THE RELATIONSHIP BETWEEN INTENSITY OF BELIEF AND LEVEL OF COGNITIVE - COMPLEXITY Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY JOHN F. FIELDER 1970 ....... IIII LIBRARY Michigan State IIIIIIIIII III III IIIIIIIIIII II 1+”:qu 3 1293 105750 University This is to certify that the thesis entitled THE RELATIONSHIP BETWEEN INTENSITY OF BELIEF AND LEVEL OF COGNITIVE-COMPLEXITY presented by John F. Fielder has been accepted towards fulfillment of the requirements for Ph.D. degree in Education . "\ ‘ A {551' a rv¢§ \\ $4 _‘ ‘1 AL \I V» i W} L I l’ o ' Date //: 7/; L, 0-169 ABSTRACT THE RELATIONSHIP BETWEEN INTENSITY OF BELIEF AND LEVEL OF COGNITIVE-COMPLEXITY By John F. Fielder The author's purposes in this study were: (1) to examine the relationship between intensity of belief about a social attitude object and the level of cognitive-complexity exhibited by subjects on a task requiring them to deal with that object, and (2) to examine the validity of two different methods, hence heuristic theoretical conceptions, in assessing level of cognitive-complexity. As predicted, a significant relationship was found to exist between intensity of belief as measured by an attitude scale and level of cognitive-complexity exhibited by subjects in writing an essay dealing with that stim- ulus domain the attitude represented. It was found that subjects with high intensity of belief about an attitude, whether for or against, exhibited a lower level of cognitive-complexity than did subjects who had a moderate intensity of belief about the attitude. Sub- jects with a moderate of belief about the attitude John F. Fielder object were found to be significantly higher in the level of cognitive-complexity exhibited (P<.Ol). The significant relationship found to exist between belief intensity and level of cognitive- complexity was obtained through application of the Schroder technique for scoring structural variables (Schroder, at al., 1968). This measure of complexity of cognitive functioning focuses on exhibited struc- tural variables taking into account, but not including, a specified stimulus domain. No such relationship was found to exist when a general measure of complexity (Barron, 1967) was given to the same subjects. Con- tingent upon the validity of the two techniques, the evidence supports a theoretical assertion of this study that cognitive-complexity should be assessed purely at a structural level and with reference to a specified stimulus domain. The second measure of cognitive-complexity, the Barron general scale (Barron, 1967), was included in this study to help determine the validity of both tech- niques through the application of certain validational procedures. Interpreting the evidence in terms of the rationale underlying these procedures led to the con- clusion that there is a reasonable basis for asserting the validity of the Schroder technique and that further study should be given to this approach to assessing L levels of cognitive-functioning. THE RELATIONSHIP BETWEEN INTENSITY OF BELIEF AND LEVEL OF COGNITIVE-COMPLEXITY BY. I . . r.\ ~. ,- John Ff Fielder A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education 1970 I? Copyright by JOHN FREDERICK FIELDER 1970 ACKNOWLEDGMENTS To all, be they friend or foe, who knowingly or unknowingly assisted in the conception and realization of this thesis, I am heavily indebted. In particular, special thanks is due my chairman, Dr. Samuel A. Moore, II, and my committee members Dr. Lee Shulman, Dr. Richard Featherstone, Dr. Ellen Strommen, and Dr. Harvey Clarizzio. And to those members of my family who watched my academic odyssey with apprehension . . . . I thank you for your patience and confidence. ii TABLE OF CONTENTS ACKNOWLEDGEMENTS . - . . . . . . . . . . Chapter I 0 INTRODUCTION 0 0 o o o I o o o a General Introduction to Cognitive- . Complexity . . . . . Problem: Theoretical Rationale for Study Problem: Assessing Level of Cognitive- Complexity . . . . . Definitions of Theoretical Variables. . . II 0 METHOD 0 O o O o c 0 c O o o o o Hypotheses . . . . . . . . Subjects . . . . . . . . . Selection of Subjects. . . . . Procedure: Attitude Scales. . . . . Procedure: Measures of Cognitive-Complexit Procedure: Scoring of Content Specific Cognitive-Complexity. . . . . . . . Data Analysis . . . . . . . . . . III. RESULTS . . . . . . . . . . . . . ' Inter-rater Reliabilities . Relationship Between Belief and Comp; exity- scores 0 I c O o 0 o O o O Instrument Validation. . . . . . . . Attitude Scales. . . . . . . . . . IV. DISCUSSION . . . . . . . . Belief Intensity and Level of Cognitive- Complexity . . . . . . Assessment of Attitudes and Cognitive- CompleXity o o o o I o 0 Educational Implications of Study. . . . REFERENCES . . . . . . . . . . . . . . APPENDICES . . . . . . . . . . . . . . Page ii 41 145 51 58 62 LIST OF TABLES Page Results of the Spearman rS Test of Correlation Between Rater's Judgements of Complexity. . . 30 Results of the Kruskal-wallis H-Test of the Significance of Difference Between Com- plexity Scores . . . . . . . . . . . 32 Results of Kruskal-wallis Post Hoc Procedure to Test for Significant Differences Between Mean Ranks. . . . A. . . . . . . . . 33 Results of Statistical Contrast Between the Two Groups of High Belief Intensity and the One Group of Moderate Intensity. . . . . . 3A Correlations (Spearman rs) Between Belief Intensity about One Attitude Area and Level of Exhibited Complexity with Respect to Another Area . . . . . . . . . . . . 35 Correlations (Spearman rs) Between Intensity of Belief in an Attitude Area and Level of Cognitive-Complexity as Assessed by a Gen- eral Measure of That Trait . . . . . . . 36 Correlational Relationships Between Birth Control Belief, Education Belief, Birth Control Complexity (Schroder), Education Complexity (Schroder) and Barron Complexity (Spearman rs). . . . . . . . . . . . 39 Correlations Between Scores on the Birth Control Attitude Scale and the Education Attitude Scale . . . . . . . . . . . 40 iv LIST OF APPENDICES Appendix Page A. Birth Control Scale . . . . . . . 63 B. Education Scale . . . . . . . . 66 C. Scale Points used for Assessing the Conceptual Level Involved in Essay Writing . . . . . . . . . . 69 D. Instructions and Letters for Test of Cognitive—Complexity . . . . . . 72 E. Barron Scale of Complexity. . . . . 77 F. Description of Experiment Given to Education 200 Students, Summer 1969 . 80 G. Attitude Scales and Filler Items. . . 84 H. Letter to Selected Subjects . . . . 89 CHAPTER I INTRODUCTION General Introduction To Cognitive—Complexity The concept of cognitive—complexity is central to theories of human information processing. Gener- ally speaking, cognitive-complexity is a construct used to indicate the manner in which a person construes his social and physical world. The existence of this contruct is posited on the assumption of mediating cognitive structures through which a person's experi- ences with his social and physical world are processed. Bieri has defined cognitive-complexity in terms of the degree of differentiation of a person's con- struct system, i.e., " . . . a person who is able to evoke more dimensions of judgement in construing others is more cognitively complex than a person who invokes fewer dimensions of judgement" (Bieri, 1966, p. 18). The idea of differentiation of mediational cognitive structures as used by Bieri is generally attributed to Kelly (1955). Harvey, 32_ al, (1961) took the complexity of a person's perceptions, as conceptualized by Bieri, to be indicative of the degree of abstractness of his cognitive structures. In Harvey's formulation, a con— tinuum of conceptual functioning was postulated. At one end of the continuum rest the simple or concrete structures, and at the other end rest the very complex or abstract structures. The concept of abstractness, as used by Harvey, refers to a person's ability to manipulate dimensions of experience without the neces- sity for their physical presence. Schroder, 22 El: (1968), in a theory of person- ality functioning, developed a conception of cognitive- complexity which appears to resemble what Kelly had in mind more closely than did the conceptualizations of either Bieri or Harvey. Although Kelly never referred specifically to "cognitive-complexity" as a construct, he did use differentiation of cognitive structures as a central construct in his theory. Together with the ability to differentiate among the dimensions of a social situation, Kelly spoke also of the ability of a person to abstract trends from these combined dimensions (Kelly, 1955). The abstracted trend represents a synthesis or integration of the differentiated dimensions of a social situation. The more complex the differ— entiation and synthesis, according to Kelly, the more likely the person will be able to successfully predict outcomes of alternative responses to a social situation. Schroder, 23. EL} (1968) also used the concept of differentiation of cognitive structures in their theory of human information processing. And they added the concept of integration. The "trend abstraction" concept of Kelly‘s may be equated with the concept of integration. However, Schroder used the title of "integrative complexity" to represent an individual's ability to construe experiences in a multidimensional and abstract manner. For Schroder, the differentiated dimensions represent the units of conceptual functioning, i.e., the elements or content of thought. Synthesis or trend abstraction (integration rules) represents the styles, schema, programs or controls for combining the units of information which are variously called cues or stimuli (Schroder, 32 al,, 1968, p. 258). In Schroder's formulation, then, conceptual structures are made up of two interdependent parts. Level of cognitive-complexity (integrative-complexity) is a function of the dimensions or units of information perceived, differentiated and utilized, and also the complexity of the rules a person has available for integrating these dimensions of experience. Essentially the concept of cognitive-complexity is relational in nature. In an information processing approach to personality functioning, a distinction is made between the content or elements of cognition, and the structure or processes of cognition. Structure is viewed as referring to the relations between the elements or content of cognition (Scott, 1963, p. 266). For example, today, at one level of analysis, Israel and Egypt represent two elements of thought. The relationship seen to exist between these two countries represents the structure of a person's cognition. Thus, level of complexity is seen to be indicative of an individual's ability to abstractly situate elements of cognition in such a manner that he might discern their differences and similarities, their one- ness yet separateness. .The cognitively-complex person, relatively speaking, sees gradation of grays while the cognitively-simple person sees but blacks and whites. For the purposes of this study the conceptualiza— tion of cognitive-complexity (integrative-complexity) conceived by Schroder, 33. al, (1968) will be used. The problem of assessment of the construct is dealt with following the next section. Problem: Theoretical Rationale for Study Much recent literature, (Barron, 19535 Berkowitz, 1957; Bieri, 1955, 1961, 1966; Scott, 1962, 1963a, 1963b; Messick and Kogan, 1966) has dealt with the con- cept of cognitive-complexity. In most cases, cognitive- complexity has been conceptualized as an independent cognitive variable which influences a person's percep- tions of his social and physical worlds. How a person construes his physical and social environments has been seen to be primarily dependent upon his overall level of cognitive-complexity. The underlying assump- tion in most instances appears to be that there is a unitary trait of cognitive-complexity which influences all perceptions in like manner. Notions as to the generality of cognitive-' complexity have varied considerably. Some writers, e.g., Bieri and Blacker (1956) have viewed cognitive- complexity as a general personality trait which influ- ences all perception in like manner. Gardner and Schoen (1962), on the other hand, assert that evidence indicates that most people can be relatively complex in some areas of perception while being relatively simple in others. Scott (1963a) asserts that evidence is weak for conceptualizing a unitary trait of cognitive- complexity which crosses all conceptual structures. Fielder (1969) has suggested that content related cognitive structures may be grouped in a finite number of constellations, each constellation varying from