50-1 (ASp.-Exp.-1) .049 .14: g032 . .058 .067 . 0 .027 .016 50-50-2 (Asp.-Exp.-2) .089 .057 .053 ‘ .11 .026 .050 .076 .101 .032 Pretest .049 .037 .034 .061 .075 .023 .025 .074 .120 Midterm .126 .070 .060 .048 .046 .037 .035 .159 .075 Instructor Extraversion .003 .004 .003 .029 .054 .104 .071 .108 .040 Instructor NeuroticiSm -.010 .009 .005 .021 .032 .084 .089 .108 .100 Instructor Marlowe—Crowne .006 .011 .007 .060 .060 .054 .100 .108 .043 Instructor Sanford-Gough .005 .007 .008 .060 .060 .058 .050 .108 .115 Instructor California-F .005 .010 .007 .045 .036 .077 .074 .090 .029 Instructor Risk .011 .013 .019 .033 .048 .045 .046 .110 .188 Instructor Course Load .014 .035 .035 .076 .076 .064 .070 .108 .079 Instructor Teaching Experience .003 .003 .003 .054 .097 .093 .068 .100 .041 Instructor ED 200 .015 .010 .003 .032 .032 .044 .082 .030 .187 N 532 213 212 1 60 1 59 7 50 47 2 25 22 6 19 TSSi/TSST 1. 000 .358 .306 .000 .087 .000 .078 .002 .056 .042 .003 .020 .016 .002 .008 MEAN 24. 615 26. 714 26. 811 6.000 24.933 16. 000 25.085 22.143 25.760 26.085 19.500 27.080 24.954 30.167 26.105 Proportion of variation in that group explainable for each predictor (BSS/TSS)i "_Yl=Split made on this variable. ”=Next best BSS/TSS. Q' =Split attempted but not made. ‘ =Final group. NA =Split not attempted. 391 TABLE E7 (cont'd.) Predictor Group Number 1 2 5 22 23 32* 33 38* 39* Sex .000 .000 .012 .000 .000 .000 Age .006 .001 .002 .OI9 .008 .0 2 Major .017 .017 .101 flit (OED Credits .006 .001 .015 I 0 .0 4 .010 Curr. Ld. .006 .004 .013 .028 .026 .013 CPA .188 3024 .029 .080 .076 .068 CQT-V .101 .061 .057 .057 .054 .055 CQT-Q .090 {£36 .022 .049 .051 .037 CQT-T .108 .059 .061 .057 .057 Extravert . 5 .022 .015 .010 .010 010 Neurotic .009 .013 .058 .087 .114 t. 140 : Eys. Lie .016 .029 .015 .029 .043 .051 A-H Anx. .055 .024 .027 .025 .022 .021 Anx.-1 .002 .000 .000 .000 .000 .000 Anx.-2 .011 .008 .005 .010 .024 .017 Anx.-3 .049 .009 .013 .009 .062 .0884 Accuracy .010 .011 .027 .047 .068 .031 Success .012 .005 .006 .010 .004 .018 Learn. Set .049 .058 .134 NE T026 NA .041 .049 NA Reason En. .006 .003 .003 .013 .030 .008 Pre-att. .001 .013 .030 .020 .014 .011 Post-att. .006 .010 .052 .031 .049 .047 Disc. Act. .004 .007 .007 .020 .014 .030 50-50-1 .049 .009 .013 .009 .017 .020 50-50-2 .089 .036 .074 .076 .056 .056 Pretest .049 .034 .047 .030 040 .035 Midterm .126 .051 .126 ' .3057 .056 Inst. Ext. .003 .004 . 6 .033 .033 .031 Inst. Neu. .010 .015 .016 .010 .008 .005 Inst. M-C .006 .010 .045 .052 .055 .058 Inst. S-G .005 .003 .045 .052 .055 .058 Inst. Ca1-F..005 .006 .064 .052 .052 .040 Inst. Risk .011 .019 .039 .014 .017 .008 Inst. Load .014 .004 .026 .022 .025 .032 Inst. Teach .003 .007 .008 .007 .022 .018 Inst. ED200 .015 .025 .016 .015 .020 .018 N 532 314 83 6 76 6 70 69 1 TSSilTSST 1.000 .449 .106 .014 .078 .005 .063 .056 .000 MEAN 24.615 23.207 25.096 20.833 25.447 22.000 25.742 25.855 18.000 Proportion of variation in that group explainable for each predictor d‘fii=Split made on this variable. =Next best BSS/TSS. =Final group. (BSS/TSS)1 4 =Split attempted but not made. NA -Sp1it not attempted. 392 TABLE E7 (cont'd.) Predictor Group Number 1 2 4 11 19 48 49* 62* 63* Sex .000 .000 .004 .013 .013 .002 .051 .026 Age .006 .001 .005 .004 .009 .067 .020 .091 Major .017 .017 .010 .027 .083 .114 .209 .159 Credits .006 .001 .008 .035 .027 .036 .190 .033 Curr. Ld. .006 .004 .008 .022 .079 .015 .123 .026 GPA .18 .024 .027 .026 .104 .128 .086 .095 CQT-V .101 .061 .010 .015 .081 .221 .059 .251 CQT-Q .090 am .023 .015 .036 .155 .2845 .095 CQT-T .10: .010 .100 .169 .122 .251 Extravert .015 .022 .031 . 3 .009 .018 .057 .125 Neurotic .009 .013 .017 .049 .054 .074 .079 .075 Eys. Lie .016 .029 .039 .007 .042 .149 .070 .054 A-H.Anx. .055 .024 .007 .028 .049 .192 .095 .070 Anx.-1 .002 .000 .001 .000 .000 .000 .000 .000 Anx.-2 .011 .008 .017 .002 .020 .025 .052 .029 Anx.-3 .006 .001 .002 .003 .026 .042 .134 .035 Accuracy .010 .011 .007 .019 .016 .010 .008 .033 Success .012 .005 .007 .008 .039 E012 .059 .008 Learn. Set. .049 .058 .052 .15 .117 .190 .112 NA Reason En. .006 .003 .004 .020 .055 .108 .042 .062 Pre-att. .001 .013 .018 .036 .077 .043 .134 .055 Post-att. .006 .010 .009 .034 .093 .050 .178 .075 Disc. Att. .004 .007 .010 .021 .028 .059 .152 50-50-1 .049 .009 .022 .015 .015 ."4 .021 .036 50-50-2 .089 .036 .020 .019 .068 .065 .063 .075 Pretest .049 .034 .034 .022 .28 .115 . 93 Midterm .126 .051 .05 .020 . .070 .035 Inst. Ext. .003 .004 .005 .004 .048 .134 .3244 Inst. Neu. .010 .015 .013 .025 .048 .034 .059 .121 Inst. M-C .006 .010 .006 .044 .048 .134 .007 .320 Inst. S-G .005 .003 .014 .022 .060 .181 .012 .271 Inst. CaleF .005 .006 .007 .022 .014 .222 .015 .194 Inst. Risk .011 .019 .017 .035 .048 .134 .059 Inst. Load .014 .004 .010 .040 .043 .110 .015 . 0 Inst. Teach .003 .007 .027 .08 .009 .034 .009 .014 Inst. Ed200 .015 .025 .023 .045 .048 .178 .059 .188 N 532 314 231 112 49 27 22 23 4 TSSi/TSST 1.000 .449 .295 .124 .042 .022 .013 .013 .003 MEAN’ 24.615 23.207 22.528 23.357 24.347 23.407 25.500 22.826 26.750 Prcporticn of variation in that group explainable for each predictor dsSplit made on this variable. Next best BSS/TSS. * =Final group. (BSS/T53)i =Sp1it attempted but not made. NA =Sp1it not attempted. 393 TABLE E7 (cont'd.) 394 TABLE E7 (cont'd.) Predictor GrOup Number 1 2 4 11 18 34 35 46* 47 54* 55* 64* 65* Sex .000 .000 .004 .013 .016 .000 .221 .007 .000 Age .006 .001 .005 .004 .008 .020 .024 .082 037 Major .017 .017 .010 .027 .040 ‘33) _221 .18 .120 Credits .006 .001 .008 .035 .081 .I55 .080 .081 .117 Current Load .006 .004 .008 .022 .013 .027 .051 .018 .003 GPA .188 .024 .027 .026 .069 .17 .102 .018 .042 CQT-V .101 .061 .010 .015 .023 .066 .043 .103 .154 CQT—Q .090 $023 .015 .028 .035 .098 .063 .084 CQT-T . 06 .010 {Jafi .040 .094 .088 .088 .077 Extraversion .015 .022 .031 .043 .12 . 25 .074 .020 .030 NeuroticiSm .009 .013 .017 .049 .044 .125 .059 .080 Eysenck Lie .016 .029 .039 .007 .013 .032 .077 .015 .062 A-H Test Anxiety .055 .024 .007 .028 .032 .036 .054 .035 .048 Test Anxiety-1 .002 .000 .001 .000 .003 .001 .000 .000 .000 Test Anxiety-2 .011 .008 .017 .002 .010 .047 .022 .035 .054 Test Anxiety-3 .006 .001 .002 .003 .040 .090 .115 .090 .085 Accuracy .010 .011 .007 .019 .013 .032 .029 .054 .042 Success .012 .005 .007 .008 .058 .008 dflflb .018 .003 Learning Set .049 .058 .052 .036 .039 .055 NA .036 .060 NA NA NA Reason Enrolled .006 .003 .004 .020 .029 .037 .051 .032 .015 Pre-attitude .001 .013 .018 .036 .020 .104 .115 .114 .121 Post-attitude .006 .010 .009 .034 .044 .112 .065 .081 .117 Discussion Attitude .004 .007 .010 .021 .059 .066 .115 .057 .045 50-50-1 (ASp.-Exp.-l) .049 .009 .022 .015 .038 .025 .014 .047 .085 50-50-2 (ASp.-Exp.-2) .089 .036 .020 .019 .027 .082 .029 .1994 Pretest .049 .034 .034 .022 .044 .130 .149 . 5 .052 Midterm .126 .051 .056 .020 .065 .067 .093 .055 .038 Instructor Extraversion .003 .004 .005 .004 .015 .047 .080 .084 .108 Instructor Neuroticism .010 .015 .013 .025 .032 .046 .067 .072 .072 Instructor Marlowe-Crowne .006 .010 .006 .044 .016 .027 .080 .057 .108 Instructor Sanford-Gough .005 .003 .014 .022 .023 .055 .080 .075 .122 Instructor California-F .005 .006 .007 .022 .008 .048 .350 .079 .103, Instructor Risk .011 .019 .017 .035 .020 .049 .513 .079 .078 W Instructor Course Load .014 .004 .010 .040 .032 .034 .167 .088 Instructor Teaching Experience .003 .007 .027 .082 .025 .010 .057 .037 .052 Instructor ED 200 .015 .025 .023 .045 .006 .013 .097 .050 .129 N 532 314 231 112 63 43 20 3 4O 37 3 16 4 TSSi/TSST 1.000 .449 .295 .124 .072 .042 .021 .002 .032 .026 .001 .006 .004 MEAN 24.615 23.207 22.528 23.357 22.587 23.326 21.000 19.000 23.650 23.973 19.667 22.063 16.750 Proportion of variation in that group explainable for each predictor (BSS/TSS)i f"3==Sp1it made on this variable. =Next best BSS/TSS. * =Final group. 