,. w‘... A MATRIX OF SOCIAL AND PERSONALITY VARIABLES FOR THE PREDICTION OF SCHOOL ACHIEVEMENT Thesis‘for the Degree of M. A. _ MICHIGAN STATE. UNIVERSITY RICHARD J. GIG-LIOTTI 1,969 ABSTRACT A MATRIX OF SOCIAL AND PERSONALITY VARIABLES FOR THE PREDICTION OF SCHOOL ACHIEVEMENT BY Richard J. Gigliotti This thesis is concerned with those personality and social variables which influence the level of achievement of high school students. The problem centers around the iden- tifying of these variables and the development of a matrix of these variables to predict the level of achievement which can be expected when certain combinations of these variables exist together. Two of the variables were the self concept of academic ability (SCAA) which a student has and; the importance to the student of his self identity student (ISIS). These were identified as the personality variables. Each of these varia- bles was developed by means of a Guttman scale. The other two variables were called the academic expec- tation level (AEL) and the degree of consensus. These were identified as the social variables. The AEL is a weighted sum of perceived parental, friends and teacher expectations. The degree of consensus is an index computed on the amount of consensus which existed about expectations among these three significant others. A high and low division was computed for each of these variables which resulted in sixteen possible Richard J. Glgliottl combinations, each of which was hypothesized to be related to a certain level of achievement. Each of these variables alone was found to be signifi- cantly correlated with school achievement (G.P.A.) When combined into a matrix, it was found that significant chi- square values were achieved for combinations where either high or low achievement was hypothesized. However, for those com- binations where moderate or moderate to low achievement was hypothesized, the results were for the most part in the desired direction, but non-significant. A MATRIX OF SOCIAL AND PERSONALITY VARIABLES FOR THE PREDICTION OF SCHOOL ACHIEVEMENT BY Richard J. Gigliotti A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Sociology 1969 DEDICATION This thesis is dedicated to all the unnamed, many forgetten, significants in my academic role set, both past and present, who have maintained high expectations. have upheld them. 11 I hope I ACKNOWLEDGMENTS I would like to acknowledge Professor Wilbur B. Brookover who as a teacher, research director, and friend, first intro- duced me to problems of concern in this thesis. The patience which he has demonstrated during my work with him is of the first order. I would also like to thank Dr. Denton horrison and Pro- fessor William Faunce for their efforts in reading this thesis and making critical suggestions on it. Finally I want to thank Carol C. Gigliotti who made my task a little easier both by typing the final draft of this thesis, and by being a positive source of encouragement. iii List of Section Section Section Section Section Section Section List of TABLE OF CONTENTS tables 1 - Problem 2 - Theory and Literature 3 - Definitions, Assumptions, and Opera- tionalizing of concepts - Hypotheses 4 5 - Analysis 6 - Discussion of Results and Conclusions 7 - Questionnaire references 13 21 39 41+ 54 LIST OF TABLES Table l - Hypothetical table for prediction of academic achievement level (SPA) 16 Table 2 - Pearson Product Moment Correlation between SCAA 10th grade and GPA 10th grade. 22 Table 3 - Pearson Product homent Correlation between AEL 10th grade and GPA 10th grade. 22 Table 4 - Pearson Product Moment Correlation between x expectations 10th grade and GPA 10th grade. f3 Table 5 - Pearson Product Koment Correlation between ISIS 10th grade and GPA 10th grade. 24 Table 6 - Range of scores for each variable in both sexes, including N's. 25 Table 7 - X2 test for association between high SCAA plus high ISIS plus high AEL with high consen- sus and GPA. :7 0 Table 8 - XL test for association between high SCAA plus high ISIS plus high AEL with a low Consen- sus and GPA. F8 Table 9 - X2 test for association between low SCAA plus high ISIS plus high AEL, with high consensus and GPA. 29 Table]£)- X2 test for association between low SCAA plus high ISIS plus high AEL with low consensus and GPA. :0 Table]J.- X2 test for association between high SCAA plus high ISIS plus low AEL with high consensus and GPA. 31 TableIJI- X2 test for association between high SCAA plus high ISIS plus low AEL with low consensus and GPA. :2 Tablelfii- X2 test for association between low SCAA plus high ISIS plus low AEL with high consensus V Table Table Table Table Table Table Table Table Table Table Table 14 15 16 17 20 X2 test for association plus high ISIS plus low and GPA. X2 test for association plus low ISIS plus high and GPA. ' X2 test for association plus low ISIS plus high and GPA. X2 test for association plus low ISIS plus high and GPA. X2 test for association between low SCAA AEL plus low consensus °3 J between high SCAA AEL with high consensus as between high SCAA AEL with low consensus between low SCAA AEL with high consensus low ISIS plus high AEL with low consensus and GPA. X2 test for association plus low ISIS plus low AEL with high consensus and GPA. X2 test for association low ISIS plus low AEL with a low consensus and GPA o X2 test for association low ISIS plus low AEL with a high consensus and GPA. X2 test for association 35 between low SCAA plus 36 between high SCAA 36 between high SCAA plus 37 between low SCAA plus low ISIS plus low AEL with low consensus and GPA. Hypotheses accepted, with n's, pendent variab%es and GPA predicted - ordered value to lowest. from highest X Hypotheses rejected, with n's, four inde- pendent variabfies and GPA predicted — ordered value to lowest. from highest X vi ,8 between low SCAA plus 38 four inde- IO 41 PROBLEM The question of why students achieve differentially in school has intrigued researchers and theorists for many years. 