the others with respect to level of cognitive-complexity exhibited. Schroder, e£_ El» (1968) and Scott (1963b) have suggested that level of cognitive functioning is more a function of the content of cognition, and any assess- ment of structural properties must be made with refer- ence to the particular stimulus domain under consideration. Both authors stress the importance of focusing upon the relatedness of content variables to structural variables. The assertion is also made that a general measure of cognitive—complexity is of relatively small value in predicting level of cognitive functioning with regard to any specified stimulus object. Schroder also has stated that evidence shows that the more concrete or simple a person's conceptual structure " . . . the more a person's functioning becomes rigidly determined by an absolute belief that structures the world in a fixed way" (Schroder, 22' al., 1968, p. 128). This suggests that an evaluative or) belief dimension may indeed be the independent variable of cognitive functioning, i.e., the level of cognitive- complexity an individual exhibits regarding any social object may be a direct function of the intensity of the belief the person has about that object. Schroder also points out that " . . . the authoritarian person, the rigid person, and the dogmatic person have all been characterized as holding relatively extreme attitudes" (Schroder, 33 al,, 1968, p.128). In the same view, Rokeach (1968) and Fishbein (1967) have both posited the existence of evaluative as well as cognitive dimensions in attitudinal structures. Also, since Osgood (1957) has shown that virtually any concept loads on an evaluative dimension, it becomes reasonable to believe that the level of cognitive- complexity exhibited regarding any social object in which a judgement is to be made would therefore be dependent upon the intensity of belief a person has about that social object. Consequently, if it is indeed the case that level of cognitive-complexity is a function of the intensity of belief that a person has about any social attitude object, it then follows that scores on any social attitude scale should be relatively good pre- dictors of the level of cognitive-complexity indivi- duals will exhibit regarding that object. It was the author's purpose in this study to examine the relationship between intensity of belief about a specified stimulus domain (social attitude object), and the level of cognitive-complexity exhibited by individuals regarding that object. The instrument used to assess level of cognitive- complexity was of a type which allowed consideration of content variables as well as structural variables. Level of cognitive-complexity was assessed with refer- ence to a specified stimulus domain--this in direct contrast to the general type of instruments which pur- portedly assess level of cognitive-complexity with reference to structural variables alone. The problem of assessment is discussed in the following section. Problem of Assessing Level of Cognitive-Complexity Many attempts have been made to determine the generality of the personality trait of cognitive- complexity. In a study conducted by Bieri and Blacker (1956), a methodological issue and a theoretical issue were considered. Of methodological concern was whether a single empirical measure could sufficiently assess level of cognitive-complexity. Of theoretical concern was whether there is a general trait of cognitive- complexity which influences perception across all cognitive structures or, whether cognitive-complexity must be specified in terms of a specific stimulus domain. For the purpose of their study, Bieri and Blacker used two different instruments to assess level of cognitive—complexity. They also made a distinction between two major divisions of stimulus environments. To test for complexity with regard to the social environ- ment (perception of people), the Role Construct Repertory Test (RCR) (Kelly, 1955) was used. A modification of the Rorschach ink blot (deter- minant-complexity and content-complexity) was used to test for complexity with regard to the nonhuman stimulus environment. The Rorschach test of complexity was measured in terms of response variability in content and deter- minants of responses. The RCR test of complexity was measured in terms of the number of different verbal constructs elicited. Significant relationships were found to exist between responses to the RCR test and the Rorschach test (from P<.005 to P<.05) which led the authors to suggest that there are relatively enduring and consistent modes of cognitive functioning which are characteristic of a person's perception across both of the environments specified in the study. Although significant relationships were found in this study, they do not appear useful in resolving the methodological nor theoretical problems posed by the authors. There is no evidence that the two instru— ments used were indeed measuring the same trait across the two stimulus environments. If there is a general trait of cognitive-complexity, then it is reasonable to expect that a single instrument measure the trait satisfactorily regardless of the stimulus domain involved. The procedure of using two different instruments to assess the same trait in two different stimulus environ- ments raises strongly the question of validity. Also, the authors failed to raise the question of generality of cognitive-complexity across different stimulus situations in either of the two environments under consideration. Had they tested for this lO possibility, it might have been found that correlations between complexity scores would not always be high. It is possible that variability within each stimulus environ- ment is quite large and a sample of only a few sub-domains is not adequate for making generalizations to the total stimulus environment. Allard and Carson (1963) developed three tests of cognitive-complexity based on the RCR test. Generated constructs for the three tests of cognitive-complexity were: (1) personal friends, (2) famous people, (3) geo- metric designs. Intercorrelations between the three stimulus domains were from .57 to .67. The findings of this study also led the authors to suggest the exis- tence of a unitary trait of cognitive—complexity. It is felt here, however, that the scope of the study was too limited for the generalizations made. Vannoy (1965) examined a battery of twenty instru- ments which were designed to measure, or which were construed to measure cognitive-complexity. A factor analysis of subject's responses to the twenty instru- ments indicated that no single dimension could be held accountable for the derived correlation. In all, eight factors were extracted, the largest accounting for 2A.3 per cent of the total variance. vannoy's conclusion from the results was that cognitive-complexity is not as general a trait as is often implied. It is doubtful, however, that this ll conclusion was warranted. In the first place, the large number of factors extracted reflects upon the construct validity of the instruments purportedly assessing level of cognitive-complexity. Alternative hypotheses would suggest that either differences in theoretical conceptualizations of the construct vary, or that none of the instruments indeed measures the construct. It was also noted in the Vannoy study that the Schroder measure of integrative complexity (sentence completion test) represented a single factor in itself. Out of the total variance, this factor accounted for 6.3 per cent. The reason the Schroder measure of com- plexity was represented on a single factor was attri- buted to the fact that all other instruments tested in the study were essentially measures of dimensional com- plexity. This would seem to be in keeping with Bieri's (1966) definition of cognitiVe-complexity which is stated in terms of differentiation of a person's con— struct system. Thus, the more differentiated or dimen- sionalized a person's construct system, the more cognitively-complex he is. It therefore appears that there are two divergent streams of thought regarding the construct cognitive- complexity. The followers of one stream see cognitive- complexity as being essentially dimensional in nature. 12 This conceptualization has been utilized by many writers in developing multidimensional scaling models for assesS- ing the dimensionality of cognition (Messick and Kogan, 1966; Jackson and Messick, 1963; Tucker and Messick, 1963). However, it has been previously noted that cognitive-complexity is seen by some writers to be more than just an ability to differentiate among informational dimensions. The second stream of though regarding this construct requires an integrative or synthesis dimen- sion, as well as the dimension of differentiation. This is in keeping with the theory of personality developed by Kelly (1955), and as most recently posited, by Schroder, gt. a}, (1968). It is consequently the case that while dimen- sional analysis can be quite useful in the study of cognitive structures, it is not in itself sufficient to the determination of level of cognitive functioning regarding a specified stimulus situation. As yet, there appears to be no evidence which would support the assertion of a general trait of cognitive-complexity. What seems to be required is a measure of cognitive- complexity which accurately depicts the ability of an individual to abstract trends from the dimensions of any situation inIa complex manner. Also, this measure should be able to take into consideration the particular stimulus domain being considered (Scott, 1963b; Schroder, 33 gin 1968). 13 Even from a purely subjective standpoint there does not appear to be sufficient evidence for postulat- ing a constant or general trait of cognitive-complexity. For example, a person might be very complex in con- struing the behavior of members of his own ethnic group, but, on the other hand, may be extremely cate— gorical when construing the behavior of persons belong- ing to some other ethnic group. One does not have to search far to find a person who is extremely categorical when it comes to religion, but who may be extremely noncategorical with respect to politics. Situationally, also, level of cognitive-complexity may vary with respect to the same object. For example, a person's perception of foreigners generally becomes much less complex in time of war. A case in point is the treatment of Japanese—Americans following Japan's attack on Pearl Harbor. Disregarding even the fact that some of these people were second generation Americans, most were categorized as "Japanese" and incarcerated as such. Considering the theoretical requirements of this study, the most promising approach to assessing level of cognitive-complexity appears to be Schroder's "general manual" for scoring structural properties of responses. In essence, " . . . the general manual represents a set of general operations for inferring the level of conceptual structure that generated the 14 response. The manual directs the rater to consider the degrees of freedom in the rules of integration in the mediating processes underlying the response" (Schroder, 33' al., 1968, p. 190). In all instances where the "general manual" has been used it has been possible to specify the stimulus domain under consideration. And, although the refer- ence is to a specified content or stimulus domain, the trained rater does not consider content referents in scoring, but is told, " . . .regardless of what the persons says, what complexity of structure would be required to generate the response?" (Schroder, 33_ al., 1968, p. 190). The "general manual" has been used to score res- ponses generated by sentence stems, (Schroder and Streufert, 1962), and also in scoring conceptual rules involved in the writing of essay answers to examination questions, (Claunch, 196A; Schroder and Phares, 1965). "The essay appears to provide the optimal opportunity (in contrast to responses generated by sentence stems) for a person to utilize high level conceptual rules" (Schroder, 1968), p. 200). Claunch (1964) examined the level of conceptual (cognitive) complexity exhibited by individuals in writing essay examinations. Using the Schroder Sentence Completion Test of Integrative Complexity (Schroder and Struefert, 1962), two groups of subjects were selected. 