4 NA =Sp1it attempted but not made. =Split not attempted. “2.94,. . . F'rxut 395 TABLE E7 (cont'd.) 396 TABLE E7 (cont'd.) Predictor ngggpjygfl§§;g 1 2 4 10 14* 15 16* 17 24 25 36 37* 52* 53* 60* 61* Sex . 000 . 000 .004 . 000 . 000 . 028 .000 . 000 .005 .002 . 000 .005 Age .006 .001 .005 .002 .008 .186 .013 .010 .000 .002 .033 .029 Major .017 .017 .010 .036 .040 .2084 .079 ”.009 .003 .010 064 .011 Credits .006 .001 .008 .020 .018 .086 .034 .086 .068 2088» .089 Current Load .006 .004 .008 .069 3' ’0. 00 .018 .008 .022 .042 1"1 .026 ,027 GPA .188 ".024 .027 .038 .015 .186 .023 .136 .014 .038 057 .059 CQT-V .101 .061 .010 .010 .011 .082 .019 .213 .018 .031 .112 .187 CQT-Q .090 .023 .038 .037 .176 .032 .146 .017 .051 .061 .187 CQT-T 108”".010 .028 .027 .154 .027 .164 .025 .077 $835 .2890 Extraversion . 5 .022 .031 .043 .052 .131. .049 .073 .10 .013 . 51 .050 Neuroticism .009 .013 .017 .082 .028 .143 .052 Q .082 .050 093 142 _ Eysenck Lie .016 .029 .039 .058' 3.'084"‘3.041 . 3 .009 .011 .028 “$101 3“ A-H Test Anxiety .055 .024 .007 .030 .033 .040 .046 .050 .057 .062 .041 ,128 Test Anxiety-l .002 .000 .001 .000 .000 .000 .000 .000 .000 .000 .000 .000 Test Anxiety-2 .011 .008 .017 .018 .017 .019 .029 .104 .053 .035 .006 .105 Test Anxiety-3 .006 .001 .002 .011 .013 .056 .020 .297 .031 .066 017 .092 Accuracy .010 .011 .007 .031 .025 .086 .041 .096 .043 .013 .094 .070 Success .012 .005 .007 .015 .046 .022 .040 .001 .012 .044 .166 Learning Set .049 .058 .036 NA .053 '15 .048 .081 .079 .087 NA .102 NA .049 NA Reason Enrolled .006 .003 .004 .004 .002 .075 .001 .080 .001 .021 .001 .089 Pre-attitude .001 .013 .018 .032 .029 .067 .052 .165 .059 .082 .084 .187 Post-attitude .006 .010 .009 .014 .029 .047 .037 .044 .044 .055 110 .116 Discussion Attitude .004 .007 .010 .017 .015 .124 .013 .115 .029 .037 .091 .085 50-50-1 (Asp.-Exp.-_1) .049 .009 .022 .036 .031 .077 .033 .146 .006 .020 .104 .070 50-50-2 (Asp.-Exp.-2) .089 .036 .020 .018 .025 .048 .025 .052 .019 .052 .048 .070 Pretest .049 .034 .034 .032 .036 .106 .031 .193 .079 .209 @053 £35) Midterm .126 .051 .056 ”.024 .017 .188 .019 .067 .016 .022 .057 . 35 Instructor Extraversion .003 .004 .005 .812 .029 .022 .026 .076 .063 .097 .136 .213 Instructor Neuroticism .010 .015 .013 .015 .019 .035 .027 .076 $3 .045 .041 .100 Instructor Marlowe-Crowne .006 .010 .006 .023 .030 .065 .040 .035 . 52 .063 .052 .012 Instructor Sanford-Cough .005 .003 .014 .056 .051 .064 .096 .057 .056 .035 .084 Instructor California-F .005 .006 .007 .007 .076 .06 .086 .043 .012 .020 .215 Instructor Risk .011 .019 .017 .052 .047 .068 .060 .054 .027 .041 .126 Instructor Course Load .014 .004 .010 .032 .029 .043 .192 .019 .095 .041 ,228 Instructor Teaching Experience .003 .007 .027 .026 .014 .033 .040 .089 .045 .041 .100 Instructor ED 200 .015 .025 .023 .010 .175 .019 .033 .039 .082 .097 .118 N 532 314 231 117 8 109 24 85 28 57 49 5 31 18 22 6 TSSi/TSST 1.000 .449 .295 .143 .003 .130 .020 .102 .029 .065 .040 .016 .016 .016 .014 .005 MEAN 24.615 23.207 22.528 21.838 18.750 22.064 23.500 21.658 22.929 21.035 21.490 18.000 20.581 23.056 23.818 19.667 Proportion of variation in that group explainable for each predictor. (BSS/TSS)i 4": =Split made on this variable. Q=Next best BSS/TSS. v =Final group. Q1=Sp1it attenmted but not made. NA=Sp1it not attempted, 397 TABLE E8 AID III: Proportion of Variation in Final Exam Total Score by Group within Branch Explainable for Each Final-Course Variable 398 TABLE E8 (cont'd ) Predictor Group Number 1 3 10 11* 14 15 18 19 26* 27 34* 35* 40* 41* Sex .000 .013 .010 .000 .003 .005 .017 .047 .015 Age .001 .008 .027 .021 .019 .081 .020 .017 .080 Major .027 .042 .070 .055 .087 .105 .063 .061 .060 Credits .019 .016 .006 .012 .030 .083 .005 .012 .085 Current Load .005 .006 .003 .005 .007 .057 .018 .032 .074 GPA .205’773042 .075 .070 .085 .032 .071 .071 .058 OQT-V .094 .135 .066 .061 .101 .049 .090 .031 .038 CQT-Q .088 @ % E080 I .086 .038 .090 .031 .040 OQT-T ' .121 .160 .1 .074 @ .054 .123‘ 5 ”.066 .049 Extraversion .021 .016 .014 .017 .052 .084 .095 45;» .053 NeuroticiSm .004 .010 .025 .040 .043 .094 .036 .123 .070 Eysenck Lie .012 .015 .018 .010 .063 .062 .095 .083 .117 A-H Test Anxiety .061 .042 .044 .034 .044 .089 .051 .084 .113 Test Anxiety-1 .000 .000 .000 .000 .001 .008 .004 .003 .019 Test Anxiety-2 .020 .019 .002 .007 .014 .049 .027 .016 .065 Test Anxiety-3 .015 .009 .017 .012 .027 .062 .012 .030 .045 Accuracy .008 .018 .014 .010 .012 .035 .018 .026 .034 Success .016 .009 .011 .018 .011 .055 .021 .021 .058 Learning Set .051 .030 .035 NA .057 NA .047 .090 NA .055 NA .041 NA .077 Reason Enrolled .004 .003 .020 .005 .098 .045 .1164 .007 Pre-attitude .002 .036 .038 .037 .065 .120 . 66 .018 .2100 Post-attitude .005 .047 .026 .061 .048 .146 .057 .062 .113 Discussion Attitude .007 .044 017 .022 .014 .088 .064 .021 .072 50-50-1 (Asp.-Exp.—1) .067 .31 .054 .019 .031 .061 .048 .078 .056 50-50-2 cAsp.-Exp.-2) .132 .118 .083 .073 .017 .175 .014 ’7?020 ‘9032 Pretest .069 .074 .075 gag» 058 .070 .070 .073 .029 Midterm .123 .115 .11 . 43 .041 .027 .042 Instructor Extraversion .005 .009 .005 .004 .027 .039 .015 .014 .010 Instructor Neuroticism .009 .053 .038 .035 .035 .084 .036 .023 .118 Instructor Marlowe-Crowne .004 .050 .030 .032 .018 .084 .020 .013 .074 Instructor Sanford-Gough .006 .022 .006 .008 .027 .040 .037 .018 .061 Instructor California-F .012 .017 .039 .031 .032 .033 .025 .005 .027 Instructor Risk .015 .026 .028 .035 .051 3 .021 .016 .025 Instructor Course Load .017 .026 .026 .047 .193' 3% 59 "b. 09 .005 @ Instructor Teaching Experience .005 .009 .011 .018 .040 .015 .032 .007 . 04 Instructor ED 200 .016 .039 .025 .044 .056 .026 .018 .009 .027 N 532 148 147 l 133 14 83 49 8 75 6 69 8 41 TSSi/TSST 1.000 .218 .149 .000 .124 .004 .074 .036 .005 .055 .004 .045 .003 .003 MEAN 53.838 58.959 59.252 16.000 58.624 65.214 57.301 60.816 63.875 56.600 51.333 57.058 56.625 61.634 Proportion of variation in that group explainable for each predictor(BSS/TSS)i "13 =Split made on this variable. Q=Next best BSS/TSS. a =Split attempted but not made. 1 =Final group. NA =Sp1it not attempted. 399 TABLE E8 (cont'd.) Predictor Group Number 1 2 5 7 46* 47 54* 55* Sex .000 .002 .014 .001 .014 Age .001 .001 .005 .030 .148 Major .027 .014 .015 .101 .360 I I Credits .019 .005 .008 .068 .148 Current Load .005 .005 .003 .091 .092 GPA .203 w (g?) .032 .096 CQT-V .094 .0 4 . 9 .210 .232 CQT-Q .088 .044 .021 .054 .148 CQT-T .121 .052 .048 .119 .l:1 Extraversion .012 .026 .014 Neuroticism .004 .006 .017 . 32 .115 Eysenck Lie .012 .026 .014 .019 .006 A-H Anxiety .061 .029 .028 .051 .148 Test Anxiety-1 .000 .001 .012 .000 .100 Test Anxiety-2 .020 .014 .020 .113 .136 Test Anxiety-3 .015 .009 .003 .020 .012 Accuracy .018 .011 .006 .166 .165 Success .016 .011 .006 .106 .120 Learning Set .051 .051 .037 .111 NA .115 NA NA Reason Enrolled .004 .004 .003 .018 .021 Pre-attitude .002 .005 .006 .079 .282 Post-attitude .005 .014 .030 .116 .027 Discussion Att. .007 .005 .007 .054 .128 50-50-1 (Asp.-Exp.).067 .011 .026 .028 .280 50-50-2 (Asp.-Exp.).132 .061 .023 .091 .092 Pretest .069 .044 .043 .027 .125 Midterm my; . 126 ’3 079" ‘3 007 . 086 Inst. Extravert .005 .015 .017 .105 .129 Inst. Neurotic .009 .008 .014 .049 .129 Inst. Marlowe-Cue. .004 .003 .006 .135 .132 Inst. Sanford-Ggh. .006 .004 .006 .123 .067 Inst. California-F .012 .008 .014 .082 .129 Inst. Risk .015 .017 .016 .074 .084 Inst. Load .017 .013 .018 .247 3 .2. 