'As with other problems in social science, this too cannot be solved by any monistic theory, although We oftentimes will hear it said that student A performs better in school than student B because he is "smarter” or in other words his I.Q. is higher. To rely on such a factor as the sole answer is of course absurd. I.Q. is singled out because it is the most popular explanation. Just what an intelligence test measures has come into serious question in recent years. At one time it was believed that they measured something close to a na« tive capacity or potential. Today it is believed that they probably measure some particular kind of performance which is neither innate or constant. 1 This suggests the possibil- ity that there might not be complete independence between one's I.Q. score and his achievement in school. The argument how- ever, is against trying to present one explanation for difi- erential achievement, and in this case there shouldn't be any question that it hasn't been accomplished to date, and may never be. In the case of I.Q., research has indeed shown 1. Carl W. Backman and Paul F. Secord, A Social Psycholog- ical View gszducation, p. 29. 2 that there is a high correlation between I.Q. score and grade achievement, but that this correlation declines significantly when other variables are partialled out.? Once we have eliminated monistic theories, as we must of course do, we open up a Pandora's box of influencing fact- ors. Attention has also been directed to the importance of social structural variables. The socio-economic status (SES) ‘of the student's family has been explored to gain insight into the influencing factors of academic achievement. Implicit in such an attempt is the idea that SES reflects the value system of the family which is influenced by the type of entertainment and recreation participated in, the type of people with which they associate, the type of occupation engaged in, and so on. Other researchers have expanded this orientation and studiei the general SE3 composition of the school, and even of the neighborhooi.3 Research such as this has its recognized merits as has been shown by the recent attempts to eliminate neighborhooi school districts which foster perpetuation of poor academic 1 performance in low SES neighborhood schools.L The problem with 7. Brookover, Wilbur B. et. al., Self Concept 23 Ability and School Achievement, III, p. 100, found that I.Q. score and GPA have a zero order correlation of 0.52 at the llth grade level, but when one's self concept of academic ability score is partialled out, the correlation is reduced to 0.99. 3. For a recent example of this type of orientation see Robert E. Herriott and Nancy Hoyt St. John, Social Class and the Urban School. A. For a discussion of this whole area see U.S. Commiss on on Civil Rights, Racial Isolation in Public Schools also United Stated District Court for District of Columbia Civil Action No. 82-86, 196". 3 much of this research however is that it is not explanatory enough. There are too many exceptions to hypotheses such as ”individuals who come from families which place a high value on academic achievement will be high academic achievers." What we need to do is identify and isolate those social and personality variables which influence an individual in his role as a student and therefore influence his academic performance level. The orientation of this thesis is that a consideration of ”innate intelligence," "fixed learning ability," "intell- ectual capacity“ or what have you, is irrelevant to success- ful explanation of academic achievement as it is formulated T A. here. an in accord with Frookover and Associates when they say that "although such fixed learning ability may exist we have no way of identifying the limits of most students learn- ing potential at this time."5 We do however know that extreme ranges of objective achievement occur within very narrow ranges of I.~. scorer. What are the causes of this pheonomenon? It is postulated that the interaction of the individual with his environment is a crucial area of exploration in attempt- ing to arrive at a satisfactory explanation. 5. Brookover, et. at., 23. cit. p. 3 HE RY AND LITERATURE The concepts of "significant other." "referent other.” and "self“ are crucial for this thesis. As initially formu- lated by G. H. head, the concept "self" refers to the idea that the individual experiences himself not directly but in- directly through interaction with significant others. In this sense the individual becomes an object which he can evaluate in the development of his ”self," by taking into account the attitudes which he perceives others have of him. The importance of "self" as head formulates it is that it ”is essentially a social structure, and it arises in social experiences."6 A "self.” or conception of self can only arise through interaction or "communication" with others. Wilbur Brookover, in the attempt to utilize the concepts of ”significant other" and "self" in research, has referred to a "significant other" as a real or imaginary person who influences the individual's beliefs about himself. In they context then of a particular role-set, those relevant