15 The first group was composed of individuals judged to be concrete (simple) in conceptual functioning, and the second group was composed of individuals judged to be abstract (complex) in conceptual functioning. The Sentence Completion Test was used as an independent variable so as to in part validate the "general manual" of structural variables to be used in scoring the essays. The essay question to be answered by the sub— jects was judged to be sufficiently related in content to the Schroder Sentence Completion Test. Content relatedness of the two tests was necessary to rule out the alternative hypothesis of differences in conceptual functioning due to differences in the stimulus domains examined. The type of essay question itself was chosen for several reasons. In the first place, the question dealt with two theories complex enough in nature that they were thought to provide an adequate number of concepts within themselves for subjects to make com- parisons, contrasts, and integrations. Secondly, the two theories to be dealt with had been presented to the subjects previously in such a way that the author could be reasonably certain that integrations and comparisons necessarily would be generated internally by the sub- jects, and not from memory alone. Finally, it was felt that asking the subjects to make comparisons between two theories provides an excellent opportunity 16 for the use of conceptual rules which would distinguish the cognitively concrete (simple) from the cognitively abstract (complex) person. The structural variables tested in this study were polarized contrasts, qualified contrasts, and integrative contrasts. It was hypothesized that the q cognitively complex individuals would use more qualified II contrasts and integrative contrasts than would the more cognitively simple individuals. This hypothesis was I'” supported in the study. It was also hypothesized that the less cognitively-complex individual would use significantly more polarized-contrasts than would cognitively-complex individuals. This hypothesis was not supported statistically at P<.05. However, results were in the predicted direction and sigificant at P<.lO. Claunch's study was important in that results showed that structural variables can be successfully used as criteria by trained raters to assess level of cognitive functioning. It also gives added support to the use of essay type exams to elicit responses which will be indicative of a person's level of cognitive- complexity regarding a specified stimulus domain. Schroder found in reviewing past research that by using " . . . stimulus situations implying conflict, uncertainty and control in a certain domain" (Schroder, 33 al,, 1968, p. 186) that more construct relevant responses can be produced. For purposes of research 17 then, it appears to be advantageous if subjects are required to work with a specified stimulus situation in which they must make judgements. The "general scoring manual" used by Claunch (196A) and Schroder and Phares (1965) for assessing level of conceptual functioning according to theoretically and operationally defined structural variables is represented on a seven point scale. At the present, four gross nodal points have been defined (1,3,5 and 7). Each nodal point represents a level of conceptual function- ing (1=low, 7=high). The operationally defined struct- ural variables used by raters to assess level of cognitive-complexity appear in Appendix C. For the purpose of this study, then, the "general manual" developed by Schroder, gt a}, (1968) was used to assess level of cognitive-complexity subjects exhibit in writing essays. Each subject wrote an essay type answer in which they made judgements about a specified stimulus domain. The particular domains con- sidered and the rationale for their selection is pre—I sented in the procedures section. Definitions of Theoretical Variables Cognitive-complexity Cognitive-complexity is a continuous cognitive variable which is indicative of the manner in which a person construes his social and physical environments. 18 The more complex a person is the better able he is to differentiate the informational dimensions of a problem, integrate these dimensions in a myriad of ways, and finally, predict the outcomes from the various integra— tions. Attitude "An attitude is the evaluative dimension of a concept . . . the attitude is the sum of beliefs about an object" (Shaw and Wright, 1965, p. 3). Also, "an attitude is defined simply as an organization of inter- related beliefs around a common object" (Rokeach, 1968, p. 116). Belief "A belief is any simple proposition, conscious or unconscious, inferred from what a person says or does, capable of being preceeded by the phrase 'I believe that'. A belief is a predisposition to action" (Rokeach, 1968, p. 113). CHAPTER II METHOD Hypotheses The following general assertion was generated out of the relationship theorized to exist between belief in a social attitude object and the level of cognitive-complexity a person exhibits toward that object. 1. The intensity of belief about a social attitude object is directly related to the level of cognitive-complexity an individual will exhibit with regard to that object. (This relationship is schematized in Figure one). From this general assertion the following working hypotheses were derived: Hl:a Individuals reporting high intensity of belief about an attitude object (high- agree and high-disagree on an attitude scale) will be judged of low cognitive- complexity on a task requiring them to deal with that object. :b Individuals reporting moderate intensity of belief about an attitude object (moderate-agree to moderate-disagree on an attitude scale) will be judged to be of high cognitive-complexity on a'taSk requiring them to deal with that object. i.e., given attitude domains A and C taken separately, with X=high intensity of belief for, Y=moderate intenSity of belief, and Z=high intensity of belief against, then HO becomes p(X>Y)=p(X>Z)=p(Y>Z)=l/2 with respect to level of complexity exhibited. l9 20 high Level of Cognitive- Complexity X Z low . high-agree moderate-agree- high-disagree disagree Belief Intensity Figure 1. Hypothesized Relationship Between Belief Intensity and Level of Cognitive-Complexity Intensity of belief about one attitude object is not significantly correlated with the level of cognitive—complexity exhibited with regard to another social attitude object. 3 Intensity of belief about a social attitude object is not significantly correlated with the level of cognitive-complexity on a general measure of that trait. Also, given the variables of this study where A = Intensity of belief toward birth control B 8 Complexity exhibited toward birth control C = Intensity of belief toward education D = Complexity exhibited toward education E - General complexity the following correlational relationships are predicted to exist. rAB>rAE>rAD rCD>rCE>rCB rBD>rBE=rDE 21 Subjects The sample for this study was drawn from a pool of 270 Michigan State University undergraduates enrolled in an education course entitled "Individual and the School". Seventy-five female subjects were 'I selected from this pool by the method indicated below. Selection of Subjects Subjects were selected on the basis of their scores on a birth control attitude scale developed by Wilke (1934). Since a similar distribution was obtained for both sexes, it was decided for practical and analytical purposes to use a homogeneous sample of female subjects only. From the 169 female undergrad- uates who originally took the attitude scale on birth control, the twenty—five who scored highest and the twenty-five who scored the lowest were selected as being respectively representative of high belief against and high belief for birth control. From the remaining 119 individuals, a third group of twenty-five was chosen at random as being representative of those with moder— ate intensity of belief toward birth control. Two weeks later, letters were then sent to those students selected asking them to be paid volunteers in the second half of the experiment (Appendix F). In this manner it was possible to eventually obtain a total N=60, twenty subjects being in each group. 22 Procedure: Attitude Scales Two attitude scales were used in this study. The birth control scale was developed by Wilke (193A) and is a 22-item Likert type scale scored from one (lsstrong agree) to seven (7=strong disagree) with scores that can range from 21 to 147. A high score indicates strong belief in disagreement with birth control, while a low score indicates strong belief in agreement with birth control. This particular attitude scale was chosen because the stimulus domain it reflects appeared to be of such a controversial nature that a wide distribution of scores would be obtained from which to choose subjects. One item on the scale was eliminated (item 6) because it was not relevant to present day concerns regarding birth control. This was ascertained not to affect the validity of the scale (Shaw and Wright, 1967). The reliability and validity values of the scale are as follows: 1. Reliability: a. split—half = .91 b. test-retest = .88 to .83 2. Validity: all items have content validity with regard to moral and pragmatic reactions to birth control. Strong face validity (Shaw and Wright, 1968, p. 136). The birth control scale is reproduced in Appendix 23 The second scale used in this study was one concerning attitudes toward progressive education developed by Kerlinger and Kaya (1959). It is a 20— item Likert type scale and is scored from -3 (strong- disagree) to +3 (strong—agree). Scores can range from -60 to +60. A positive score indicates progressive attitudes toward education, and a negative score indi- cates traditional attitudes toward education. The reliability and validity of the scale are as follows: 1. Reliability: a. Split-half = .83 b. test-retest = .76 2. Validity: validity estimates considered satisfactory (Shaw and Wright, 19683 p0 84). The education scale (Appendix B) was included in this study for analytical reasons and its purpose is explained in the section on method validation. The birth control and education attitude scales were administered together. All items from both scales were randomly mixed with 19 filler items so that the total number of items each subject initially responded to was equal to sixty (Appendix G). Procedure: Measures of Cognitive—Complexity'_ Two different assessment techniques were used to ascertain level of cognitive-complexity. One of these was the Barron scale of complexity (Barron, 1967). This scale contains twenty true and false items and is 24 generally regarded as an overall or general measure of cognitive-complexity. The Barron scale of complexity was administered to all sixty subjects and, like the education scale, was included for analytical purposes (Appendix E). This will be explained in the section .on method validation. The second assessment technique used to ascertain level of cognitive-complexity was that developed by Schroder, g£_ al., (1968). This technique consists of having trained raters use the "general scoring manual of structural variables" (Appendix C) to score essay type responses of subjects dealing with a specified stimulus domain. The use of this assessment technique required that subjects assume the role of a physician at one time and an educator at another. The task each subject was to perform under these assumed roles was to respond in an "Ann Landers" fashion to a query in letter form from a person seeking advice about a specific topic. Therefore, each subject wrote two letters. The first letter was a response to a woman seeking advice about birth control (Appendix D). The second letter was a response to a woman seeking advice about progressive and traditional education (Appendix D). To guard against any kind of serial effect from the presentation, half of the subjects received the birth control letter first and the other half the 25 education letter. Since subjects appeared for the experiment at their convenience, the requirement of randomization was considered met. Subjects were given 20 to 25 minutes to complete each letter. However, they were instructed to take more time if it was felt to be necessary. Procedure: Socring of Content Specific Cognitive-Complexity In all, 120 essay type answers were collected. Sixty of these dealt with the topic of birth control, and the other sixty dealt with the topic of progressive- traditional education. Each essay was scored by two raters using the "general scoring manual" already described. This manual is reproduced in Appendix C. As previously explained, although content was important for the purposes of this study, raters were instructed to consider not what the person said, but how he said it. The focus was on structural rather than content variables. As ". . . the major requirement of reliable and valid scoring is a thorough grasp of the theoretical variables describing structural variation and a consideration of each response in these terms" (Schroder, 23, al,, 1968, p. 187), it was necessary to initially train the two raters using simple materials and discussions of the concept of cognitiveécomplexity in structural terms. 