35 Inst. Teaching .005 .015 .017 .148 .034 Inst. ED 200 .016 .015 .015 .136 .044 N 532 .379 239 32 12 20 5 15 TSSi/TSST 1.000 .575 .279 .031 .005 .019 .003 .009 MEAN 53.838 51.839 53.569 57.563 60.833 55.600 60.800 53.867 Proportion of variation in that group explainable for each predictor a =Split made on this variable. =Next best BSS/TSS. =Final group. (BSS/TSS)1 4 =Split attempted but not made. NA =Sp1it not attempted. TABLE E8 (cont'd.) 400 Predictor Group Number 1 2 5 6 9 22 23* 32* 33* Sex .001 .002 .014 .019 .009 .002 .001 Age .001 .001 .005 .002 .016 I014 .011 Major .027 .014 .015 .036 .085 . 12 .003 Credits .019 .005 .008 .009 .027 .059 .020 Curr. Ld .005 .005 .003 .001 .012 .022 .029 GPA .203 .066 ”.020 .021 .006 CQT-V .094 . 4 .0 9 .029 .027 .025 .014 CQT-Q .088 .044 .021 .026 .018 .033 .035 CQT-T .121 .052 .048 .040 .029 .046 .044 Extravert .021 .008 .018 .024 .052 .041 .025 Neurotic .004 .006 .017 .012 .022 .009 .015 Eys. Lie .012 .026 .014 .015 .040 .035 .045 A-H Anx. .061 .029 .028 .030 .047 .043 .038 Anx.-l .000 .001 .012 .012 .052 .086 .109} Anx.-2 .020 .040 .020 .027 .067 .044 .059 Anx.-3 .015 .009 .003 .008 .006 .021 .031 Accuracy .008 .011 .006 .012 .042 .032 .045 Success .016 .011 .006 .008 .014 .015 .014 Learn. Set .051 .051 .037 .043 .083 .077 NA .057 NA Reason En. .004 .004 .003 .002 .012 .005 .011 Pre-att. .002 .005 .006 .008 .019 .014 .017 Post-att. .005 .014 .030 .036 .065 .035 .o 0 Disc. Att. .007 .005 .007 .007 .030 .062 50-50-1 .067 .011 .026 .019 .017 .023 .135 50-50-2 .132 .061 .023 .016 .074 .093 .099 Pretest .09 .044 .043—fl .093 .062 Midterm m. .120 ‘.07 .020 . 9 .027 .032 Inst. Ext. .oos .015 .017 .011 .042 .043 .039 Inst. Neu. .009 .008 .014 .011 .048 .030 .016 Inst. M-C .004 .003 .006 .011 .021 .005 .022 Inst. S-G .006 .004 .006 .011 .007 .021 .033 Inst. Cal-F .012 .008 .014 .013 .052 .098 .027 Inst. Risk .015 .017 .016 .034 .052 4g39 .007 Inst. Load .017 .013 .018 .023 .080 .103 3.6I8 ‘ Inst. Teach .005 .015 .017 .006 .042 .043 .039 Inst. ED200 .016 .015 .015 .012 .005 .005 .003 N 532 379 239 207 99 81 18 75 6 TSSi/TSST 1.000 .575 .279 .226 .095 .065 .021 .056 .003 MEAN 53.838 51.839 53.569 52.952 54.404 55.099 51.278 55.520 49.833 Prcporticn of variation in that group explainable for each predictor "1t=6plit made on this variable. =Next best BSS/TSS. * =I=Fina1 group. (BSS/TSS)i =Sp1it attempted but not made. NA =Split not attempted. 401 TABLE E8 (cont'd.) 402 TABLE E8 (cont'd ) Predictor Group Number 1 2 5 6 8 20 21 24* 25 42 43* 48* 49* 50 51* 52* 53* Sex .000 .002 .014 .019 .027 .015 .106 .059 .001 .001 003 Age .001 .001 .005 .002 .007 .013 .022 .017 .008 .017 _ .013 Major - .027 .014 .015 .036 .047 .060 .393’ .011 .138 .181 a 1' .240 Credits .019 .005 .008 .009 .022 .032 .032 .013 .048 .021 .069 Current Load .005 .005 .003 .001 .001 .002 .025 .016 .008 .024 .006 GPA .203 4'33 flip .066 ”.013 .029 .146 _031 .051 .086 089 CQT-V .094 .0'4 .0 9 .029 .034 .028 .035 .043 .064 .087 042 CQT-Q .088 .044 .021 .026 .053 .038 .114 .089 .047 .074 .161 CQT-T .121 .052 .048 .040 .074 3051 3133 .1209 .078 .087 .054 Extraversion .021 .008 .018 .024 .025 .026 .045 .052 .194"‘$037 ‘* .015 NeuroticiSm .004 .006 .017 .012 .031 .029 .049 .042 .077 .073 .103 Eysenck Lie .012 .026 .014 .015 .008 .009 .080 .053 .053 .165 .150 A-H Test Anxiety .061 .029 .028 .030 .030 .048 .080 .052 .058 .041 .082 Test Anxiety-l .000 .001 .012 .012 .000 .000 .000 .004 .000 .000 .000 Test Anxiety-2 .020 .014 .020 .027 .030 .025 .026 .022 .024 .007 .017 Test Anxiety-3 .015 .009 .003 .008 .016 .026 .026 .043 .072 .026 .069 Accuracy .008 .011 .006 .012 .010 .016 .071 .017 .062 .040 .072 Success .016 .011 .006 .008 .016 .017 .091 .030 .011 .033 .049 Learning Set .051 .051 .037 .043 .046 .038 .093 .101 .046 .044 NA NA NA 049 NA NA NA Reason Enrolled .004 .004 .003 .002 .020 .007 .040 .024 .022 .040 .013 Pre-attitude .002 .005 .006 .008 .064 .053 .026 .048 .075 .04; . Post-attitude .005 .014 .030 .036 .052 .056 .113 . 98 .120 .170 .316 d” 77‘ Discussion Attitude .007 .005 .007 .007 .031 .057 .032 .026 @ -126 50-50-1 (Asp.-Exp.-l) .067 .011 .026 .019 .027 .008 .043 .025 .030 .030 .109 50-50-2 (Asp.-Exp.-2) .132 .061 .023 .016 .038 .029 .166 .067 .107 .163 .075 PreteSt .069 .044 .043 $6420 .055 . .101 .056 .095 .098 .099 Midterm .12‘0 3'. 079 .020 .044 .017 .062 .040 .142 InStructor Extraversion 005 .015 .017 .011 .025 .028 .0 1 .034 .051 .046 .009 Instructor Neuroticism .009 .008 .014 .011 .011 .033 .061 .018 .020 .033 .009 Instructor Marlowe-Crowne .004 .003 .006 .011 .014 .023 .127 .026 .097 .170 000 Instructor Sanford-Gough .006 .004 .006 .011 .015 .022 .108 .024 .120 .235 .009 Instructor California-E .012 .008 .014 .013 .060 .052 .074 .040 .036 .055 .003 Instructor Risk .015 .017 .016 .034 .059 .146 .063 .017 .033 .009 Instructor Course Load .017 .013 .018 .823 .1 7 .028 .050 .025 .036 .071 .003 Instructor Teaching Experience .005 .015 .017 .006 .020 .09 .031 .005 .020 .020 .000 Instructor ED 200 .016 .016 .015 .012 .039 .043 .022 .009 .011 .033 .009 N 532 379 239 207 108 76 32 40 36 31 5 16 16 25 6 14 11 TSSi/TSST 1.000 .575 .279 .226 .116 .078 .030 .035 .036 .028 .004 .006 .012 .021 .001 .010 .004 MEAN 53.838 51.839 53.569 52.952 51.620 50 671 53.875 49.150 52.361 51.452 58.000 57.000 50.750 50.280 56.333 47 929 53 273 Proportion of variation in that group explainable for each predictor (BSS/13S)i "73 =Split made on this variable. =Next best BSS/TESS. Q, =Final group. NA =Split attempted but not made. =Split not attempted. 403 TABLE E8 (cont'd.) 404 TABLE E8 (cont'd.) Predictor Group Number 1 2 4 13 16 17 28 29* 36* 37* 38* 39* Sex .000 .002 .036 .023 .063 .032 .037 .005 .128 .061 Age .001 .001 .007 .002 .021 .003 .059 .039 .016 .029 Major .027 .014 .033 431 .134 .081 .141 .134 .140 Credits .019 .005 .026 . 1 .032 .132 .048 .043 .209 .2699 Current Load .005 .005 .013 .021 .041 .064 .021 .028 .209 .019 GPA 203’". .039 .059 .044 .065 .074 .123 .209 .139 CQT-V .094 .044 .058 .053 .091 .090 .082 .143 .189 .184 CQT-Q .088 .044 on. .020 .032 .071 .036 .047 .139 .057 CQT-T .121 .052 . o .023 .041 .111 .057 .111 .084 .145 Extraversion .021 .008 .036 .024 .119 .037 fig? .205 .114 .081 Neuroticism .004 .006 .042 .043 .31 .061 .021 .034 .159 ._079 Eysenck Lie .012 .026 .063 .090 .056 .323’ .066 .083 .065 Y ’.250 A-H Test Anxiety .061 .029 .035 .043 .049 .064 .054 .071 .3164 .080 Test Anxiety-l .000 .001 .004 .004 .000 .000 .000 .000 .000 .000 Test Anxiety-2 .020 .014 .017 .027 .033 .036 .043 .024 .054 .120 Test Anxiety-3 .015 .009 .021 .033 .063 .016 .060 .2164 .036 .112 Accuracy .008 .011 .047 .064 .043 .049 .037 .209 .050 Success .016 .011 .006 .032 . 5 .008 .043 .022 .090 .016 Learning Set .051 .051 .051 .042 .022 .071 .046 .027 .167 NA .209 Reason Enrolled .004 .004 .009 .054 .130 .064 .053 .141 .027 .100 Pre-attitude .002 .005 .033 .057 .056 .055 .071 .050 .112 .118 Post-attitude .005 .014 .017 .020 .029 .073 .057 .043 .083 .078 Discussion Attitude .007 .005 .011 .025 .085 .141 .027 .026 .195 191 50-50-1 (Asp.-Exp.-1) .067 .011 .012 .008 .001 .035 .002 .071 .039 .058 50-50-2 (Asp.-Exp.-2) .132 .061 .028 .075 .043 .109 .060 .045 .121 .123 Pretest .069 .044 .036 .072 .109 {:9 .150" 3.021 ".074 Midterm .120 .084 .11 .058 .134 .130 .119 .178 .031 Instructor Extraversion .005 .015 .008 .026 .079 .081 .087 .067 Q? .112 Instructor Neuroticism .009 .008 .007 .016 .063 .138 .069 .192 . 5 .165 Instructor Marlowe-Crowne .004 .003 .006 .034 .049 .040 .054 .023 .137 .053 Instructor Sanford-Cough .006 .004 .006 .024 .025 .084 .024 .079 .137 .045 Instructor California-F .012 .008 .007 .019 .057 .117 .054 .152 .109 .083 Instructor Risk .015 .017 .010 .022 .017 .074 .025 .152 .153 .046 Instructor Course Load .017 .013 .010 .015 .031 .035 .046 .152 .137 .049 Instructor Teaching Experience .005 .015 .016 .021 .076 .067 .034 .192 .068 .126 Instructor ED 200 .016 .015 .006 .019 .025 .095 .045 .081 .023 .208 N 532 379 137 99 51 48 50 l 27 23 17 31 TSSi/TSST 1.000 .575 .217 .130 .071 .044 .049 .000 .023 .018 .008 .022 MEAN 53.838 51.839 49.007 50.283 48.314 52.375 48.800 24.000 46.963 50.957 56.177 50.290 Prcportion of variation in that group explainable for each predictor (BSS/TSS)i ”'1! =Split made on this variable. =Next best BSS/TSS. =Final group. 