indiv- iduals who are significant for ego. will play an important part in determining how ego conceives of his ability and performance in the particular role. Brookover notes that 6. George Herbert Mead, Mind. Self and Society. p. 1&0. u 5 ego will tend to behave in terms of what he perceives others expect of him and in terms of how he perceives they evaluate his behavior. The crucial variable here is ego's percep- tion of the significant other's expectations and evaluations. That the perception of expectations is important in influen- in; a student's performance, is illustrated by the results of an experiment conducted by Rosenthal and Jacobson.7 They admin- istered Flanigan's nonverbal intelligence test to all children in an elementary school and disguised it as a test designed to predict academic "blooming" or intellectual gain. There were six grades with three classrooms for each grade. In each of the the eighteen classes an average of twenty percent of the children were assigned to the experimental condition by means of a table of random numbers. The names of these children were given to each teacher and they were told that the scores for these children on the "test for intellectual blooming" indicated that they would show un- usual intellectual gains during the academic year. Eight months after the experimental conditions were instituted, all the child- ren were retested with the same I.Q. test and a change score was computed for each child. The results showed that for the school as a whole, those children from whom the teachers had been led to expect great intellectual gain, showed a significantly greater gain in I.Q. score and school performance than did the control group. This was particularly evident in the very early years 7. Robert Bosenthal and Lenore Jacobson Psychological Reports Vol. 19., 1966. ' 6 where teacher attitudes about the ability of specific child- ren were not yet well formed. The variable of expectations needs further exploration by taking into consideration the differential expectations of significants as well as the varying influences of significant others on ego. In this respect then, it is also necessary to consider the degree of consensus about expectations in the role set. It seems logical to assume that the probability of ego achieving all A's in school would be higher if he perceived all of his significants as expecting him to perform at that level then if only a few of them did. In the former, the expectations of each significant are reinforced by the others in the role set. In the latter, there is a degree of dissensus which must be re- solved in some way. The personality variables which are considered to be import- ant in influencing the level of academic achievement are the self concept which the individual has of his academic ability, and the importance which he attaches to his self identity as a student. Brookover hypothesizes that for the expectations of others to be functional in a particular individual's behavior, they must be internalized and become part of ego's conception of him— self. Thus, in Brookover's conception, ego's self concept of academic ability is an intervening variable which is a necessary but not sufficient condition for determining academic achieve- ment. Ego's self concept of academic ability functions to limit the quality and quantity of learning attempted. He hypothesizes 7 that "to the extent that a society is increasingly character- ized by students who have learned that they all ought to acquire high levels of academic skill, but that only a few of them are so capable, an increasing association may be observed between academic achievement and self concept of academic ability."8 The fourth variable of concern and the second personality variable is that of the importance of the identity, student, to ego. Faunce9 proposes a tendency towards isomorphism be- tween the personality structure and the social structure. He holds that in a society characterized by an immense number of vertically and horizontally stratified roles one will find personality structures which are similar. That is, an indiv- idual will have many roles, and many self identities. The self identities will be vertically ordered in terms of impor- tance to the individual's self-esteem maintenance process. Therefore those identities which an individual regards as im- portant will be highly conscious and will receive considerable investment of self. It will be those identities which are most important to the individual which will receive the most atten- tion in terms of time and ego investment. Therefore, the indiv— idual who regards his self identity of a student as being import- ant theoretically will consciously strive to perform at a level which will maintain a high self esteem. It is, however, entirely possible for an individual to attach a low importance to his self 8. Brookover, et. al., p. 13 9. Lecture notes from a graduate course in "Social Structure and Personality" taught by Faunce in the Winter term of 1968 at Michigan State University. 