26 An additional analysis of the essay responses was also performed. Those essays which exceeded more than one and one-half handwritten pages in length, thereby indicating a prolific writer and not necessarily a more cognitively-complex person, were divided in half and reread separately. A split-half reliability a coefficient was then computed to ascertain whether “1 length of response influenced rater's scores. Data Analysis a. Inter—rater reliability coefficients An inter—rater reliability coefficient was com- puted to test the reliability of judgements made by the two trained raters regarding the complexity of the collected essay answers. In addition, a split-halfs reliability coefficient was computed to ascertain whether or not length of a response was an important variable affecting rater's scores as to level of cog- nitive functioning. b. Tests of differences between groups To test for complexity differences between the three groups examined in this study, the Kruskal-Wallis H-Test of variance by ranks was used. (Siegel, 1956, 184-193). This nonparametric technique tests the null hypothesis that k-samples come from the same population under the assumption that the variable under study is part of an underlying continuous distribution. 27 The Kruskal-Wallis technique was applied to the complexity scores obtained for each of the groups under consideration in this study: a) complexity scores obtained from rater's judgements of essays written by subjects on the topic of birth control (Schroder analytic technique), b) complexity scores obtained from rater's judgements of essays written by subjects on the topic of education (Schroder analytic technique), and c) complexity scores obtained from the administration of the Barron general scale. Since the Kruskal—Wallis technique does not indicate the direction when a difference is found to exist, it was found necessary to carry out further analysis using the post hoc procedures for the Kruskal— Wallis test (Marascuilo and McSweeney, 1967). To test for differences of significance, the Kruskal-Wallis post hoc procedures contrast the mean ranks of the groups under consideration and indicate direction and magnitude of difference. c. Method validation The rationale underlying the multimethod-multitrait matrix for convergent and discriminant validation (Campbell and Fiske, 1959) was seen to be particularly useful in this study for helping to show the validity of the Schroder technique for assessing level of cognitive-complexity. Since concern in this study was 28 with structural functioning with reference to specified content variables, it was important to show as strongly as possible that the scoring procedure for structural variables was indeed measuring level of cognitive functioning. In essence, discriminant validation is used for the justification of novel trait measures, for the validation of test interpretation, or for the establish— ment of construct validity. The rationale underlying the procedure for discriminant validation is that tests can be invalidated by too high correlation with other tests from which they were supposed to differ. The basic requirement of this procedure is that there are at least two different traits to be measured. The two traits considered in this study were: 1) complexity with regard to birth control, and 2) complexity with regard to education. Convergent validation is confirmation by independ- ent measurement techniques and requires at least two measures of the variable under consideration. This requirement was considered met since there were two, measures of cognitive-complexity used in this study (Schroder and Barron). If the correlation between the two tests is high this is interpreted to mean that both tests are measur- ing the same variable. However, if the correlation is 29 low it is necessary to examine the evidence in favor of several alternative propositions: l) neither method is adequate for measuring the trait, 2) the trait is not a functional unity, the response ten- dencies involved being specific to the nontrait attri- butes of each test, or 3) one of the two methods does not measure the trait. Wfisfu‘ A“; III CHAPTER III RESULTS ‘Inter-rater Reliabilities To check the reliability of judgements made by the two raters with respect to the complexity of essays, a Spearman rS was computed using the two rater's scores on both birth control and education complexity. Table 1 shows the correlations derived from these scores. The correlations indicate an acceptable level of reliability in the raters judgement. TABLE l.--Results of the Spearman r Test of Correlation Between Rater's Judgements of Complexity. * birth . group control education I .76 .86 II .85 .87 III .87 I .8: * Group I = high belief intensity for attitude object Group II = moderate belief intensity toward attitude object Group III = high belief intensity against attitude object 30 31 Ten of the essays were arbitrarily categorized as being of such a length as to possibly affect the judgement of raters as to the level of complexity exhibited. It was feared that while not being more complex, a prolific writer may be more likely of being judged more complex. The results of a split-halves correlation in which the longer essays (more than a page and one-half in length) within each attitude area were split in two and separately re- judged as to level of cognitive-complexity exhibited are as follows: for birth control essays rS = .63 (N=6), for education essays r8 = .59. Correlations were not signifi- cant (p>.Ol) which might indicate that length of an essay response may have some influence on rater's judgements of complexity of the response. Relationship Between Belief and Complexity Scores Following the Kruskal-Wallis procedure outlined by Siegel (1956, pp. l8A-l93), complexity scores derived from rater scoring according to the Schroder technique (birth control and education attitude areas) and complexity scores derived from the Barron scale of general complexity were analyzed separately to test for differences between the three groups (each representing a different intensity of belief in the attitude objects under consideration). 32 Table 2 shows the results of the Kruskal-Wallis H-Test of significance between groups. As expected, complexity scores derived from essays on the topic of birth control show a significant difference (p<.Ol). However, it was not possible to reject the null hypothesis regarding complexity differences between groups of vary- I, ing intensity of belief as determined on the education attitude scale. Nor was it possible to reject the null hypothesis regarding complexity differences between the I'? three groups in either the birth control or education attitude areas using the Barron scale of general complexity. TABLE 2.--Results of the Kruskal-Wallis H-Test of the Significance of Differences Between Complexity Scores. Complexity birth control education Barron N 20 20 20 degrees of freedom 2 2 2 a H 16.8A 5.67 1.71 H corrected * for ties 16.97 5.74 1.72 * denotes significance at p<.Ol. Table 3 shows the results of a Kruskal-Wallis post hoc analysis done in accordance with the technique out- lined by Marascuilo and McSweeny (1967). Using this 33 technique differences in birth control complexity scores between the three groups of varying belief intensity were identified using the contrast procedure comparing the mean ranks of the groups. As expected, a significant difference was found between the mean ranks of groups one and two (p<.Ol) where group one is high belief intensity for birth control and group two is moderate belief inten- sity against birth control. A significant relationship was also found between groups two and three (p<.Ol) where group three is high belief against birth control. TABLE 3.--Results of Kruskal-Wallis Post Hoc Procedure to Test for Significant Differences Between Mean Ranks. COMM“ IIIIII IIIIIII - ._ * r1 - r2 -33.57 -19.17 51 - 53 40.149 3.89 52 - 53 15.88 30.27* * denotes significance at p<.Ol. Since no differences were found between the complexity scores of groups one and three (both high intensity of belief), it was tested whether, as a group, the mean ranks of groups one and three differed from the mean rank of group two (moderate intensity of belief). Table A shows the two groups representing high intensity of belief and the one group representing moderate intensity of belief 3A about birth control differ significantly (p<.01) in the level of cognitive-complexity exhibited on the task. TABLE A.--Results of Statistical Contrast Between the Two Groups of High Belief Intensity and the One Group of Moderate Intensity. , lower upper I: contr°5t limit limit “‘ 2' ' ‘ u 10 80* r2 - r1 - r3 7. 51. I ll * denotes significance at p<.Ol. Based upon the results stated above, it is possible to reject the null hypothesis concerning the predicted relation- ship between high intensity of belief about an attitude object (birth control) and a low level of exhibited congnitive- complexity. With respect to the same attitude area, it is also possible to reject the null hypothesis concerning the relationship between moderate belief intensity and high level of exhibited complexity. However, it was not possible to reject the null hypothesis for the above two relationships based upon the findings relating belief intensity about education and level of exhibited complexity. As predicted, no significant correlation was found between intensity of belief about one attitude area and the level of exhibited cognitive—complexity with respect to 35 another attitude area (p>.Ol). It was therefore impossible to reject the null hypothesis of no relationship. The results of the correlational computations are shown in Table 5. Also, as predicted, it was not possible to reject the null hypothesis of no relationship between belief intensity about a social attitude object and level of complexity as determined by a general measure of that trait (p>.Ol). Table 6 shows the results of the corre- lational computations. TABLE 5.--Corre1ations (Spearman rs) Between Belief Inten- sity About One Attitude Area and Level of Exhibited Com- plexity with Respect to Another Area. Group r r AD CB I .415 .350 II .627* .413 III .370 .365 * denotes significance at p<.Ol. Intensity of belief toward birth control Complexity exhibited toward education Intensity of belief toward education Complexity exhibited toward birth control (EQUIP 36 TABLE 6.——Correlations (Spearman rs) Between Intensity of Belief in an Attitude Area and Level of Cognitive-Complexity as Assessed by a General Measure of That Trait. Group rAE rCE I .3A8 .162 II .365 .375 III .A06 .205 Intensity of belief toward birth control -General complexity Intensity of belief toward education ODIID IIII ll Instrum nt Validation Using the rationale underlying the concepts of discriminant and convergent validation (Campbell and Fiske, 1959), correlations between the variables described below were compared in order to ascertain in part the validity of the instruments used in this study to assess level of cognitive-complexity. With respect to discriminant validation, it was necessary to show that the correlations found to exist between birth control belief intensity and exhibited complexity (rAB); education belief intensity and exhibited complexity (rCD), should be higher than those found between either birth control or education belief intensity and the level of complexity as measured on the Barron scale (r AE and rCE). These correlations are reported in Tables 7a ‘- s“ 7"- '71., 37 and 7b. Except for group three where rCDrBE= rDE' Table 7c shows the correlations found to exist and supports the prediction. F? Tables 7a and 7b show the correlations found to fi‘ exist between the variables AB, AE, AD and CD, CE, CB . respectively. It was predicted that AB>AB>AD and CD>CE>CB. v” However, only in group three in Table 7a AB>AB>AD (.600> .AO6>.370). In all other areas AB>AECE.Ol). Table 8 shows the results of the correla- tional computations. TABLE 7.