0' NA =Split attempted but not made. =Split not attempted. 405 TABLE E8 (cont'd.) Predictor Group Number 1 2 4 12 30 31* 44* 45* Sex .000 .002 .036 .002 .003 .039 Age .001 .001 .007 .030 .031 .052 Major .027 .014 .033 .011 .117 .176 Credits .019 .005 .026 .051 .076 .099 Current Load .005 005 .013 .035 .106 .103 GPA .20 .039 .067 .076 .099 CQT-V .094 .044 .058 .164 .055 .219 CQT-Q .088 .044 45;” .113 .125 .115 CQT-T .121 .052 . co . 42 .139 .193 Extraversion .021 .008 .036 .136 .086 .066 Neuroticism .004 .006 .042 .121 .081 @ Eysenck Lie .012 .026 .063 .064 .038 .3524 A-H Anxiety .061 .029 .035 .124 .149 .109 Test Anxiety-1 .000 .001 .004 .000 .000 .000 Test Anxiety-2 .020 .014 .017 .149 .224 Test Anxiety-3 .015 .009 .021 .007 .093 .052 Accuracy .008 .011 .047 .044 .103 .102 Success .016 .011 .006 .025 .049 .003 Learning Set .051 .051 .051 .105 .088 NA .099 NA Reason Enrolled .004 .004 .009 .050 .142 .100 Pre-attitude .002 .005 .033 .079 .157 .140 Post-attitude .005 .014 .017 .088 .128 .182 Discussion Att. .007 .005 .011 .079 .099 .327 50-50-1(Asp.-Exp.) .067 .011 .012 .024 .092 .125 50-50-2(Asp.-Exp.) .132 .061 .028 .070 .025 .024 Pretest .069 .044 .036 .105 .155 .263 Midterm .12fl84 .132 .100 .209 Inst. Extravert .005 .015 .008 .121 125 .089 Inst. Neurotic .009 .008 .007 .399 $125 I .099 Inst. Marlowe-Cue. .004 .003 .006 .142 .115 .089 Inst. Sanford-Ggh. .006 .004 .006 .142 .130 .097 Inst. California-F .010 .008 .007 .215 .013 I099 Inst. Risk .015 .017 .010 .273 .2667 .099 Inst. Load .017 .013 .010 .090 .099 .090 Inst. Teaching .005 .015 .016 .044 .039 Inst. ED 200 .016 .015 .006 .064 .027 .039 N 532 379 137 38 36 2 20 16 TSSi/TSST 1.000 .575 .217 .065 .035 .004 .014 .012 MEAN 53.838 51.839 49.007 45.684 46.694 27.500 49.050 43.750 Proportion of variation in that group explainable for each predictor “=Split made on this variable. * =Final group. =Next best BSS/TSS. (BSS/TSS)1 =Split attempted but not made. NA =Split not attempted. 406 TABLE E9 AID III: Proportion of Variation in Final Exam Grade by Group _____n1thin_Bzangh_Expl§ip§ble for Each Final-Course Variable Predictor Group Number 1 3 12 13 20* 21* 24 25* 28* 29* Sex .001 .015 .004 .005 .004 .002 .009 Age .001 .011 .012 .050 .006 .038 .052 Major .023 .050 .113 .057 .094 .081 .108 Credits .014 .015 .035 .033 .017 .058 .040 Curr. Ld. .006 .005 .002 .030 .016 .043 .037 GPA .211 ?051 .020 .047 .051 .047 .026 CQT-V .085 .079 4:22 .002 .056 .047 .082 CQT-Q .096 .161 .120 .019 .053 059 CQT-T .128 .159 .17? .056 ‘.036 " . 50 Extraversion .018 . 19 .071 .030 .059 .029 .100 Neuroticism .005 .023 .042 .018 .051 .029 .079 Eys. Lie .013 .023 .023 .012 .031 .036 .043 A-H Anxiety .061 .054 .083 .043 .058 .019 .019 Anxiety-l .000 .000 .000 .006 .000 .003 .005 Anxiety-2 .017 .012 .007 .055 .096 .036 .036 Anxiety-3 .018 .019 .050 .077 .027 .045 .071 Accuracy .007 .014 .049 .010 .056 .005 .019 Success .018 .009 .042 .020 .051 .022 .034 Learning Set .050 .041 .072 .030 NA .1043 .058 NA NA .087 Reason Enrol .003 .002 .003 .001 .014 .008 .058 Pre-attitude .003 .045 .088 .022 .066 .066 .123 Post-attitude .008 .024 .111 .024 .063 .013 .041 Discussion At..009 .025 .032 .025 .033 .039 .031 50-50-1 .072 .099 .179 .067 .033 .029 .105 50-50-2 .137 .088 .097 .070 .058 .084 .038 Pretest .080 .063 .025 .049 .027 .075 .144} Midterm .089 .182 .039 ‘ "o. 70 .043 .047 Inst. Ext. .006 .009 .028 .067 .027 .068 .052 Inst. Neu. .008 .033 .051 Q9 .095 .074 Inst. M-C .004 .029 .039 .007 .041 .o 9 .036 Inst. S-G .012 .012 .029 .014 .042 .010 .041 Inst. Cal-F .017 .019 .113 .064 .086 .088 .063 Inst. Risk .017 .023 .119 .077 (QZE) .109”* ‘—tsr-~s019 Inst. Load .021 .029 .031 .044 .056 .030 .020 Inst. Teach .006 .008 .022 .027 .058 .055 .051 Inst. ED 200 .013 .029 .089 .024 .059 .065 .052 N 532 148 63 85 10 52 71 14 16 55 TSS1/TSST 1.000 .175 .075 .072 .013 .049 .059 .000 .012 .041 MEAN 3.270 4.074 3.698 4.353 2.900 3.846 4.225 5.000 3.813 4.345 Proportion of variation in that group explainable for each predictor "‘I.=Split made on this variable. (BSS/TSS)1 =Next best BSS/TSS. ’1'=Split attempted but not made. * =Final group. NA =Split not attempted. 407 TABLE E9 (cont'd.) 408 TABLE E9 (cont'd.) Predictor Group Number 1 2 5 6 8* 9 10 11* 18* 19 32 33* Agggx 41* Sex .001 .005 .008 .005 .001 .000 .000 .012 .005 .015 Age .001 .000 .001 .005 .004 005 .015 .013 .013 .006 Major .023 .016 .012 .033 .04 .012 . 11 .006 .104 .033 Credits .014 .006 .006 .006 .003 .006 .017 .018 .025 .010 Current Load .006 .005 .002 .001 .005 .007 .006 .007 .099 .020 GPA .211 .061 .049 .029 .022 .018 .050 .033 .044 .126 CQT-V .085 .042 .048 .015 .094 .029 .132 .130 (P CQT-Q .096 .043 .031 .I37 .024 .157 .029 3 3051 .31 . 2 CQT-T .128 .061 (fi‘) .049 .041 .110 .214 .078 Extraversion .018 .007 .17 .025 . 1 . 2 .047 .070 .091 .2081 NeuroticiSm .005 .015 .011 .019 .013 .092 .020 .156 5.049 .‘099 Eysenck Lie .013 .026 .052 .029 .020 .038 .025 .074 .256 .016 . A-H Test Anxiety .061 .027 .031 .034 .029 .098 .035 .078 .142 .044 Test Anxiety-l .000 .001 .003 .000 .000 .000 .000 .000 .000 .000 Test Anxiety-2 .017 .014 .012 .023 .022 .087 .019 .028 .008 .089 Test Anxiety-3 .018 .010 .004 .008 .006 .028 .029 .041 .153 .030 Accuracy .007 .005 .006 .014 .011 .026 .026 .027 .044 .019 Success .018 .013 .011 .009 .010 .013 .011 .013 .040 .014 Learning Set .050 .065 .071 .098 .021 .051 .035 NA .033 .044 .064 NA NA Reason Enrolled .003 .002 .006 .009 .009 .023 .022 .028 .044 .040 Pre-attitude .003 .006 .003 .010 .016 .036 .0503 .045 077 .064 Post-attitude .008 .021 .023 .021 .017 .016 .029 .050 .144 .024 Discussion Attitude .009 .007 .007 .015 .012 .007 .027 .099 .064 50-50-1 (Asp.-Exp.-l) .072 .012 .014 .007 .005 .006 . 2 .015 .044 .124 50-50-2 (Asp.-Exp.-2) .137 my .045 .016 .023 .051 .034 .124 .223 .064 Pretest .080 .053 .047 .030 .012 .021 .013 .013 _150 .043 Midterm .125—_‘2078 ”.013 .004 .056 .016 .018 .041 .015 Instructor Extraversion 006 .014 .017 .018 .016 .039 .014 .100 .256 .115 Instructor Neuroticism .008 .018 .025 .035 .022 .051 .005 .081 .172 Instructor Marlowe-Crowne .004 .005 .008 .008 .003 .013 .022 .081 _ 0 .042 Instructor Sanford-Gough .012 .007 .012 .008 .003 .014 .017 .051 .130 .040 Instructor California-F .017 .012 .013 .013 .008 .028 .007 .100 .262 .109 Instructor Risk .017 .017 .014 .013 .017 .052 .013 _094 .091 _179 Instructor-Course Load .021 .018 .018 .016 .008 .028 .009 .051 .091 .083 Instructor Teaching Experience .006 .014 .018 .018 .016 .039 .014 .041 .150 _051 Instructor ED 200 .013 .015 .015 .011 .015 .057 .010 Q? .144 .189 N 532 379 319 196 15 176 75 101 14 61 27 34 18 9 TSSi/TSST 1.000 .609 .436 .251 .025 .185 .078 .100 .017 .049 .021 .022 .012 .003 MEAN 3.270 3.034 3.182 2.990 2.133 3.085 3.267 2.950 2.643 3.410 3.704 3.176 3.444 4.222 Proportion of variation explainable in each group for each predictor (BSS/T5571 ‘r'fii=Sp1it made on this variable. <:::)=Next best BSS/TSS. ’1 =Sp1it attempted but not made, * =Final grOup. NA =Split not attempted. 409 TABLE E9 (cont'd.) 410 TABLE E9 (cont‘d.) Predictor Group Number 1 2 5 7 14 15 22* 23 26* 274 30 314 36* 37* Sex .001 .005 .008 .010 .000 .094 .002 .001 .219 Age .001 .000 .001 .011 .022 .040 .019 .010 .135 Major .023 .016 .012 .016 .057 .081 .417 a " Credits .014 .006 .006 .005 .030 .c 3 .u .045 .061 current Load .006 .005 .002 .005 .001 .024 .001 .010 .061 CPA .211‘ T061 .049 .065 .104 .120 .024 .048 .101 CQT-V .085 .042 .048 .091 .051 .081 .119 5 .Im8 .121 CQT-Q .096 .043 .031 .028 .021 .167 .025 .042 .175 CQT-T .128 .061 (%:3) .11- .031 . 