8 identity student and still perform at a high level. The fact that an individual may have a high self con- cept of academic ability, that is that he feels he has high ability to achieve academically, does not necessarily mean that he also regards his student identity as important. Theoretically, an individual may have the conception of him- self as one with high academic ability but perform poorly because he does not regard his student identity as important and therefore does not invest time and effort. DEFINITIONS, ASSUMPTIONS, AND OPERATIONALIZING OF CONCEPTS The four variables which I have isolated as being the most important in influencing the level of academic achieve- ment are: l. The Academic Expectation Level defined as the combined and weighted expectation level which ego perceives his significant academic others hold for him. 2. The Degree of Consensus defined as the degree of consensus among the units of the role set regarding the perceived expecta- tions for ego. The units are defined as each role occupant in the role set. 3. Self Concept of Academic Ability defined as ego's con- ception of the level at which he is able to perform academi- cally. h. Importance of the Self Identity Student defined as the relative degree of investment placed in the identity student, for self esteem maintenance. ' The assumptions accepted for this thesis are: 1. That ego has no physiological or emotional problems which will inhibit his ability to learn under normal classroom procedures. 2. That the orientation or value in the American society is to high academic achievement. By this I mean that the signif- 9 lO cant others do not value low academic achievement. It does not imply that expectations may not be low. 3. That individuals strive to maintain a favorable self image. The model originally intended for this thesis could not be strictly used because the data which are needed were not collected. Instead, an existing pool of data was used. This data approximates the data needed for the model, enough so that it could be used, however it is anticipated that distor- tions will occur. Envisioned was a role set with ego as the central role occupant. In the role set there were to be six individuals whom ego would identify as being most significant for him in his role as a student. Instead, data is available for only three significants, parents, who are considered as one, friend and teacher. What was envisioned was a method of determining the relative degree of influence which each significant had on ego. This was to be done by forcing ego to make a judge- ment on several different dimensions. For each identified significant other it was intended to have ego answer several questions in the attempt tc rank order and hypothetically weight the influence of each significant. In so doing it was hoped that we could approach the influence of the respective expectations on a weighted basis. That is, an influence fac- tor for each significant was to be developed and this was to be used to weight the perceived expectations for that signif- icant. By approaching the problem this way it was hoped that we could arrive at a more accurate index of the influence which ll expectations have on grades. In addition we would be able to use one expectation score which considered the whole role set, instead of considering several different expectations. As was mentioned earlier, the data which I had would not allow me to do this. Instead, I was able to take the ex- pectation score for each of the three significants and weight it according to a crude indicator of influence which was oun— strued out of research findings by Brookover. Parents' ex- pectations were given a weight of 1, friends were given a weight of .10, and teachers were given a weight of .HO. These combined, weighted, expectation scores yielded a single score which I call the Academic Expectation Level of ego's role set. The second variable, that of the "degree of consensus " was derived from the first by calculating a consensus factor from the perceived grade expectations of each significant. This was done by calculating the mean of the three expectation scores and determining the standard deviation from that. The lower the standard deviation the higher the consensus. The third variable, SCAA, was derivedby the development of a Guttman Scale with eight items. These items are one through eight in the questionnaire. Then the total score of the eight items was used as the index. The fourth variable, that of importance of self identity student (ISIS) was constructed similarly from existing items in the pool of data. These items are nine through fifteen and thirty four through thirty eight. For this variable, as for the degree of structuring, the items were not the same as those 12 which would have been constructed if original data were used, however they came close enough to the intended purpose to allow their use. The data used in this study was part of an ongoing lon- gitudinal study carried out by Brookover and his associates, of a high school population in a midwestern city. The study began while the students were in seventh grade and continued until two years after high school graduation. The data for this thesis was taken from the results of the tenth grade study. The selection of the tenth grade sample was done for several reasons. The first is that a high drop out rate should not yet have occured. This allowed for a wider range of academic achievement levels. Also considered is the fact that the tenth grade is located in the middle of the high school years, a point at which the student is sufficiently stabilized in the high school environment, yet not so far along that he is beginning to envision a life outside of the school environment. The N for this study is 1290 students, with females equaling 606 and males equaling 59h. HYPOTHESES It is recognized that the four variables to be used may not be independent from one another. For example, the 'Academic Expectation Level in ego's role set probably influences both his self concept of academic ability (as Brookover has demonstrated) and