--Correlational Relationships Between Birth Control Belief, Education Belief, Birth Control Com- 39 plexity (Schroder), Education Complexity (Schroder) and Barron Complexity (Spearman rs) Group A Pia PAE rAD I .562* .348 .415 II .523 .365 .627* III .6oo* .uos .370 Group B rCD rCE rCB I .u95 .162 .350 II .568* .365 .ui3 III .119 .205 .365 Group 9 rBD r'BE rDE * ,* , I .750 .620 .055 II .556* .uiu .522 III .750* .555* .511 * denotes significance at p<.Ol. ['UUOCIJID II II II II II Intensity of belief toward birth control Complexity exhibited toward birth control Intensity of belief toward education Complexity exhibited toward education General Complexity 40 TABLE 8.-—Correlations Between Scores on the Birth Control Attitude Scale and the Education Attitude Scale (Spearman rs). Group rAC I .18u II .396 III .315 A = Intensity of belief toward birth control C = Intensity of belief toward education CHAPTER IV DISCUSSION Belief Intensity and Level of Cognitive—Complexity Ea The results of this study strongly support the general assertion that there is a direct relationship 1 between intensity of belief in a social attitude object ' and the level of cognitive-complexity individuals exhibit toward that object. Evidence points to a relationship in which individuals with low belief intensity about an attitude object are better able to deal with that object in more cognitively-complex manner. These indi— viduals see more than one side to a problem and are more willing to consider what could be described as conflicting alternatives. The evidence also demonstrates that individuals with high belief intensity about an attitude object are more restricted in the kinds of responses they may make in a situation requiring the use of that object. Persons of low complexity appear to be much more categorical and subjective in their Judgements and evaluations. They tend to see but one viewpoint and are more likely to defend it exclusively. Al 42 For purposes of method validation, two different tests of cognitive-complexity were used in this study. The Schroder technique disclosed a positive significant relationship between belief intensity in birth control and the level of cognitive-complexity exhibited toward that attitude object (p<.Ol). Again using the Schroder technique, a relationship was found to exist between belief intensity in education and level of complexity exhibited (p<.lO). However, the results from the Barron test of complexity did not statistically lend support to these findings. In the following paragraphs the results of this study are interpreted in terms of the rationale underlying the concepts of convergent and discriminant validation (Campbell and Fiske, 1959). The procedure followed with respect to discriminant validation is generally used to justify the use of novel trait measures or to establish construct validity. Essen- tially, a test is considered invalidated if it correlates too highly with a test or tests from which it was supposed to differ. In terms of this study, this was construed to mean that, to be discriminantly valid as a trait measure, the correlations found to exist between birth control belief intensity (A) and exhibited complexity (B); education belief intensity (C) and exhibited complexity (D), should be higher than those found to exist between either birth control or education belief intensity and the level of complexity as measured on a general scale (AE and CE). 43 The evidence was interpreted to lend support to the validity of the Schroder techniques as a method of assess- ing level of cognitive-complexity. Except for group three in Table 7b where r .ll9<.205), the correlation CD.Ol). Group two (moderate belief intensity) taken alone did show a signifi- cant relationship with rS = .396 (p<.O5). Although this evidence is weak, it can be interpreted to suggest that more complex individuals are more likely to have attitude domains which intersect each other where there are common or similar stimulus elements. I‘D“. "g -.‘-' 4A The procedure for convergent validation requires that there be two different measures of the structural variable under consideration. In this particular instance, the Schroder measure of complexity was used with reference to two specified stimulus domains. It is contended here that, while the general manual of structural variables was the same for the analysis of both attitude areas considered, it can logically be construed as two different tests because the attitude areas were different. Therefore, if it is shown that the Schroder technique can efficiently discrim- inate between belief levels in more than one attitude domain then the test can be to some extent convergently validated. A strong significant relationship was found to exist between belief intensity in birth control and the level of complexity exhibited on the required task (p<.Ol). Although not as strong (p<.lO), a relationship was found also between belief intensity in education and level of complexity exhibited. Also, correlations between belief intensity in one attitude domain and complexity exhibited with respect to the other domain (AD and CB) are nonsignificant except for group two in Table 7a where the correlation between belief intensity about birth control was significantly related to complexity exhibited toward education (p<.Ol). 45 According to Campbell and Fiske (1959), the low correlations found to exist as noted above can lead to one of three alternative propositions. One of these propositions simply states that perhaps one of the methods used does not measure the trait under investigation. Since it was shown that the Schroder technique was best able to statistically discriminate between levels of cognitive functioning as theoretically defined in this study, and in two separate attitude domains, it is contended that this evidence partially confirms the validity of the Schroder technique for assessing level of cognitive-complexity and suggests that further study be given this area. In summary, the evidence from this study may be inter- preted as supporting the hypothesis of this study that cognitive-complexity is not a general trait and must be assessed with reference to a specified stimulus domain. The use of two theoretically different tests of complexity gave evidence which may be interpreted as favoring the Schroder technique. Evidence also supports the hypothesis that there is a strong relationship between intensity of belief about an attitude object and the level of complexity exhibited toward that object on an essay task. Assessment of Attitudes and Cognitive-Complexity Scores obtained from the attitude scales used in this study were seen to be indicative of the intensity of belief that subjects had about the attitude areas under consideration. 46 By using a highly controversial stimulus domain (birth control) it was possible to select three disparate groups each characteristic of a different intensity level. A basic problem engendered by the use of such scales is that of all the scales in existence, none is recommended for other than group measures (Shaw and Wright, 1967; Fishbein, 1967) and are therefore of little use in pre- dicting individual attitudes. Attitude scales represent an abstraction of the attitude r": construct and all that it implies. An individual generally responds to a scale at a level that is most general, unspe- cific and inconsiderate of situational variables. Careful and extensive qualifications are not allowed by the very response modes generally used by those who construct the scales. Consequently, what is obtained from the administration of an attitude scale is a mean abstraction of the attitude construct representing the total group from which the score was derived. This mean score is generally used to generalize to a larger population from which the test sample has been selected and is for the most part quite misleading as to what it actually implies about the population. The current status of attitude theory, in particular the "school" that views attitudes as learned concepts which are in turn defined as information processing structures (Schroder, et al., 1963) and the measurement “7 techniques available make this assessment problem essen- tially one of both measurement and methodology. The usual measurement techniques such as those develOped by Likert (I932); Thurstone (I929); Osgood (l957); and Guttman (19A?) stress the use of a unidimen- sional scale of favorableness-unfavorableness. An individual is typically located somewhere on the con- tinuum with respect to some attitude. The arguement for using unidimensional scales has generally been one of pragmatism (Fishbein, 1967). Aside from the fact that none of the existing attitude scales is recommended for anything but group measures, two prominent problems are raised by the "traditional" approaches to attitude assessment. The first of these has methodological consider- ations. That both situational and dispositional factors can vary at any given time makes the usual methods of assessment highly unreliable. For example, a person will not necessarily "act out" a social attitude in a strange crowd as often as he might in the confines of his own home and with friends. It becomes necessary to recognize the difference between kinds of behavior when assessing social attitudes. LaPiere (1934) was among the first to point out that a distinction can and should be made between verbal behavior and actual behavior when studying social attitudes. Himmelstrand (1960) and Green (1954) have since stressed the need for this 48 distinction. These writers conclude that there is no necessary relationship between the different kinds of behavior. Peer pressures, religious background, or any number of situational or dispositional variables may cause a person to act much differently in a purely symbolic situation (the usual kind of attitude questionnaire) than he would were he actually involved in the hypothetical— general "situation" the attitude questionnaire represents. Closely connected with the problem of different kinds of behavior is the fact that most individuals answering attitude questionnaires are usually quite aware of the attitude on which they are being tested. Depend- ing on the particular situation, this awareness may indeed bias the responses made. This alone raises the question of whether or not it is ever justifiable to extrapolate from the verbal behavior of respondents to the prediction or "guessing" of their behavior in any situation. What appears to be necessary is the development of a method which will closely approximate or simulate the situation in which a social attitude might be used. For example, if concern were with the functioning of ethnic biases in teachers then, in order to determine to what extent any individual's actions were based on an ethnic bias, it would be useful to construct a situation which would simulate a classroom activity in which the teacher might engage. In a simulated activity a person would “9 have the opportunity to directly express behaviorally those attitudinal biases which he or she felt. And, if properly conducted, the person under investigation would not necessarily be aware that the primary purpose of the simulation was to examine the behavioral manifesta— tions of their attitudes. A possible and seemingly promising approach to the solution of this methodological problem is that offered by the "Teacher's In-Basket" (Shulman, 33 al., 1968). Essentially, the in-basket is a technique which was developed to study the inquiry behavior of indivi- dual teachers-in-training. As in the present study, the subject assumes a role. As a teacher she is placed at a simulated teacher's desk which holds many potential pro- blems in the guise of phone messages, memoranda from faculty, school newsletters, etc. From these materials the subject proceeds to identify problems, or perhaps potential problems and then using the informational resources available to her attempts to work out viable solutions. The various informational materials available concern her new students, access (through an intercom) to a school secretary (an experimenter), a school principal and a "reference memory" (another experimenter). Using these sources the subject, after first identifying a pro- blem, goes about a search for information which will aid in the resolution of the problem. 