89 .033 .058 .160 Extraversion .018 .007 .~17 .028 .005 .112 .018 .014 .155 Neuroticism .005 .015 .011 .029 .034 .116 .049 .054 .127 Eysenck Lie .013 026 .052 .024 .012 .123 .010 .015 .126 A-H Test Anxiety .061 027 .031 .041 .037 .042 .041 .034 Test Anxiety—l .000 .001 .003 .024 .031 .000 .045 .648 .000 Test Anxiety-2 .017 .014 .012 .024 .031 .056 .045 .048 .057 Test Anxiety-3 .018 .010 .004 .005 .047 .025 .045 .089 .126 Accuracy .007 .005 .006 .024 .042 .061 .056 .066 Success .018 .013 .011 .027 .041 .049 .056 .066 .037 Learning Set .050 065 .071 .079 .052 .123 NA .051 NA .1150a .140 NA NA NA Reason Enrolled .003 .002 .006 .036 .044 .008 .041 .012 .043 Pre-attitude .003 .006 .003 .009 .051 .114 .053 .026 .212 Post-attitude .008 .021 .023 .016 .033 .064 .009 .056 .048 / Discussion Attitude .009 .007 .007 .025 .031 .080 .019 .011 .137 50-50-1 (Asp.-Exp.-l) .072 .012 .014 .060 .017 .151 .016 .026 .179 50-50-2 (ASp.-Exp.-2) .137 '”‘, .045 .036 .041 .114 -020 .015 209 Pretest .080 . .047 p .069 .041. .20f .055 .065 “‘.053fi' Midterm ‘33» .120 7078 it p "_ .108 .059 .031 .155 t Instructor Extraversion .006 .014 .017 “ .636 .003 .031 .047 .033 ‘ Instructor NeurotiCiSm .008 .018 .025 013 .021 .078 .014 .037 .038 Instructor Marlowe—Crowne .004 .005 .008 015 .011 .067 .020 .021 .107 Instructor Sanford-Gough .012 .007 .012 .022 .011 .151 .020 .016 .164 Instructor California-F .017 .012 .013 020 .018 .041 .031 .032 .046 Instructor Risk .017 .017 .014 .018 .014 .083 .024 .036 .060 Instructor Course Load .021 .018 .018 027 .033 .166 .031 .045 .177 Instructor Teaching Experience .006 .014 .018 016 .028 .033 .018 .022 .009 Instructor ED 200 .013 .015 .015 .030 .012 .058 .034 .071 .101 N 532 379 319 123 82 41 5 77 11 66 25 16 9 16 TSSi/TSST 1.000 .609 .436 .151 .075 .059 .006 .061 .009 .045 .037 .010 .007 .014 MEAN 3 270 3.034 3.182 3.488 3.293 3.878 2.400 3.351 3.909 3.258 3.560 4.375 4.333 3.125 Proportion of variation in that group explainable for each predictor (BSS/TSS)i "“i=8p1it made on this variable. =Next best BSS/TSS. ’1 * =Final group. NA éSplit attempted but not made. =Sp1it not attempted. 411 TABLE E9 (cont'd.) Predictor Group Number 1 2 4 16 17 34* 35* 38* 39* Sex .001 .005 .048 .052 .043 Age .001 .000 .104 .41 .008 Major .023 .016 .116 .164 .053 Credits .014 .006 .072 .152 .075 Current Load .006 .005 .096 .050 .001 GPA .211 T061 .027 .058 .132 CQT-V .085 .042 .056 .217 .075 OQT-Q .096 .043 .044 .150 .061 CQT-T .128 .061 .059 197 .185 Extraversion .118 .007 .164—@125 Neuroticism .005 .015 .020 . 58 .107 Eysenck Lie .013 .026 .074 .037 .107 A-H.Anxiety .061 .027 .035 .100 Test Anxiety-1 .000 .001 .000 .000 . 00 Test Anxiety-2 .017 .014 .035 .037 .107 Test Anxiety-3 .018 .010 .033 .078 .032 Accuracy .007 .005 .035 .036 .032 Success .018 .013 .011 .134 .132 Learning Set .050 .065 .043 .065 .107 NA NA NA Reason Enrolled .003 .002 .004 .027 .017 Pre-attitude .003 .006 .056 .104 .062 Post-attitude .008 .021 .023 .050 .153 Discussion Attitude .009 .007 .036 .120 .038 50-50-1 QASp.-EXP.) .072 .012 .018 .064 .122 50-50-2- (ASp.-Exp.) .137 .027 .037 .069 Pretest .001 .53_@ .227 .206 Midterm .120 .074 .071 .24? 5 ‘ Instructor Ext. .006 .014 .027 .051 .149 Instructor Neu. .008 .018 .027 .050 .069 Instructor MPC .004 .005 .027 .030 .033 Instructor S-G .012 .007 .019 .030 .033 Instructor Cal-F .017 .012 .014 .011 .043 Instructor Risk .017 .017 .028 .022 .037 Instructor Load .021 .018 .025 .019 .037 Instructor Teach. .006 .014 .008 .052 .210 Instructor ED 200 .013 .015 .027 .037 .033 N 532 379 57 27 27 15 6 21 TSSi/TSST 1.000 .609 .087 .043 .025 .004 .021 .007 .012 MEAN 3.270 3.034 2.281 1.926 2.667 1.250 2.467 2.000 2.857 Prcporticn of variation in that group explainable for each predictor “=Split made on this variable. =Next best BSS/TSS. * =Final group. (BSS/TSS)1 Q. =Split attempted but not made. NA =Split not attempted. 412 TABLE E10 AID III: Proportion of Variation in Instructor Grade by Group within Each Branch Explainable for Each Final-Course Variable 413 TABLE E10 (cont'd.) Predictor GrouLNumber l 3 7 12 13 26* 27 32* 33 36* 37* 38* 39 48* 49* Sex .003 .010 .000 .009 .000 .003 .000 .001 .006 .013 Age .005 .003 .007 .009 .012 .018 .032 .041 .012 .032 Major .013 .010 .026 .044 .060 .062 .043 .034 .088 .066 Credits .024 .011 .017 .022 .025 .020 .044 .032 .040 .051 Current Load .007 .003 .002 .001 .011 .001 .004 .003 .011 .012 GPA .129"‘¥031 .058 .041 .029 .024 .018 .080 080 CQT-V . 0 .021 .019 .057 .024 .073 .098 .052 .048 .038 CQT-Q .024 .026 .022 .095 .025 5 i043 .041 .058 .021 035 CQT-T .019 .013 .037 .027 .035 .031 .041 .034 .039 .037 Extraversion .009 .021 .058 .046 .109 .048 .062 .086 .091 .091 Neuroticism .008 .016 .029 .071 .086 ‘ '?b41 .013 .014 . 3 Eysenck Lie .011 .023 .027 .037 .106 .047 .033 .031 .149 $028 ” A-H Test Anxiety .013 .015 .022 .040 .035 .078 .0960 .110 .111 Test Anxiety-l .002 .001 .000 .000 .000 .001 .001 .002 .000 .001 Test Anxiety-2 .006 .017 .008 .039 .023 .007 .016 .039 .023 .024 Test Anxiety-3 .010 .009 .009 .013 .011 .008 .011 .014 .035 .037 Accuracy .003 .003 .037 .054 .023 .056 .065 .069 .041 .036 Success .005 .005 .004 .010 .011 .009 .021 .027 .022 .012 Learning Set .006 .030 .024 .059 .087 .060 .090 Reason Enrolled .002 .006 .010 .016 .062 NA .020 NA .021 NA . 1 NA .054 .080 NA Pre-attitude .005 .011 .047 .051 .086 .073 .127 .032 .094 .084 Post-attitude .004 .013 .040 .078 .023 .054 -633 044 .038 .023 Discussion Attitude .006 .009 .038 .033 .206"' .031 .036 .036 ' ”‘77044 .077 50-50-1 (Asp.-Exp.-1) .052 .029 .026 .104 .020 .023 .034 50-50-2 (Asp.-Exp.-2) .088 .054 .022 .013 .065 .015 .020 .045 .119 .1239 Pretest .032 .019 .013 .023 .044 .030 .044 .067 .047 .039 Midterm .115 .043 .045 .054 .086 .056 .065 .069 .077 .075 Instructor Extraversion .015 .007 .027 .016 .010 .024 .023 .015 .033 .021 Instructor Neuroticism .037 .011 .011 .014 .032 .005 .011 .037 .024 009 Instructor Marlowe-Crowne .007 .016 .044 .014 .008 .024 .016 .037 .027 .021 Instructor Sanford-Gough .010 .016 .027 .023 .010 .036 .026 .032 .028 .025 Instructor California-F .024 .026 .036 .018 .041 .066 .050 .032 .024 Instructor Risk .008 .021 .037 .036 .015 .041 .066 .050 .027 .021 Instructor Load .007 .007 .034 {072 .030 .062 .049 .075 .049 .018 Instructor Teaching Experience .018 .032 .106 .020 .010 .014 .015 .010 .003 .002 Instructor Education 200 .012 .016 .026 .014 .005 .009 .015 .036 .028 .025 N 532 243 149 86 63 4 82 7 73 2 71 3 60 59 1 TSSi/TSST 1.000 .362 .186 .098 .068 .002 .086 .005 .069 .000 .060 .002 .052 .045 .000 MEAN 3.790 4.078 4.268 4.093 4.508 3.250 4.134 3.571 4.164 3.000 4.197 3.333 4.566 4.593 3.000 PrOportion of variation in each group explainable for each predictor (BSS/TSS)i fl=Split made on this variable. (:::;>=Next best ESS/TSS. =Final group. =Split attempted but not made. NA =Sp1it not attempted. 415 TABLE E10 (cont'd.) Predictor Grgup Number 1 3 6 18* 19 28* 29 42 43* 44* 45 54* 55 64* 65* Sex .003 .010 .015 .011 .044 .044 .056 .108 .001 Age .005 .003 .006 .020 .022 .006 .009 .046 .092 Major .013 .010 .018 .037 .055 .088 @ .124 .138 Credits .024 .011 .014 .042 .058 .042 .0 6 .081 .120 Current Load .007 .003 .003 .026 .022 .050 .060 .154 .193 GPA .129 T036 .058 .045 .036 .047 .107 .204 CQT-V . .021 .059 6? .055 .050 .076 .111 .132 CQT-Q .024 .026 .057 . 