also the importance to ego of his self iden- tity student. But the degree of consensus probably also influ- ences these two personality variables. But, influencing is not determining. There are many factors which combined in a matrix, determine these two personality variables. But an att— empt to isolate all of them would be like trying to isolate all of those factors, which combined would determing the overall personality of an individual. Brookover, utilizing the Meadian theory of symbolic in— teraction has shown in his research 10 that the following pro- positions are true: Academic expectation level for ego ego's perception of his academic ability ’Tego's perception of his AA‘51‘SCAA ’rAcademic expectation level for ego‘91\SCAA TS CAA-i 'rAcademic achievement (GPA) ’rAcademic expectation level for ego—91\Academic achievement (GPA) 10. 92. Cit. 13 la However by Brookover's own admission, that is not suff- icient. There is a time factor involved from the point at which expectations change, are perceived, become part of the self concept and materialize in level of academic achieve- ment. This can be diagramed as follows: Tl _ T2 T3 T4 Expectations ------ perceptions ------ self concept —-- behavior Brookover's research has shown that a time lag of two or more years is not unusual before a significant association between even self concept and behavior occurs. This fact alone justifies using Academic Expectation Level and SCAA as separate variables. Theoretically then SCAA should be more highly assoc- iated with school achievement (GPA) than is the Academic Expecta— tion Level, Thus: fiypothesis 1: The self concept of academic ability (SCAA) of a student is positively associated with his level of school achievement. (GPA) Hypothesis 2: The Academic Expectation Level of a student's academic significant others role set is positively associated with his level of school achievement. GPA) The other two variables, i.e. degree of consensus and importance of self identity student have never been empirically tested. That they will play a significant part in accounting for much of the remaining variance is theoretically implied but not empirically proven. Thus: fiyppthesis 3: The importance to a student of his self iden- tity as a student is positively associated with his level of school achievement. (GPA) 15 The fourth variable, the degree of consensus cannot by itself be associated with school achievement for its referent is the consensus which surrounds the expectations of the units in the role set. It must be considered in conjunction with the Academic Expectation Level. Theoretically an individual's school achievement would be high if the Academic Expectation Level was high and there was a high consensus about this. Similarly an individual's school achievement would be low if the Academic Ex- pectation Level was low and there was a high consensus about this. The other possible combinations would yield varying levels of achievement. I If these two variables were put into a matrix with the other two variables the result would be sixteen possible combin- ations, each of which would be associated with a level of school achievement. The matrix would be as follows, with the dependent variable being grade point average. (see Table l on the next page.) Such a formulation would yield the following hypotheses: Hypothesis 4: A high SCAA plus a high ISIS plus a high AEL with a high degree of consensus is positively associated with a high GPA. flypothesisgj: A high SCAA plus a high ISIS plus a high AEL with a low degree of consensus is positively associated with a high GPA. In a situation such as this it would not seem unlikely that there are probably one or two significants who exert consider- ably more influence on ego than the others in the role set. These significants would hold very high expectations. 16 TABLE 1 Hypothetical table for Prediction of Academic Achievement Level (G.P.A.) HI LO SCAA lISIS AELJ. HI LO HI LO Consensus HI HI HI MOD MOD HI LO MOD NOD LO HOD-LO HI HI-MOD MOD MOD-LO LO LO LO MOD-LO MOD-LO LO LO Hypothesis 6: A low SCAA plus a high ISIS plus a high AEL with a high degree of consensus is positively associated with a moder- ate GPA. In this situation ego has probably experienced consistent and relatively defined poor grades in school for some reason which cannot be explained by the theory (poor eyesight? personality quirks?) Hypothesis 2: A low SCAA plus a high ISIS plus a high AEL with a low degree of consensus is positively associated with a moderate GPA. Here ego is probably expected by some important significant to perform at a high level but there is not consensus about this. 17 He feels that being a good student is important but is giving only a limited effort because of the dissensus existing in the role set. Hypothesis 8: A high SCAA plus a high ISIS plus a low AEL with a high degree of consensus is positively associated with a mod- erate GPA. Such a situation, if it exists, should not occur very often. With a low Academic Expectation Level, ego should have or be tend- ing towards a low SCAA. However, a high degree of consensus about low expectations theoretically would not allow a high SCAA or high ISIS unless there are other significants who haven't been included in the role set, and hold high expectations. The more likely sit- uation is that ego is at the beginning of a significant change to a lower SCAA and ISIS. Hypothesis 9: A high SCAA plus a high ISIS plus a low AEL vith a low degree of consensus is positively associated with a moder- ate GPA. With a low AEL and a high SCAA it