50 The subject is isolated in the simulated setting and is observed by two trained scorers through a one—way mirror. As the subject proceeds through the inquiry process she must verbalize at all times what she is think- ing. This protocol is recorded for further analysis. Scorers are responsible for monitoring the mani- festations of five variables during the course of the simulation. These five variables are: l) the number of built in potential problems actually identified and taken up by the subject, 2) the number of "bits" of information used by the subject, 3) the sources of information the subject uses, A) the level of understand- ing the subject attains with respect to each problem, and 5) the amount of time the subject gives to inquiry. Scorers are trained to high reliabilities in observing and recording these variables. It has been found using this approach in the study of inquiry that two basically different groups of inquirers can be identified among teachers-in—training. Interesting enough, these two groups roughly correspond in characteristics with what has been termed in this present study the cognitively—complex and the cognitively— simple. Respectively these were called dialectical and didactic in inquiry style. It appears that with the necessary adaptations the in—basket technique may offer a unique approach to the 51 study of attitude manifestations in simulated settings. As with the problem of ethnic biases already discussed, the inclusion of an "ethnic dimension" may result in a different approach to problem resolution by a subject. This in turn may indicate the presence of a bias and suggest experiences which may aid in changing the attitude. In any event, the approach of simulation is closer to actual situations where an attitude might be manifested and is therefore more likely to be reliable in terms of predicting behavior. Further refinement of simulation techniques appears to be an excellent and perhaps rewarding path to explore for those interested in the assessment of attitude functioning. Directly allied to the first problem is that of the measurement techniques used in the traditional approaches to attitude assessment. The unidimensional scale is most common and has been criticized by Abelson (195A- l955) as being inadequate to the complexity of measuring attitudinal structures. As shown above, a more complex assessment of cognitive-functioning is used for the in-basket technique. This is in keeping with a general movement away from the attempt to parsimoniously describe psychological variables as if they were reducible and capable of being so described. Viewed from the theoretical frameworks of 52 Kelly (1955), Harvey, 33’ Bi; (1963) and Schroder, 32 al. (1968), attitudes can be seen as conceptual structures which function through the selection and integration of information from the several informational dimensions usually available to an individual in any given situation. Cognitively speaking, attitudes can therefore be seen as conceptual structures which range from being very simple to being extremely complex in nature. A simple structure can be characterized as being concrete in nature in that an individual is restricted in his judgements which are based on only a few or one dimen- sion of information. When more than one dimension is used the cognitively—simple person has fewer ways in which to integrate the dimensions. This type of struc- ture might be indicated, for example, if an individual classified persons of the basis of skin color without the consideration of other information. On the other hand, a complex attitudinal structure can be characterized by judgements which are made using several dimensions of information and integrating them in a manner which best fits the situation at hand. At the complex level the attitudinal strtcture serves more the function of information searching than simple cate— gorization of that information most prominent. In addition to the assessment technique used by Schulman, et_ al (I968), an additional solution to the measurement problem may be forthcoming in the form of 53 multidimensional scaling techniques. Although they are only in the developmental stages, they have been useful in situations where no assumptions have been made about the number or nature of the informational dimensions an individual uses in making judgements about some group of stimulus objects. Tucker and Messick (1963) have developed a multidimensional scaling model which allows ! the dimensions of information an individual uses in making similarity judgements to be described and spatially mapped. A factor analysis of subjects responses permits a deter- mination of dimensional weights thereby giving some indication of the priorities a person uses in making his judgements. Coupled with a content and process analysis of situational and dispositional variables present in the simulated situation, the multidimensional analysis would offer added information about the cognitive-processes in which individuals engage. .In summary, the current status of attitudinal assessment techniques leads one to search for new methods and ways of measuring which are more relevant and useful to the prediction of behavior. It has been suggested here that the "in-basket" offers a potentially useful method for assessing the functioning of attitudes in simulated settings that approximate conditions where an attitude is likely to be used. It has also been suggested 54 that closer attention be given to multidimensional scaling techniques for analysis of the informational dimensions persons use in making judgements. Educational Implications of Study Teachers are involved in an enterprise which requires them to search for and evaluate an enormous amount of information about their students. This infor- mation is derived from both formal and informal sources. They are expected to search for information relevant to the understanding of each student's peculiar needs and capabilities and their relationship to his per- formance behavior and also to his potential for achieve- ment. This process of selecting relevant information about students and basing decisions upon it is central to the function of teaching. At times, however, certain highly specific and irrelevant information may play an inordinate role in the judgemental processes of teachers. In the case of a teacher who has an attitudinal bias toward a particular ethnic group or groups, the immediately recognizable characteristic of skin color or another phenotypic mark may serve to put the student into an evaluative category which better serves the needs of the teacher than those of the student. That attitudes function, at least in part, to serve the personality needs of the individual 55 has been reported by Katz (1960) and Smith, 33 a1, (1956). Representing many contemporary school critics, Schrag (1968) has leveled the charge of both racial and social biases in many teachers. If it is the case that certain prejudicial biases exist in teachers, whether they be favorable or unfavor- able in nature, it follows that they may be potentially or actually harmful to the students of these teachers. What appears desirable at this point is to identify those teachers during their training who harbor such potentially harmful attitudinal biases and to develop experiences that may precipitate attitude change. It has been known for a long time that attitudes affect the behavior of individual toward attitude objects. This study has shown that the intensity of belief about an attitude object hears some relation to the level of cognitive-complexity an individual will exhibit toward that object. It has also been pointed out that traditional approaches to attitude measurement are useless for the prediction of individual behavior because of instrument limitations. What is here suggested in the adaptation and experimentation with the "in—basket" technique as a method of assessing attitude manifestation in situations which closely approximate, or simulate situations in which an attitude may be used. As this approach has 56 been given some use (Shulman, at 31. 1968) in an educa— tional Setting, it would seem that, as suggested in previous sections, that the inclusion of different informational dimensions which might reflect specific attitudes would give information through the task activities of a subject's biases. Also, it is suggested glans-1'." that attention be given also to multidimensional scaling techniques for gaining extra information about how and * in what manner a person integrates informational - r‘ dimensionsin making judgements. NCES C m F E 2... 57 REFERENCES Abelson, R. P. "A technique and a model for multi- dimensional attitude scaling", Public Opinion Quarterly, Winter l95A-55, pp. 3054318. I.M. Fishbein (Ed-), Readings in Attitude Theory and Nea- surement, New York, Wiley, 1967, pp. 147-156. Allard, M. and Carlson, E. "The generality of cognitive complexity" Journal of Social Psychology, 1963, 59, pp. 73-75. Barron, F. "Complexity-simplicity as a personality dimension" Journal of Abnormal and Social Psychology, 1953. A8, pp? 163-172. . Creativity and Personal Freedom, Princeton, New Jersey: Van Nostrand and Co., 1968. Berkowitz, L. "Leveling tendencies and the complexity- simplicity dimension" Journal of Personality, 1957, 22:: pp- 743-751. Bieri, J. "Cognitive complexity—simplicity and pre- dictive behavior" Journal of Abnormal and Social Psychology, 1955, ii, pp. 263—268. . and Blacker, E. "The generality of cognitive complexity in the perception of people and inblots" Journal of Abnormal and Social Psycholgy, 1956, 5;, pp. 112-117. . "Complexity—simplicity as a personality variable in cognitive and preferential behavior" In D. W. Fiske and S. R. Maddi (Eds) Functions of Varied Experience, Honewood, Illinois: Dorsey Press, Inc., 1961, pp. 355-379. Campbell, D. and Fiske, D. "Convergent and discriminant validation by the multitrait-multimethod matrix" Psychological Bulletin, March, 1959, 56, 2, pp. 81- 105. 58 59 Claunch, N. "Cognitive and motivational characteristics associated with concrete and abstract levels of conceptual complexity" Joing ONR and NIMH report, Princeton University, 1962. Fielder, J. F. "Subjective and objective knowing/not knowing: an interpretation in terms of cognitive— complexity" Michigan State University, 1969, Unpublished Paper. Fishbein, M. "A behavior theory approach to the relations between beliefs about an object and attitude toward the object" In M. Fishbein (Ed.) Readings in Attitude Theory and Measurement, New York: Wiley, 1967, pp. 389-900. Gardner, R. and Schoen, R. "Differentiation and abstration in concept formation" Psychological Monographs, 1962, 16, (Al all of No. 560). Green, B. F. "Attitude measurement" In G. Lindzey. (Ed.), Handbook of Social Psychology, Cambridge, Mass.: Addison-Wesley, 1959, I, Chapter 9. Guttman, L. "The cornell technique for scale and intensity analysis" Educational Psychological Measurement, 19147. 1. pp. 217-279. Harvey, O. J., Hunt, D., and Schroder, H. Conceptual Systems and Personality Organization, New York: Wiley, 1961. Himmelstrand, Ulf. "Verbal attitudes and behavior" Public Opinion Quarterly, 1960, 23, pp. 224-250. Jackson, D. and Messick, S. "Individual differences in social perception" British Journal of social and Clinical Pschology, 1963, 2, pp. 1—10. Katz, D. "The functional approach to the study of attitudes" Public Opinion Quarterly, 1960, 23, pp. 224-250. Kelly, G. A Theory of Personality, New York: Norton, 1963. Kerlinger, F. N. and Kaya, E. "The construction and factor analytic validation of scales to measure attitudes toward education" Educational Psychological Measure- ment, 1959, 19, pp. 13-29. In M. Shaw and J. wright (Eds.) Scales for the Measurement of Attitudes, New York: McGraw—Hill, 1967. 6O IaPiere, R. T. "Attitudes versus actions" Social Forces, 193A, 15, pp. 230-237. In M. Fishbein (Ed.) Readings in Attitude Theory and Measurement, New York: Wiley, 1967, pp. 26-31. Likert, R. A. "A technique for the measurement of attitudes" Archives of Psychology, 1932, 55, pp. 5-55. Marascuilo, L. and McSweeney, M. "Nonparametric post hoc comparisons for trend" Psychological Bulletin, Messick, S. and Kogan, N. "Personality consistencies in judgement: dimensions and role constructs" Multi— variate Behavioral Research, April, 1966, pp. 165-175. Osgood, C., Suci, G., and Tannenbaum, P. The Measurement of Measurement of Meaning, Urbana: University of Illinois Press, 1957. Rhine, R. J. "A concept-formation approach to attitude acquisition" Psychological Review, 1958, 55, pp. 362-370. In M. Fishbein, (Ed.) Readings in Attitude Theory and Measurement, New York: ’Wiley,fl967, pp. 382-3881 Rokeach, M. Beliefs, Attitudes and Values, San Francisco: Jossey-Bass, 1968. Schrag, P. "Why our schools have failed" Commentary, Schroder, H. and Phares, J. Untitled. Unpublished manu- script, Princeton University, 1965. (Summary in Schroder, g: 5;}, Human Information Processing, New York: Holt, Rinehart and Winston, 1968, 186, 200. . Driver, M., and Struefert, S. Human Information Processing, New York: Holt, Rinehart and Winston. '1968. Scott, W. A. "Cognitive complexity and cognitive flexi- bility" Sociometry, 1962, 55, pp. AO5-hlu. . "Conceptualizing and measuring structural pro- perties of cognition" In Harvey, O. J. (Ed.) Motivation and Social Interaction, New York: Ronald Press, 1963b, pp. 266;288. . "Cognitive complexity and cognitive balance" Sociometry, 1963a, 25, pp. 66-74. 