3 .079 .14 .113 .074 .101 CQT-T .019 .013 .048 .046 .034 .040 .078 .129 .079 Extraversion .009 .021 .026 .057 .045 .057 .072 .081 .120 Neuroticism .008 .016 .084 g 7006 .011 .018 .046 .27l S 7132 Eysenck Lie .011 .023 .053 .058 .038 .029 .033 .037 .096 A-H Test Anxiety .013 .015 .039 .028 .051 .050 .092 .054 .031 Test Anxiety-1 .002 .001 .002 .004 .009 .013 .010 .005 .000 Test Anxiety-2 .006 .017 .050 .016 .013 .023 .030 .016 Test Anxiety-3 .010 .009 .028 .054 .009 .013 .010 .020 .074 Accuracy .003 .003 .028 .016 .013 .004 .020 .023 .016 Success .005 .005 .043 .079 S T013 .012 .030 .010 Learning Set .006 .030 .057 NA .048 NA .031 .041 NA NA .072 NA .107 NA .086 Reason Enrolled .002 .006 .033 .028 .012 .002 .006 .023 .058 Pre-attitude .005 .011 .039 .016 .031 .036 .053 .074 .041 Post-attitude .004 .013 .031 .028 .059 .057 .076 .108 .138 Discussion Attitude .006 .009 .026 .031 .045 .057 .045 .081 .255Q 50-50-1 (Asp.-Exp.-1) .052 (ED .007 .016 .020 i023 .137 .006 .058 50-50-2 (Asp.—Exp.-2) .088 .054 .053 .048 .095 .098 .107 .074 .041 Pretest .032 .019 .028 .016 . .040 .064 .045 .121 .079 Midterm .115 T043 .033 .020 .008 .086 .060 .154 .193 Instructor Extraversion .015 .007 .006 .046 .044 .064 .103 .140 .041 Instructor Neuroticism .037 .011 .015 .015 .031 .040 .085 .123 .132 Instructor Marlowe-Crowne .007 .016 .016 .049 .051 .050 .060 .140 .132 Instructor Sanford-Cough .010 .016 .026 .049 .051 .050 .082 .092 Instructor California-F .024 .026 .004 .034 QED @ .147—_T—g, 56 .092 Instructor Risk .008 .021 .020 .030 .0 5 .057 .051 .074 .193 Instructor Course Load .007 .007 .011 .028 .059 .033 .046 .111 .101 Instructor Teaching Experience .018 .032 .002 .006 .040 .064 .078 .045 .086 Instructor ED 200 .012 .016 .018 .033 .030 .021 .051 .140 .101 N 532 243 90 13 77 16 59 55 4 11 44 15 29 7 22 TSSi/TSST 1.000 .362 .127 .026 .090 .019 .064 .055 .002 .003 .044 .010 .027 .005 .016 MEAN 3.790 4.078 3.756 4.231 3.675 4.000 3.576 3.527 4.250 3.091 3.636 3.333 3.793 3.286 3.955 Proportion of variation in each group explainable for each predictor (BSS/TSS)i /_1I=Split made on this variable. <::;>=Next best BSS/TSS. Q! =Split attempted but not made. =Final group. NA =Split not attempted, 414 TABLE E10 (cont'd.) 416 TABLE E10 (cont'd.) 417 TABLE E10 (cont'd.) Predictor Group Number 1 2 5 9 20* 21 24 25* 30 31* 46* 47 52* 53* Sex .003 .001 .000 .000 .015 .010 .000 .002 .038 Age .005 .014 .019 .013 .009 .008 .033 .051 .058 Major .013 .020 .032 .038 .040 .050 .055 .055 .012 Credits .024 .016 .024 .001 .013 .012 .038 .055 .050 Current Load .007 .008 .009 .021 .030 .036 .001 .002 .001 GPA @ .052 ”.013 .033 .029 .022 .062 .1754 CQT-V .020 .003 .010 .105 .030 .036 .046 .0 8 .082 CQT-Q .024 .012 .012 .034 .040 @ .055 .118 CQT-T .019 .006 .013 .032 .034 .036 .062 .055 .106 Extraversion .009 .030 .035 .032 .040 .050 .074 .118 .093 Neuroticism .008 .011 .013 .054 .046 .017 .038 .043 Eysenck Lie .011 .007 .013 .027 .039 .055 .080 .16 .062 A-H Test Anxiety .013 .011 .015 .105 .123 $072 8 .046 .052 .043 Test Anxiety—1 .002 .002 .004 .000 .000 .000 .000 .000 .000 Test Anxiety-2 .006 .001 .002 .002 .005 .008 .033 .054 .038 Test Anxiety-3 .010 .004 .016 .019 .048 .050 .151 S 5050 .092 Accuracy .003 .010 .018 .026 .040 .050 .055 .055 .036 Success .005 .009 .014 .004 .009 .012 .008 .010 .004 Learning Set .006 .017 .021 .035 NA .031 .031 NA .043 NA NA .038 .014 NA Reason Enrolled .002 .004 .005 .024 .016 .009 .009 .021 .006 Pre—attitude .005 .013 .017 .032 .048 .035 .046 .058 .082 Post-attitude .004 .007 .013 .032 .034 .040 .043 .081 .082 Discussion Attitude .006 .018 .040 .067 .054 .118 3048 I .055 .064 50-50-1 (ASp.-Exp.-1) .052 .015 .026 .009 .008 .020 .023 .015 .024 50-50-2 (Asp.-Exp.-2) .088 .018 .029 .045 .034 .036 .046 .048 .081 Pretest .032 .013 .016 .030 .040 .050 .055 .058 Midterm .1157—TTO37 .055 .015 .007 .002 .003 .019 . 18 Instructor Extraversion .015 .039 .033 .085 .060 .031 .036 .033 .032 InStructor NeuroticiSm .037 .038 .036 .013 .039 .022 .020 .013 _018 Instructor Marlowe-Crowne .007 .021 .022 .040 .011 .002 .010 .024 .047 Instructor Sanford-Cough .010 .012 .004 .027 .019 .033 .033 .027 .040 Instructor California-F .024 .027 .023 .11 .059 .036 .023 .015 .024 Instructor Risk .008 .005 .004 .008 .014 .023 .041 .056 Instructor Course Load .007 .005 .027 .111 .017 .012 .014 .021 .018 Instructor Teaching Experience .018 (1:2) .037 .013 .023 .031 .036 .033 .020 Instructor ED 200 .012 .011 €43) .053 .048 .050 .046 .030 .038 N 532 285 193 70 5 65 64 1 50 14 2 48 41 7 TSSi/TSST 1.000 .515 .334 .120 .006 .100 .088 .000 .054 .002 .000 .046 .034 .004 MEAN 3.790 3.551 3.668 3.900 3.000 3.969 4.000 2.000 3.880 4.429 5.000 3.833 3.927 3.286 Proportion of variation in each group explainable for each predictor (BSS/TSS)i "EN =Split made on this variable. Q=Next best BSS/TSS. Q =Split attempted but not made. ' =Final group. NA =Split not attempted. 418 TABLE E10 (cont'd.) 419 TABLE E10 (cont'd.) Predictor Gr0up Number 1 2 5 8 11 16 17 22 23 50 51* 56* 57* 68* 69* 70* 71* Sex .003 .001 .000 000 .011 .014 .027 .010 .000 .010 Age .005 .014 .019 .031 .029 .014 .115 .051 .023 .051 Maior .013 .020 .032 .034 .023 .034 .186 .051 .050 .321 Credits .024 .016 .024 .084 “0. 12 .028 .2727 .071 .028 .130 ”W“ 7' (urrenr Load .007 .008 .009 .014 .020 .041 .000 .025 .064 .035 CPA .052 2013 .034 .027 .059 .081 .050 .071 .089 CQT-V . 0 .003 .010 .021 .030 .067 .077 .077 .077 .289 CQT-Q .024 .024 .012 .024 .056 .085 .077 .107 .070 .089 CQT-T .019 .006 .013 .027 ‘33) -090 .077 .051 .132 .211 Extraversion .009 .030 .035 ‘13? .577 .067 .117 .064 .114 Neuroticism .008 .011 .013 .015 .052 .090 .240 . 67 .100 .109 Eysenck Lie .011 .007 .013 .005 .016 .020 I072 :19; I{083 .288 A-H Test Anxiety .013 .011 .015 .059 .067 .101 .088 .29 . .186 ’W 7“ Test Anxiety-1 .002 .002 .004 .005 .000 .000 .000 .000 .000 .000 Test Anxiety-2 .006 .001 .002 .009 .013 .005 .091 .071 .030 .036 Test Anxiety-3 .010 .004 .016 .022 .017 .034 .127 .051 .028 .187 Accuracy .003 .010 .018 .011 .019 .013 .186 .169 .033 .089 Success .005 .009 .014 .035 .041 .045 .057 .165 .009 .255 Learning Set .006 .017 .021 .040 .087 .077 .091 .132 .183 NA NA NA NA NA NA . NA Reason Enrolled .002 .004 .005 .028 .036 .012 .061 .037 .091 .067 Pre-attitude .005 .013 .017 .048 .040 .026 .061 .195 .058 .043 Post-attitude .004 .007 .013 .038 .027 .037 .108 .049 .070 .267 Discussion Attitude .006 .018 .040 .038 .059 .044 .077 .088 .040 50-50-1 (Asp.-Exp.-1) .052 .015 .026 .038 .041 .060 .026 .113 .091 .I 5 50-50-2 (Asp.-Exp.-2) .088 .018 .029 .014 .011 .006 .077 .092 .040 .044 Pretest .032 .013 .016 .035 .063 .045 .240 .057 .064 .156 Midterm .115 *037 .055 T055 .064 .087 .107 .070 .088 Instructor Extraversion .015 .039 .033 .044 .050 010 .027 .077 .132 .035 Instructor Neuroticism .037 .038 .036 .076 .11 .032 .013 .077 .132 .187 Instructor Marlowe-Crowne .007 .021 .022 .043 .055 .027 .027 .041 .037 .106 Instructor Sanford-Cough .010 .012 .004 .024 .014 .053 .027 .028 .213 .358 ‘1» "‘_1U Instructor California-F .024 .027 .023 .039 .030 .019 .027 .077 4219 151 . Instructor Risk .008 .005 .004 .025 .033 .033 .013 .077 . ' .071 Instructor Course Load .007 .005 .027 .050 .025 .048 .053 .051 .069 .255 Instructor Teaching Experience .018 .037 .067 .063 .011 .013 .049 .034 .077 Instructor ED 200 .012 .011 .039 .033 .026 .007 .015 .172 .106 N 532 285 193 123 91 66 25 29 37 23 10 13 TSSi/TSST 1.000 .515 .334 .196 .139 .102 .022 .041 .050 .020 .003 .010 MEAN 3.790 3.551 3.668 3.537 3.660 3.803 3.280 3.552 4.000 3.739 4.429 3.929 3.200 3.471 2.875 4.100 3.462 Proportion of variation in that group explainable for each predictor (BSS/TSS)i "‘S =Split made on this variable. =Next best BSS/TSS. CL =Sp1it attempted but not made. * =Final group. NA =Sp1it not attempted. 