would seem likely that either a change has occurred in the attitudes of ego's signifi- cant others or the significants have changed. However, with a low degree of consensus there is considerable dissensus about the expectations for ego. It is likely that the perceived parental expectations have changed from high to low with the other signi- ficants maintaining reasonably high expectations. There should be a tendency toward the lowering of SCAA and ISIS. Hypothesis 10: A low SCAA plus a high ISIS plus a low AEL with a high degree of consensus is positively associated with a low 18 GPA. A situation such as this where the expectations are low and the consensus about them high, coupled with a low SCAA would outweigh the fact that ego places a high investment in his self identity as a student. His role-set gives him neg- ative support for high achievement. Hypothesis 11: A low SCAA plus a high ISIS plus a low AEL with a low degree of consensus is positively associated with a moderate to low GPA. Here we have a situation where some significant (prob- ably friend and/or teacher) regards the student role as impor- tant and feels that ego can achieve highly. Ego regards his student role as important but because of his uncertainty about this ability and the dissensus in his role set, he performs at a moderate to low level. Hypothesis 12: A high SCAA plus a low ISIS plus a high AEL with a high degree of consensus is positively associated with a high to moderate GPA. Here is a situation where ego probably has other iden- tities which are more important to him for self esteem main- tenance than the identity student. He feels that he has the ability to perform well but is probably investing less time and effort than he would if the identity student was very impor- tant to him. Hypothesis 1): A high SCAA plus a low ISIS plus a high AEL with a low degree of consensus is positively associated with a moderate GPA. The combined expectations for ego are marginally high with 19 some significant (8) having very high expectations and some low, yielding much dissensus. It is probable that ego's parents have the high expectations and the other significants have low expectations. Hypothesis 14: A low SCAA plus a low ISIS plus a high AEL with a high degree of consensus is positively associated with a moderate to low GPA. I A situation such as this would not be very likely unless ego regards the student role as so very unimportant that the significants in his role set are unimportant in the rest of his life space. Hypothesis l5: A low SCAA plus a low ISIS plus a high AEL with a low degree of consensus is positively associated with a low GPA. The high expectations are probably marginal as indicated by the low degree of consensus which exists in the role set. With a situation such as this, performance should be slightly less than that in Hypothesis 1%. Hypothesis 16: A high SCAA plus a low ISIS plus a low AEL with a high degree of consensus is positively associated with a moderate to low GPA. A situation such as this should not occur too often. Ego does not place much importance in his student identity, and this is reinforced by high consensus in his role set regarding low expectations. Yet ego still feels that he has high academic ability. This is an unstable situation in which ego's self concept should be decreasing. 20 Hypothesis 18: A low SCAA plus a low ISIS plus a low AEL with a high degree of consensus is positively associated with a low GPA. This should be one of the stable and frequent situations. All of the variables are consonant and the conditions are perfect for low achievement. Hypothesis 19: A low SCAA plus a low ISIS plus a low AEL with a low degree of consensus is positively associated with low GPA. This situation is essentially the same as that for Hypoth- esis 18 except that the consensus for low expectations is low. Because the personality variables are low however, this should still result in low achievement. ANALYSIS The entire analysis for all nineteen hypotheses was carr- ied out separately for males and females. Previous research of this nature has shown that quite different results often do occur on the same variables for males and females. Hypothesis one stated that "the self concept of academic ability of a student is positively associated with his level of school achievement." As Table 2 shows, the correlation coefficients for both males and females are very high and significant beyond the 0.001 level. Brookover reports similar results 11 in the sample which he used. He found correlations of 0.62 for the males with an N of approximately 255, and 0.53 for the females with an N of 307. For both males and females the samples which I used were over twice as large as the ones which Brookover used, and yet it's interesting to note the results which he initially obtained have been fairly well main- tained in this retest. 