61 Siegel, S. Nonparametric Statistics, New York: McGraw- Hill, 1956. Shaw, M. and wright, M. Scales for the Measurement of Attitudes, New York: ‘MCGraw-Hill, 1967. Shulman, L., Loupe, M., and Piper R. Studies of the Inquiry Process, Final Report, Office of'Education_Pr03ect NumEer 5-50597, July 1968. Smith, M. B., Bruner, J. S., and White, R. W. Opinions and Personality, New York: Wiley, 1956. Thurstone, "The measurement of social attitudes" Journal of Abnormal and Social Psychology, 1931, 55, pp. 2149-2690 Tucker, L. and Messick, S. "An individual differences scaling model for multidimensional scaling" Pscho- metrika, Dec. 1963, 55, No. A, pp. 333-367. Vannoy, J. S. "Generality of cognitive complexity- simplicity as a personality construct" Journal of gggsonality and Social Psychology, 1965, 3, pp. 385- Wilke, W. H. "An experimental comparison of the speech, the radio, and the printed page as propoganda devices" Archives of Psychology, 1934, No. 169. In M. Shaw and—Jl—Wright (Eds.), Scales for the Measurement of Attitudes, New York: McGraw—Hill;“1967. 62 3'“ APPENDIX A BIRTH CONTROL SCALE 53 ll. l2. l3. 1“. BIRTH CONTROL SCALE We should be absolutely opposed to birth control. Birth control reduces the marital relation to the level of vice. We ought to approve of birth control because of the advantages to women's health resulting from the correct spacing of children. Uncontrolled reproduction should be opposed on the grounds that it is a fundamental cause of crime. We should not approve of women taking the health risks involved in birth control. The present depression and the attendent problem of unemployment makes more desirable than ever the gen- eral approval of birth control. Wide-spread acceptance and approval of birth control is imperative. The practice of birth control is equivalent to murder. Birth control is race suicide. Uncontrolled reproduction should be discouraged since it leads to many social evils. Birth control is a legitimate health measure. Our laws should prohibit giving, even to adults, infor— mation concerning birth control. We should not only allow but strongly urge birth con- trol to limit the size of families of low income. Effective measures should be taken to prevent any sale of birth control devices. 15. l6. 19. 20. 21. 22. 65 Birth control would help to solve many of our social problems. Birth control is highly desirable for women who must earn a living. The possible benefits of bir rth control do not alter the fact that it is morally wrong. Wide-spread knowledge of birth control methods should be opposed as likely to lead to the spread of social disease. We should approve as socially desirable the program of those organizations supporting the movement for birth control. The practice of birth control evades man's duty to propagate the race. Birth control increases the happiness of married life. Every normal healthy couple should have as many children as is physiologically possible. * These items are against birth control and weights for their response alternatives must be reversed for scoring. The same response categories are used for all items. EDUCATION SCALE 66 10. ll. 12. EDUCATION SCALE The goals of education should be dictated by child- ren's interests and needs, as well as by the larger demands of society. No subject is more important than the personalities of the pupils. Schools of today are neglecting the three R's. The pupil—teacher relationship is the relationship between a child who needs direction, guidance, and control and a teacher who is an expert supplying direction, guidance, and control. Teachers, like university professors, should have academic freedom-—-freedom to teach what they think is right and best. The backbone of the school curriculum is subject matter; activities are useful mainly to facilitate the learning of subject matter. Teachers should encourage pupils to study and crit- icize our own and other economic systems and practices. The traditional moral standards of our children should not just be accepted; they should be examined and tested in solving the present problems of students. Learning is experimental; the child should be taught to test alternatives before accepting any of them. The curriculum consists of subject matter to be learned and skills to be acquired. The true view of education is so arranging learning that the child gradually builds up a store house of knowledge that he can use in the future. One of the big difficulties with modern schools is that discipline is often sacrificed to the interests of children. 67 l3. l6. 170 190 20. 68 The curriculum should contain an orderly arrangement of subjects that represent the best of our cultural heritage. Discipline should be governed by long—range interests and well—established standards. Education and educational institutions must be sources of new social ideas; education must be a social program undergoing continual reconstruction. Right from the very first grade, teachers must teach the child at his own level and not at the level of the grade he is in. Children should be allowed more freedom than they usually get in the execution of learning activities. Children need and should have more supervision and discipline than they usually get. Learning is essentially a process of increasing one's store of information about the various fields of knowledge. In a democracy, teachers should help students under- stand not only the meaning of democracy but also the meaning of the ideologies of other political systems. APPENDIX C SCALE POINTS USED FOR ASSESSING THE CONCEPTUAL LEVEL INVOLVED IN ESSAY WRITING 69 SCALE POINTS USED FOR ASSESSING THE CONCEPTUAL LEVEL INVOLVED IN ESSAY WRITING* Scale Point Description 1 Presents only one side of a problem ignores differences, similarities, and gradations. 2 One side of the problem presented and supported much more fully than the other. Opposing views perceived as compartmentalized or negative. No interrelationships considered. 3 Two or more views clearly different- iated. Similarities and differences implied or presented. One view can be opposed, but it is understood. A Includes all involved under scale point 3 but begins to "consider" the similar- ities and differences between views. At this level, consideration is expressed . . . .as qualifications of each . . . (for example, "similar, but . . . ."). That is, the simultaneous effects of alternate views becomes apparent in the writer's thinking. 5 Considers alternate and conflicting reasons for perceived similarities and differences between views in pro— ducing the essay. 6 Begins to consider relationships, not only among direct similarities and dif- ferences between sides of the problem, but also relationships between alter- nate reasons as to why the differences and similarities occur. 70 71 Scale Point Description 7 The consideration of notions which include relational linkages between alternate views. Such notions are open to all conflicting components and express attempts to see these as parts of a more inclusive "con- struction" of the problem. * Taken from Schroder, g: 5;. Human Information Processing, New York: Holt, Rinehart and Winston, Inc., 1968, p. 201. APPENDIX D INSTRUCTIONS AND LETTERS FOR TEST OF COGNITIVE-COMPLEXITY 72 INSTRUCTIONS This booklet is composed of three parts. Carefully read the instructions for each part before beginning. The times given for completing your answers are not meant to be a constaint. They are only suggestions based on pre- F vious tests of this kind. Feel free to take more time if necessary. PART ONE: completion time 20 to 25 minutes For the purposes of part one you are asked to assume the role of a prominent physician whose specialty is the study of birth—related problems. Your task is to respond to a letter from a young couple seeking advice about birth con- trol. Respond in an "Ann Landers" fashion giving your advice as a person of knowledge and authority. If you have any questions please ask them before beginning. Please turn page an‘ 9 O m 09 H :5 73 7A 7 July 1969 S. L. Sulzberg, M.D. 517 National Bldg. F New York Medical Center ' New York, New York Dear Dr. Sulzberg, My fiance and I are to be married in the very near future and would greatly appreciate some advice from you regard- ing birth control. We would like not to have any children right away for both financial and personal reasons. However, we have not been able to decide whether it is acceptable for us in a moral sense to use artificial means for preventing conception. Because of our religious background the decision is quite difficult. We think it would help us a great deal if you would write giving us some advice as to how you think we should proceed in resolving our dilemma. Thank you for your time. Sincerely yours, Mary Diffel 8341 Eight Mile Rd. Detroit, Michigan 75 PART TWO: Completion time 20 to 25 minutes For the purposes of part two you are asked to assume the role of a prominent educator. Your task is to respond to a letter from a parent concerned with placing her child in a particular school. Respond in an "Ann Landers" fashion giving your advice as an educator of knowledge and authority. If you have any questions please ask them before beginning. PLEASE TURN PAGE AND BEGIN. 76 7 July 1969 Professor H. J. Laurence School of Education F Harvard University Cambridge, Mass. Dear Professor Laurence, i My husband and I are currently in the process of moving to a new city. We have found two houses located in this city which are situated in two different neighborhoods. Both houses are equally acceptable to us. However, these houses are located in two different school districts and one of the districts is much more progress— ive and permissive in its attitudes toward education than the other. The progressive school district feels that learning is experimental and that children should be taught to test alternatives before accepting any of them. They also believe that children should be allowed more freedom than they normally get in the execution of learning activities. The other school district feels that children need and should have more supervision and discipline. We are unable to decide whether or not it would be better for our child to attend the school located in the more progressive district. Consequently, we would appreciate your advice concerning what we should do about our decision. Thank you. Sincerely yours, Mrs. Howard Greene, 2218 Appel Ct. E. Lansing, Michigan APPENDIX E BARRON SCALE OF COMPLEXITY 77 PART THREE: Completion time 5 to 10 minutes Part three is a questionnaire composed of 30 statements which you are asked to agree or disagree. For each state- ment, respond according to the following key: (1) TRUE (2) FALSE Use the attached IBM score sheet for your responses. BEGIN WITH NUMBER 61. l. I like to have a place for everything and everything in its place. 2. Some of my friends think that my ideas are impractical, if not a bit wild. 3. I don't like to undertake any project unless I have a pretty good idea how it will turn out. A. For most questions there is just one right answer, once a person is able to get all the facts. 5. Politically I am_probably something of a radical. 6. Perfect balance is the essence of all good composition. 7. I prefer to engage in activities from which I can see definite results rather than those from which no tang- ible or objective results are apparent. 8. I find that a well-ordered mode of life with regular hours is not congenial to my temperament. 9. The unfinished and the imperfect of ten have greater appeal for me that the completed and the polished. .10. I like to listen to primitive music. lJ.. I have always had goals and ambitions that were impractical or that seemed impossible for me to realize. 312. When a teacher lectures on something other than what he originally announced, I feel uneasy. 78 79 13. Trends toward abstractionism and the distortion of reality have corrupted much art of recent years. 1A. It bothers me to have different news commentators give different interpretations of the news. 15. I like to fool around with new ideas, even if they turn out later to have been a total waste of time. 16. I don't like to work on a problem unless there is a possibility of coming out with a clear-cut unambiguous answer. ' 17. I have always hated regulations. 18. Many of my friends would probably be considered unconventional by other people. 