420 TABLE E10 (cont'd.) Predictor Group Number 1 2 5 8 10 58* 59* Sex .003 .001 .000 .000 .029 Age .005 .014 .019 .031 .016 Major .013 .020 .032 .034 .078 Credits .024 .016 .024 .080070 Current Load .007 .008 .009 .014 .009 GPA .052’3013 .034 .113 CQT-V . .003 .010 .021 .032 CQT-Q .024 .012 .012 .024 .112 CQT-T .019 .006 .013 .027 .053 Extraversion .009 .030 .035 .077 Neuroticism .008 .011 .013 .015 Eysenck Lie .011 .007 .013 .005 .035 A-H Test Anxiety .013 .011 .015 .059 .112 Test Anxiety-1 .002 .002 .004 .005 .002 Test Anxiety-2 .006 .001 .002 .009 .122 Test Anxiety-3 .010 .004 .016 .022 .078 Accuracy .003 .010 .018 .011 .067 Success .005 .009 .014 .035 .114 Learning Set .006 .017 .021 .040 .101 NA NA Reason Enrolled .002 .004 .005 .028 .036 Pre-attitude .005 .013 .017 .048 .108 Post-attitude .004 .007 .013 .038 .099 Discussion Attitude .006 .018 .040 .038 .190 50-50-1 CA8p.-Exp.-l) .052 .015 .026 .038 .008 50-50-2 0A3p.-Exp.-2) .088 .018 .029 .014 .017 Pretest .032 .013 .016 .035 .078 Midterm .11f"3o37 .osfioss .068 Instructor Extraversion .015 .039 .033 .044 .060 Instructor Neuroticism .037 .038 .036 .076 .066 Instructor Marlowe-Crowne .007 .021 .022 .043 .131 Instructor Sanford-Cough .010 .012 .004 .024 .146 Instructor California-F .024 .027 .023 .039 .109 Instructor Risk .008 .005 .004 .025 .132 Instructor Load .007 .005 .027 .050 .234 Instructor Teaching Exp. .018 .037 .067 .245 ' ' Instructor ED 200 .012 . 11 ® .039 .187 N 532 285 193 123 32 16 16 TSSi/TSST 1.000 .515 .334 .626 .040 .018 .013 MEAN 3.790 3.551 3,668 3.537 3.188 2.875 3.500 Proportion of variation in each group explainable for each predictor “=Sp1it made on this variable. (BSS/TSS)i ®=Next best BSS/TSS, 4 -Split attempted but not made. =Final group. NA =Split not attempted. 421 TABLE E10 (cont'd.) 422 TABLE E10 (cont'd,) Predictor Group Number 1 2 4 14 15 34 35 40* 41 60* 61* 62* 63* 66* 67* Sex .003 .001 .022 .010 .035 .016 .009 .051 .001 .003 Age .005 .014 .007 .033 .045 .145 .028 .073 .008 .193 Major .013 .020 .073 .089 .209 .189 (£13 .281 .2240 255%} credits .024 .016 .019 .019 .088 .096 .076 .273 .122 .041 Current Load .007 .008 .010 .033 .003 .033 .037 .175 .059 .101 GPA .052 ”.009 .061 .105 .161 .119 .223 .191 .138 CQT-V O 0 .003 .020 .051 .092 .044 .064 .272 .059 .073 CQT-Q .024 .012 .046 .043 .102 .067 .066 .206 .059 086 CQT-T .019 .006 .022 .039 .185 .073 .123 .146 .119 .227 Extraversion . 009 . 030 .073 . 148 .092 3. 068 ”.069 . 280 .149 . 120 Neuroticism .008 .011 .035 .062 .104 .062 .273 .078 .231 Eysenck Lie .011 .007 .009 .027 .109 .161 .080 .164 084 .138 A—H Test Anxiety .013 .011 .059 .022 @ .064 .044 .57 .043 .088 Test Anxiety-l .002 .002 .000 .000 .000 .000 .000 .000 .000 .000 Test Anxiety-2 .006 .001 .026 .036 .041 .146 .153 .023 .180 .192 Test Anxiety—3 .010 .004 .028 .027 .030 .145 157 .031 .178 .193 Accuracy .003 .010 gnu) .040 300* .077 .182 3 7044 @ .107 Success .005 .009 .011 .033 .060 .073 .045, 054 .059___= .120 Learning Set .006 .017 .054 .089 .220 .122 .251 NA .273 F059 NA NA NA NA .083 Reason Enrolled .002 .004 .019 .010 .032 .020 .048 .036 .037 .003 Pre-attitude .005 .013 .013 .038 .052 .035 .119 .097 .191 .092 Post—attitude .004 .007 .032 .065 .032 .090 .061 .051 .112 .193 Discussion Attitude .006 .018 .029 .033 .022 .043 .037 .109 .059 .084 50—50-1 (Asp.—Exp.-1) .052 .015 .026 .030 .085 .084 .066 .073 .053 .101 50-50-2 (Asp,-Exp.—2) .088 .018 .022 .036 .035 .122 .037 .011 .059 .132 u Pretest .032 .013 . .068 .057 .092 .096 .156 .273 .178 @ 1‘. Midterm .115 3.037 .027 .074 .077 .199 .068 .040 _ 31 Instructor Extraversion .015 .039 .052 .018 .032 .112 .009 003 .018 .075 Instructor NeuroticiSm .037 .038 cam .032 .137 .029 .127 .059 .088 Instructor Marlowe-Crowne .007 .021 .10 .033 .032 .162 L001 .127 .016 .088 Instructor Sanford-Cough .010 .012 .089 .094 .032 .233 .048 .164 .037 J €138 Instructor California-F .024 .027 .046 .032 .076 .200 .011 .103 .040 .079 Instructor Risk .008 .005 .023 .013 .038 .103 .005 .073 .025 .079 Instructor Course Load .007 .005 .021 .024 .032 .079 .036 .164 .040 _088 Instructor Teaching Experience .018 ® .046 .035 .048 .079 .004 .273 .076 .088 Instructor ED 200 .012 .011 .005 .102 .048 .166 .029 .273 .027 .138 N 532 285 91 61 3O 29 29 9 21 28 1 l9 2 7 22 TSSi/TSST 1.000 .515 .153 .072 .066 .028 .034 .013 .033 .025 .000 .014 .000 .005 .016 MEAN 3.790 3.511 3.297 3.131 3.633 2.897 3.379 4.333 3.333 3.321 5.000 3.158 5.000 2.429 3,045 Proportion of variation in each group explainable for each predictor (BSS/TSS)i "1!==Split made on this variable. (L =Split attempted but not made. Q=Next best BSS/TSS, NA =Split not attempted. '=Mhfl.gmm. I [I\ [Dir stillilrlll'lllll I . n 423 TABLE Ell AID III: Proportion of Variation in Final Course Grade by Group within Branch Explainable for Each Final-Course Variable 424 TABLE E11 (cont'd.) Predictor Group Number 1 3 7 10 11 24* 25 26 27* 30* 31* 38* 39* Sex .004 .000 .021 .026 .027 .002 .052 .001 .4010 .098 Age .002 .015 .017 .013 .025 .118 .055 .452’ .018 .009 " * Major .040 .042 .030 .075 .074 .102 .203 .347 .036 .303 Credits .036 .027 .011 .006 .025 .042 .013 (3:69 .021 .295 Current Load .007 .003 .011 .024 016 127 .030 .094 .014 .024 GPA 225—"3101 .17 .043 .015 .162 .085 .014 .295 CQT-V .088 .081 .069 .075 .074 .035 .125 .136 .027 .159 CQT-Q .073 .067 .061 .072 .048 .102 .055 .139 .023 .063 CQT-T .103 .082 .088 .088 .063 .135 .085 .179 .053 .074 Extraversion .025 .013 .038 .049 .142 .070 .055 .276 .071 .213 NeuroticiSm .003 .013 .014 .021 .022 .132 .066 .136 .042 .089 Eysenck Lie .012 .007 .010 .028 .021 .085 .024 .085 .052 .170 A-H Test Anxiety .052 .045 .035 .004 .043 .095 .098 .169 .083 .3573 Test Anxiety-1 .001 .006 .010 .036 .008 g .013 .000 .007 .004 Test Anxiety-2 .019 .030 .015 GED .045 . '- .030 .029 .029 .141 Test Anxiety-3 .023 .019 .010 .022 .019 .028 .042 .085 .072 .170 Accuracy .009 .004 .013 .076 .034 .050 .098 .273 .007 .141 Success .021 .006 .017 .031 .091 .060 .060 .237 .007 .191 Learning Set .041 .040 .038 .061 .045 .176 .093 .179 .042 NA .194 NA NA Reason Enrolled .004 .013 .017 .049 .002 .058 .068 061 021 .174 Pre-attitude .005 .014 .021 .013 .046 .048 .21 .063 .187 Post-attitude .010 .023 .021 .048 .095 .058 .128 .085 . 91 .165 Discussion Attitude .009 .012 .008 .048 .025 I023 I039 I085 .048 .051 50-50-1 (Asp,-Exp.-1) .085 .116 .095 .098 .18 .023 .045 1276 .020 (g;§) 50-50-2 (Asp.-Exp.-2) .192 .130 CIZE> .078 .157 .116 .096 .217 .089 . 4 Pretest .082 .062 .055 010 .042 .089 .284. .072 .171 Midterm 33% .14 .033 .018 .010 .040 .029 .056 .089 Instructor Extraversion .011 .004 .019 .001 .027 .045 .006 .085 .028 .063 Instructor Neuroticism .020 .004 .029 .012 .131 .049 .041 .226 .1414 .029 Instructor Marlowe-Crowne .007 .007 .029 .012 .133 .035 .019 .109 .111 .035 Instructor Sanford-Gough .097 .010 .006 .010 .014 .015 .019 .103 .014 .063 Instructor California-F .219 .040 .035 .020 .088 .095 .104 .307 .071 .129 Instructor Risk .219 .040 .026 .017 .048 .162 .059 .226 .071 .042 Instructor Course Load .239 .040 .017 .013 .044 .052 .028 .237 .071 .042 Instructor Teaching Experience .109 .005 .018 .025 .085‘ .058 .058 .139 .063 .028 Instructor Education 200 .186 .006 .032 .012 .116 .075 .056 .168 .090 .014 N 532 242 149 81 68 38 43 21 47 19 22 10 ll TSSi/TssT 1.000 .282 .127 .058 .051 .014 .034 .018 .024 .012 .013 .003 .007 MEAN 3.633 4.149 4.389 4.198 4.618 3.976 4.395 4.286 4.766 4.632 4.136 3.900 4.636 Preportion of variation in each group explainable for each predictor (BSS/TSS)1 "‘I=Split made on this variable. =Split attempted but not made. @Next best BSS/TSS. NA =Split not attempted, Final group. 425 TABLE E11 (cont'd.) Predictor Group Number 1 3 6 18* 19* Sex .004 .000 .033 .0908'.001 Age .002 .015 .042 .012 .036 Major .040 .042 .029 4339 .076 Credits .036 .027 .058 .059 .097 Current Load .007 .003 .002 .001 .063 GPA .229’?080 .070 .184 CQT-V .088 .081 .019 .055 .143 CQT-Q .073 .067 .024 .039 .184 CQT-T .103 .082 .035 .038 .143 Extraversion .025 .013 .053 .045 .143 Neuroticism .003 .013 .019 .029 .076 Eysenck Lie .012 .007 .040 .039 .4129 A-H Test Anxiety .052 .045 .013 .018 .076 Test Anxiety-1 .001 .006 .019 .019 .000 Test Anxiety-2 .019 .030 .035 .049 .123 Test Anxiety-3 .023 .019 .071 .010 Accuracy .009 .004 .036 .038 .0 8 Success .021 .006 .023 .029 .241 Learning Set .041 .040 .048 .038 .259 Reason Enrolled .004 .013 .044 .020 .075 Pre-attitude .005 .014 .018 .028 .189 Post-attitude .010 .023 .008 .038 .107 Discussion Attitude .009 .012 .039 .085 .076 50-50-1 (ASp.