11. Ibid. p. 88 21 22 TABLE 2 Pearson Product Moment Correlation between SCAA 10th grade and GPA 10th grade Males Females N=591+ N=696 r 0.59 0.58 Both sig. beyond 0.001 Hypothesis two states that "the AEL of a student's academic significant others role set is positively associated with his level of school achievement." Table three shows that this is indeed the case for both males and females as expected. TABLE 3 Pearson Product Moment Correlation between AEL 10th grade and GPA 10th grade Males Females N=594 N=696 r 0.55 0.58 t=16.02 t=18.76 Both sig. beyond 0.001 The variable AEL you will remember, is a composite of the weighted expectations of ego's parents, friends and teachers. Correlations were run between the mean expectations of each of the three and GPA. As Table 4 shows, we were able to achieve 23 a greater positive association between expectations and GPA when the variable AEL was developed then when the unweighted expectations of each significant were considered separately. The difference, though not great, is in the direction hoped for. If we keep in mind that the variable AEL had to be dev- eloped in a way which I feel does not allow for its best predic- tive value, then the possibility that a much greater association will result, is likely, when it is formulated the way it was originally intended to be. TABLE 4 Pearson Product Moment Correlations between i expectations and GPA 10th grade Males Females N2594 N=696 Parents 0.53 0.56‘ Friends 0.U2 0.58 Teachers 0.51 0.5U Hypothesis three tested whether "the importance to a student of his self identity as a student is positively associated with his level of school achievement." Table 5 shows that this is indeed the case although the association is not as great as that of SCAA or AEL. 24 TABLE 5 Pearson Product Moment Correlation between ISIS 10th grade and GPA 10th grade Males Females N=594 N=696 r 0.30 0.30 t=7.65 t-8.28 Both sig. beyond 0.001 The main attempt of this thesis is of course to find out what type of association and potential prediction exists be- tween different combinations of the four independent variables and grade point average. High and low for the independent variables was arrived at separately for the male and female samples by taking the top half of all the scores for each and calling that high and taking the bottom half and calling that low. Thus, for each of the variables we had a different range for males and females. Similarly, the dependent variable of GPA was divided into three sections for males and females, with the lower third corresponding to low GPA, and the top third corresponding to high GPA. The range for all the variables for both sexes is shown in Table 6. SCAA AEL o Cons GPA High Low High Mod. Low . 25 TABLE 6 Range of scores for each variable in both sexes, including n's Males Females N: 594 N: 696 Range n Range n 28-40 297 28-40 348 8-28 297 8-28 348 32-43 297 31-43 348 8-32 297 8-31 348 44.80+ 297 43.60+ 348 44.80- 297 43.60- 348 .00-.37 297 .00-.35 348 .37-l.60 297 .35-1.48 348 2.25-4.00 198 2.75-4.00 232 1.50-2.25 198 2.00-2.75 232 0.25-1.50 198 0.25-2.00 232 26 As we can see, for the cutting point on the four inde- pendent variables, there is not a great deal of difference. However, for the cutting points for the dependent variable of GPA there is a considerable amount of variance. The fe- male sample shows that for moderate and high achievement the base point is 0.50 higher than it is for the males. The question arose as to how one determines what con- stitutes high, moderate and low achievement. For me to de- termine it, or anyone else would of course introduce consid- erable bias. The other, and probably more important considera- tion is that of relative definition. The assumption which was necessary to work on was that achievement levels are defined relatively and informally by the populations of different schools. That is, high achievement in School A may be regarded as a 3.00 or better, whereas in School B it might be 2.50 or better. To circumvent this problem, I divided the samples into three equivalent groups for males and the same for females, by tak- ing the highest third, middle third and bottom third in GPA. The result is a relatively defined situation of high, moderate and low achievement, not only on the basis of the sample, but also in terms of sex. Because each of the three categories of the dependent variables (GPA) had exactly the same n, I was able to use this as a basis of analysis. That is, if there was no association then theoretically each cell should have one-third high, one- third moderate, and one-third low achieving students. Using this fact, I was able to establish an expected frequency by taking the n which occured in each cell, dividing it into 27 three and using each third as the expected frequency for high, moderate, and low achievement. The observed frequency then, was what actually occurred in the three categories of the dependent variable. It was then possible to use a chi-square statistic on a lflN table to test for association. Most of the hypotheses seek an association between the four independent variables and one of the three categories of the dependent variable. When this occurred, the remaining two categories were combined for analysis and assumed an expected frequency of two-thirds n. In some of the hypotheses the dependent variable was two of the three categories and similarly the expected frequency was two-thirds of n and the observed frequency was combined for the two. TABLE 7 X2 test for Association Between High SCAA plus High ISIS plus High AEL with high consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. High 50 29 56 31 Mod-Lo 37 58 36 ‘ 61 n38? n=92 x = 22.81 x2 = 30.41 def. = 1 def. = l Sig. beyond 0.001 Sig. beyond 0.001 28 Table 7 shows the results for hypothesis four which stated that a high GPA would exist when these four independent variables occur together. As expected, a high degree of posi- tive association exists for both males and females with the female sample showing a greater positive association than the male sample. Hypothesis five also predicted a high GPA. As Table 8 shows, the hypothesis was well supported. The only surprise which exists is that the chi-square is greater here for the men than it was in