19. It doesn't bother me when things are uncertain and unpredictable. 20. My way of doing things is apt to be misunderstood by others. 21. Facts appeal to me more than ideas. 22. I have had strange and peculiar thoughts. 23. I don't like things to be uncertain and unpredictable. 2A. The worst thing an instructor can do is to make very specific plans for each lesson. 25. It is a good rule to accept nothing as certain or proved. 26. I dislike following a set schedule. 27. Usually, I prefer known ways of doing things rather than trying out new ways. 28. I like to go alone to visit new and strange places. 29. I much prefer friends who are pleasant to have around to those who are always involved in some difficult problem. 30. I have had very peculiar and strange experiences. APPENDIX F DESCRIPTION OF EXPERIMENT GIVEN TO EDUCATION 200 STUDENTS, SUMMER 1969 80 TO: Ed. 200 Students, SS'69 FROM: J. Fielder RE: The following is a brief description of the experi- ment in which most of you participated either once or twice. If you are interested in a more detailed description or have any questions about the project you want answered, feel free to see me at AOlG Erickson Hall. Regardless of whether it was for altruistic or financial reasons, I greatly appreciate the response of those students who volunteered for the second half of the experiment. The purpose of this study is to extend the range of con- venience of the concept of cognitive-complexity. Gen— erally speaking, the concept of cognitive-complexity is used to indicate the manner in which a person perceives his social world. The cognitively-simple person has been characterized as being rigid, categorical and holding rela- tively extreme attitudes. When making judgements about any social situation, this person tends to examine only a few dimensions of the problem and rarely seeks additional information. In terms of the research on inquiry done by Shulman, Loupe and Piper (1968), the cognitively simple teacher might be deemed "didactic" in inquiry style. In contrast, the cognitively-complex person uses more dimensions of information when making judgements, has more ways of integrating the dimensions, and is more ‘abstract. Again in terms of the above mentioned study, the cognitively—complex teacher might be deemed as being "dialectical" inquiry style. Most studies of cognitive-complexity have tended to search for a unitary trait which affects all of a person's per- ceptions in the same manner. Recent research has shown, however, that it is unlikely that there is such a unitary trait of cognitive-complexity. One important theoretician (Schroder, 1968) has stated that to assess level of cognitive- complexity, one must direct measurements to the stimulus domain under consideration. His assertion is that cognitive- complexity of teachers in perceiving various ethnic groups 81 82 measurements would have to be directed to the specific ethnic group of interest. The thesis of this study is that the attitude scale offers an efficient, if not potential, way of predicting the complexity of a person's perception toward any social situation. It is theorized that intensity of belief determines one's perception of a social situation. Since an attitude is defined as an organization of beliefs about an attitude object (Rokeach, 1968), it is feasible to hypothesize that scores on an attitude scale will allow us to predict level of cognitive—complexity regarding the stimulus domain the attitude scale represents. This study is important since we know that teachers may categorize or stereotype students on the basis of an easily recognizable phenotypic characteristic such as skin color. It would be of great value if procedures could be developed which would allow for identification of individual teachers-in-training in which an ethnic bias exist, and most important, affects their judgements about students. To test the thesis of this study, it is necessary to have subjects who evidence varying degrees of belief intensity on some attitude scale. Therefore, all of the approximately 300 students enrolled in ED. 200 will be initially tested as to intensity of belief about some social attitude object. On the basis of scores obtained from the attitude scale, approximately 60 subjects will be chosen for further experi— mentation. There will be three groups each composed of 20 subjects. The first group will be composed of indi- viduals who exhibited a high amount of belief in agreement with the attitude object. The second group will be com- posed of individuals who exhibited a moderate amount of belief toward the attitude object. And lastly, the third group will be composed of individuals exhibiting a high amount of belief in disagreement with the attitude object. To test for complexity toward the stimulus domain the attitude object represents, each subject will be asked to perform a task. The task in this instance will be the writing of a letter in an "Ann Landers" fashion in response to a query of the same kind. This letter(s) will reflect the stimulus domain of the attitude scale previously admin- istered. Results are expected to show that there is a strong relationship between intensity of belief about an attitude 83 object and the level of cognitive-complexity exhibited by individuals in performing a task requiring them to deal with the stimulus domain the attitude scale represents. If indeed this prediction is correct, future research can be conducted in a manner that will allow for the develop- ment of more sophisticated methods of assessing intensity of belief and level of cognitive—complexity toward a myriad of social situations, especially in the classroom. APPENDIX G ATTITUDE SCALES AND FILLER ITEMS 811 USING THE KEY PROVIDED BELOW PLEASE INDICATE YOUR OPINION ON EACH OF THE FOLLOWING STATEMENTS (Numbers 1—60). USE THE ATTACHED ANSWER SHEET TO INDICATE YOUR OPINION. strong moderate slight slight moderate strong agree agree agree disagree. disagree disagree (1) (2) (3) (A) (5) (6) l. The present over population and the attendant problem of unemployment makes more desirable than ever the ‘ general approval of birth control. 2. Law controls the conduct of citizens. 3. Practical considerations should come first, beauty second. A. The possible benefits of birth control do not alter the fact that it is morally wrong. 5. One of the big difficulties with modern schools is that discipline is often sacrificed to the interests of children. 6. Wide-spread acceptance and approval of birth control is imperative. , 7. Teachers, like university professors, should have academic freedom.....freedom to teach what they think is right and best. 8. The most effective teaching is by private tutoring. 9. Discipline should be governed by long-range interests and well-established standards. 10. Regardless of sex, there should be equal pay for equal work. 11. The true view of education is so arranging learning that the child gradually builds up a store house of knowledge that he can use in the future. 12. Birth control is a legitimate health measure. 85 13. 1A. 15. l6. 17. 18. 19. 20. 21. 22. 23. 2A. 25. 26. 27. 28. 29. 86 We should approve as socially desirable the program of those organizations supporting the movement for birth control. Learning is experimental; the child should be taught to test alternatives before accepting any of them. We should be absolutely opposed to birth control. The practice of.birth control is equivalent to murder. We should not approve of women taking the health risks involved in birth control. Most young people are getting too much education. Children need and should have more supervision and discipline than they usually get. Young people should be able to use their leisure time as they please. The traditional moral standards of our children should not just be accepted; they should be examined and tested in solving the present problems of students. The pupil-teacher relationship is the relationship between a child who needs direction, guidance, and control and a teacher who is an expert supplying direction, guidance and control. Effective measures should be taken to prevent any sale of birth control devices. Children should be allowed more freedom than they usually get in the execution of learning activities. The goals of education should be dictated by children's interests and needs, as well as by the larger demands of society. The curriculum consists of subject matter to be learned and skills to be acquired. Birth control reduces the marital relation to the level of vice. Teachers are the molders of society. No subject is more important than the personalities of the students. 30. 31. 32. 33. 3A. 35. 36. 37. 38. 39. A0. A1. A2. A3. AA. “5. A6. A7. A8. 87 The practice of birth control evades man's duty to propagate the race. The functioning of law results in the prevention of behavior harmful to others. Every normal healthy couple should have as many children as is physiologically possible. Birth control is race suicide. Most great fortunes are made honestly. One the whole, our economic system is just and wise. A young person should be restricted only when he is infringing upon the rights of others. Schools of today are neglecting the three R's. Birth control would help to solve many of our social problems. Teachers should encourage pupils to study and criticize our own and other economic systems and practices. The backbone of the school curriculum is subject matter; activities are useful mainly to facilitate the learning of subject matter. We should not only allow but strongly urge birth control to limit the size of families of low income. Education in artistic things is a waste of public funds. It is doubtful whether education has improved the world or not. Uncontrolled reproduction should be discouraged since it leads to many social evils. Uncontrolled reproduction should be opposed on the grounds that it is a fundamental cause of crime. Birth control increases the happiness of married life. War is necessary to avoid overpopulation of nations. Our laws should prohibit giving, even to adults, information concerning birth control. ,m: ‘uurunun- A9. 50. 51. .52. 53. 5A. 55. 56. 57. 58. 59. 60. 88 International disputes should be settled without war. The curriculum should contain an orderly arrangement of subjects that represent the best of our cultural heritage. We ought to approve of birth control because of the advantages to women's health resulting from the correct spacing of children. Education and educational institutions must be sources of new social ideas; education must be a social program undergoing continual reconstruction. Right from the first grade, teachers must teach the child at his own level and not at the level of the grade he is in. Participation in intensive competition develops leadership. Learning is essentially a process of increasing one's store of information about the various fields of knowledge. Wide—spread knowledge of birth control methods should be opposed as likely to lead to the spread of social disease. I believe teaching is the most interesting of the professions. In a democracy, teachers should help students under- stand not only the meaning of democracy but also the meaning of the ideologies of other political systems. Young people should obey their parents because they are their parents. Birth control is highly desirable for women who must earn a living. PLEASE MAKE CERTAIN YOUR NAME, STUDENT NO., SUMMER CAMPUS ADDRESS, AND SUMMER PHONE NUMBER ARE ON THE ANSWER SHEET. THANK YOU. APPENDIX H LETTER OF SELECTED SUBJECTS 89 Dear On June 30th, you were 90 7 July 1969 a participant in the first part of a research project focusing on the teachers-in- training in Education 200. to participate as a paid volunteer in the second half of this project. For one hour of your time, at your convenience, you will be paid $2.00. If you wish to participate, please come at one of the following times and places. July 1A 5 July 15 5 July 16 10 July 17 10 July 18 10 If none of these times special arrangements. Again, this project will only require one hour of your time for which you will be compensated. that you participate since this research project deals p.m. - 9 p.m. A01 p.m. — 9 p.m. A01 a.m. - 3 p.m. A52 a.m. - 3 p.m. A52 a.m. - 3 p.m. A52 Erickson Erickson Erickson Erickson Erickson >3>§I>QQ are convenient and you wish to participate, please phone me at the number below to make with the problem of training teachers and cannot be completed without your help. Thank you. Sincerely yours, John F. Fielder, Project Director A01 G Erickson Hall 353-3798 This letter is an invitation Hall Hall Hall Hall Hall It is important "AAAAAAAAI