-Exp.-l) .085 .016 .019 .019 .023 50-50-2 (ASp.-Exp.-2) .192 .130 (8713 .035 .076 Pretest .082 062 . 7 53 .184 Midterm .335’2229 .255 ”.004 ”.143 Instructor Extraversion .011 .004 .010 .028 .286 Instructor Neuroticism .020 .004 .011 .031 .176 Instructor Marlowe-Crowne .007 .007 .026 .018 .184 Instructor Sanford-Gough .097 .010 .022 .011 .143 Instructor California-F .219 .040 .037 .045 .139 Instructor Risk .219 .040 .037 .059 .259 Instructor Course Load .239 .040 .037 .024 .418 Instructor Teaching Experience .109 .005 .010 .036 .021 Instructor ED 200 .186 .006 .018 .012 .076 N 532 242 89 69 20 TSSi/TSST 1.000 .282 .086 .052 .012 MEAN 3.633 4.149 3.753 3.594 4.300 Proportion of variation in each group explainable for each predictor v‘fiI=Sp1it made on this variable. Q=Next best BSS/TSS. ' =Final group. (BSS/T35)i ¢=Sp1it attempted but not made. 426 TABLE Ell (cont'd.) 427 TABLE E11 (cont'd,) Predictor Group Number 1 2 S 9 14 15* 20 21 28* 29* 34* 35* Sex .004 .004 .007 .012 .007 .012 .196 .006 _003 Age .002 .005 .008 .010 032 .012 @ .016 ,004 Major .040 .032 .056 .14 .021 .053 .052 $2 .179 Credits .036 .013 .021 .005 .017 .038 .027 09 .045 Current Load .007 .008 .008 .008 .014 .004 .027 .060 .004 GPA <§:§> .072 .059 C:::> .116 .4635 $069 T073 .049 CQT-V 88 .046 .051 .072 .055 .290 .096 .060 .202 CQT-Q .073 .041 .041 .056 .064 .073 .096 .096 _070 CQT-T .103 .051 .049 .056 .290 .096 .093 .061 Extraversion .025 .020 .023 .070 .058 .172 .057 .072 .061 NeuroticiSm .003 .009 .022 .003 .081 .078 .064 .26f '.219: ' Eysenck Lie .012 .031 .023 .049 .025 £39 .067 .036 .058 A-H Test Anxiety .052 .039 .045 .046 .081 .093 .122 .136 .196 Test Anxiety-1 .001 .000 .001 .000 .000 .000 .000 .000 .000 Test Anxiety-2 .019 .025 .015 .019 .023 .031 .112 .020 .003 Test Anxiety-3 .023 .028 .007 .004 .040 .172 .016 .142 .202 Accuracy .009 .004 .033 .004 .013 .012 .009 .060 .148 Success .021 .015 .008 .008 .007 .041 .010 ,011 Learning Set .041 .030 .017 .025 .018 .290 .139 .050 NA NA .061 NA Reason Enrolled .004 .002 .017 .013 .012 .073 .096 .093 .135 Pre-attitude .005 .014 .013 .053 .046 .172 .205 .036 .047 Post-attitude .010 .011 .012 .016 .027 .241 .069 .067 .061 Discussion Attitude .009 .015 .015 .019 .028 .143 .067 .033 .172 50-50-1 (ASp.-Exp.-1) .085 .036 .003 .013 .005 .065 .079 .012 .061 50-50-2 (ASp.-Exp.-2) .192 0130 .013 .020 .014 .189 .020 .093 .028 Pretest .082 .043 .048 .050 .215 .037 .061 .196 Midterm .335 a3172.089 I.003 .009 .001 .052 .004 .009 Instructor Extraversion .011 .010 .006 .026 .054 .031 .138 .122 .091 Instructor NeuroticiSm .020 .012 .005 .018 .075 .172 .138 .068 .135 Instructor Marlowe-Crowne .007 .008 .010 .042 .044 .065 .083 .072 ,055 Instructor Sanford-Cough .097 .006 .004 .043 .038 .143 .083 .024 .061 Instructor California-F .219 .006 .011 .021 .046 .055 .175 .042 .095 Instructor Risk .219 .021 .011 .021 .014 .030 .143 .039 .137 Instructor Load .239 .017 .012 .028 .031 .132 .143 .039 Instructor Teaching Experience .109 .018 .006 .007 .046 .172 .027 061 , 61 Instructor Education 200 .186 .008 .008 .011 .044 .151 .26 .072 _055 N 532 286 226 102 79 23 29 50 17 12 36 14 TSSi/TSST 1.000 .378 .244 .104 .077 .013 .036 .032 .015 .011 .014 .010 MEAN 3.633 3.200 3.345 3.459 3.671 3.130 3.414 3.820 3.706 3.000 3.972 3.429 Preportion of variation in each group explainable for each predictor (BSS/T8831 "‘8=Split made on this variable. =Sp1it attempted but not made. Q—Next best BSS/TSS. NA =Sp1it not attempted. 3 =Final group. 428 TABLE E11 (cont'd.) Predictor Group Number 1 2 5 8 12 13* 16 17* 36* 37* Sex .004 .004 .007 .003 .005 .039 .012 .024 Age .002 .005 .008 .002 .006 .093 .019 .107 Major .040 .032 .056 .019 .037 116 .048 .099 Credits .036 .013 .021 .083 '.001 ’.012 .004 Curr. Ld. .o 7 .008 .008 .o37 .062 .111 .039 .069 GPA (3:3) .072 .059 .055 .123 .131 CQT-V . .046 .051 .063 .052 .035 . o .024 CQT-Q .073 .041 .041 .060 .057 .093 .046 .113 CQT-T .103 .051 .049 .050 .050 .037 .035 .081 Extraversion .025 .020 .023 .020 .022 .020 .037 .034 Neuroticism .003 .009 .022 .017 .020 .129 .025 .091 Eys. Lie .012 .031 .023 .024 .034 69% A-H Anxiety .052 .039 .045 .051 .056 .23 .038 .004 Anxiety-1 .001 .000 .001 .001 .001 .001 .000 .070 Anxiety-2 .019 .025 .015 .008 .009 .066 .013 .088 Anxiety-3 .023 .028 .007 .030 .017 .010 .050 .063 Accuracy .009 .004 .033 .001 .006 .055 .003 .023 Success .021 .015 .019 .023 .139 .026 .171 Learning Set .041 .030 . .012 .016 NA .133 .037 .184 NA Reason Enrol .004 .002 .017 .017 .040 .117 .031 .115 Pre-attitude .005 .014 .013 .018 .020 .046 .023 .094 Post-attitude .010 .011 .012 .028 .011 .129 .026 .070 Disc. Att. .009 .015 .015 61% .085 .021 .002 50-50-1 .085 .036 .003 .11 .018 .066 .030 .008 50-50-2 .192 ms .013 .038 .026 .093 .021 .107 Pretest .082 .043 .039 .021 .014 .093 .045 .107 Midterm .335"?172'3708§"8027 .027 .034 .023 .023 Inst. Ext. .011 .010 .006 .029 .030 .105 .025 .107 Inst. Neu. .020 .012 .005 .004 .010 .094 .018 .107 Inst. M-C .007 .008 .010 .002 .008 .170 .006 Inst. S-G .097 .006 .004 .009 .018 .013 . G Inst. Cal-F .219 .006 .011 .007 .017 .063 .084} .036 Inst. Risk .219 .021 .011 .014 .012 .095 .021 .121 Inst. Load .239 .017 .012 .014 .016 .039 .052 .113 Inst. Teach. .109 .018 .006 .027 .030 .035 .025 .030 Inst. ED 200 .186 .008 .008 .006 .006 .042 .012 .041 N 532 286 226 124 122 1 36 86 34 2 TSSi/TSST 1.000 .378 .244 .118 .103 .000 .028 .066 .021 .000 MEAN 3.633 3.200 3.345 3.177 3.189 1.000 2.944 3.291 2.882 4.000 PrOportion of variation in that group explainable foTLeach predictor -Split made on this variable. @=Next best BSS/TSS . =Final group. NA (BSS/TSS)1 =Split attempted but not made. =Split not attempted. 429 TABLE E11 (cont'd.) Predictor GroupiNumber 1 2 4 22 23* 32* 33* Sex .004 .004 .031 .038 .009 Age .002 .005 .151 .141 .112 Major .040 .032 .073 .187 .2800 Credits .036 .014 .014 .118 .026 Current Load .007 .008 .028 .001 .013 GPA GT) .072 .129 .126 CQT-V .088 .046 .046 . 53 .112 CQT-Q .073 .041 .042 .160 .074 CQT-T .103 .051 .064 245 .089 Extraversion .025 .020 .21 .42 . 70 Neuroticism .003 .009 .002 .169 Eysenck Lie .012 .031 .064 .089 .045 A-H Test Anxiety .052 .039 .040 .102 .093 Test Anxiety-1 .001 .000 .000 .000 .000 Test Anxiety-2 .019 .025 .027 .087 .042 Test Anxiety-3 .023 .028 .035 .077 .073 Accuracy .009 .004 .021 .047 .032 Success .021 .015 .090 .014 .023 Learning Set .041 .030 .031 .046 NA NA .074 Reason Enrolled .004 .002 .012 .066 .013 Pre-attitude .005 .014 .082 .182 .126 Post-attitude .010 .011 .082 .113 .181 Discussion Attitude .009 .015 .041 .039 .071 50-50-1 (Asp,-Exp.) .085 .036 .044 .047 .044 50-50-2 (A8p.-Exp.) .192 .079 .169 .244 Pretest .082 .043 .115 .143 .172 Midterm .339172 .143 .173 Instructor Extraversion .011 .010 . 80 .045 .071 Instructor Neuroticism .020 .012 .065 .033 .038 Instructor Marlowe-Crowne .007 .008 .045 .028 .064 Instructor Sanford-Gough .097 .006 .045 .047 .116 Instructor California-F .219 .006 .043 .057 .116 Instructor Risk .219 .021 .045 .089 .217 Instructor Course Load .239 .017 .031 .049 .026 Instructor Teaching Exp. .109 .018 .123 .045 .071 Instructor ED 200 .186 .008 .036 .038 .044 N 532 286 60 33 24 7 26 TSSi/TSST 1.000 .378 .066 .033 .017 .002 .017 MEAN 3.633 3.200 2.644 2.394 3.000 3.143 2.192 Prcporticn of variation in each group explainable for each predictor O-Split made on this variable. @Next best BSS/TSS . =Final group. NA (BSS/TSS)i =Split attempted but not made. =Split not attempted.