Table 7. Theoretically this should not have occurred because unlike the previous situation where high expectations were supported by a high consensus, here they were supported by a low consensus. Therefore, the asso- ciation should have been slightly less, as it is for the fe- males. TABLE 8 X2 test for association between high SCAA plus high ISIS plus high AEL with a low consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. High 52 27 60 34 Mod. + Lo 30 55 ”3 69 n§82 ”£103 X = 34051 X :: 29.68 def. = l d.f. = l 318- beyond 0-001 Sig. beyond 0.001 29 Hypothesis six stated that a moderate GPA would result when a matrix of low SCAA, high ISIS, high AEL, and high degree of consensus existed. Table 9 shows that although the results are in the positive direction, they are not signi- ficant at the .05 level which serves as a base level of signi- ficance for all the hypotheses. TABLE 9 X2 test for association between low SCAA plus high ISIS plus high AEL with a high consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. Mod. 7 5 9 6 High + Lo 8 10 10 13 n=15 n=l9 x2 = 1.20 x2 = 2.19 def. = 1 def. = 1 11.8. nose The results for hypothesis seven which predicted that a moderate GPA would result, again are not significant. Again, the results are in the positive direction, slightly for the males, more so for the females. Lower chi-square values should be expected here than those reported in Table 9 for hypothesis six. This is expected for the reason that in this present case the high expectations are supported by low consensus. As we see, the chi-square value is indeed lower for the males but higher for the females 30 TABLE 10 X2 test for association between low SCAA plus high ISIS plus high AEL with a low consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. Mod. 7 6 8 5 High+ L0 11 12 6 9 11:17; nfilu x2 = 0.25 x = 2.8 def. = 1 def. = J. n.s. n.s. Hypothesis eight predicts a moderate GPA given the situation stipulated. The results in Table 11 show this to be the case for both the male and female sample. It's in- teresting to note however that the chi-square value for the males is considerably higher. Because the samples are not very large however, it probably is not safe to speculate as to the reasons for this. 31 TABLE 11 X2 test for association between high SCAA plus high ISIS plus low AEL with high consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. Mod. 15 6 10 6 High + Lo u 13 7 11 n=l9 n=l7 x2 = 19.73 x2 = 4.12 def. - 1 def. = l sig. at 0.001 sig. at 0.05 The situation in hypothesis nine is the same as that in eight except that now there is disensus as to the high expec- tations for ego. Therefore, while moderate achievement was expected it was anticipated that the positive association with moderate achievement would not be as high as in eight. This is indeed the case as Table 12 shows. However, as one can see by the very low chi-square values and the non-significance resulting, the positive association is much lower than antici- pated for the females and in the negative direction for the males. 32 TABLE 12 X2 test for association between high SCAA plus high ISIS plus low AEL with low consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. Mod. 5 6 6 5 High+Lo 12 ll 8 9 n=l7 nfilu X = 0.26 x = 0.31 def. = 1 def. = l n.s. n.s. The situation in hypothesis 10 calls for low achieve— ment. Table 13 shows that the association for both samples is in the hypothesized direction with, however, only the male sample having a significant chi-square value. TABLE 13 X2 test for association between low SCAA plus high ISIS plus low AEL with high consensus and GPA Males Females GPA OBS. EXP OBS. EXP. Low 19 11 19 16 High+hod 13 21 29 32 n§32 n=48 X = 8-87 x2 - 0.84 def. = l dofo = l sig. at 0.05 n.s. 33 The next situation predicts a moderate to low achieve- ment given a situation where a lack of consensus exists about low expectations. Here again the hypothesis (11) was upheld although the results were non significant for the male sample. TABLE 14 X2 test for association between low SCAA plus high ISIS plus low AEL with low consensus and GPA Hale Females GPA OBS. EXP. OBS. EXP. Mod-Lo 21 18 36 27 High 6 9 5 14 n—27 n=4l x2 = 1.50 x2 = 8.79 dof. '3 1 def. = l n.s.? sig. at 0.05 Hypothesis twelve calls for a high to moderate achieve- ment level. For both samples the chi-square values are significant and the hypothesis can be safely accepted as Table 15 shows. 30 TABLE 15 X2 test for association between high SCAA plus low ISIS plus high AEL with high consensus and GPA Males Females GPA _, OBS. EXP. OBS. EXP. High-Mod. 2G 19 27 19 Low 4 9 2 10 n=28 n=29 x2 = n.09 x2 = 9.77 def. = 1 def. = 1 sig. at 0.05 sig. at 0.05 Hypothesis thirteen predicted that a moderate GPA would exist given the combination of independent variables set up. Table 16 shows that the hypothesis cannot be accepted on the basis that the null hypothesis of no difference was upheld. As can be seen, the observed and expected frequencies for the female sample were the same, and almost the same for the males. TABLE 16 X2 test for association between high SCAA plus low ISIS plus high AEL with low consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. Mod. 13 12 21 21 High+Lo 22 23 #1 41 n§35 n=62 X = 0.13 x2 = 0.00 def. = 1 def. = l n.s. n.s. 35 The same situations exist for hypotheses fourteen and fifteen. In both cases the hypotheses were not upheld. The results for hypothesis fourteen which predicted a mod-low GPA are reported in Table 17, and show a chi—square value of 0.00 for the females and a very low one for the males. Similarly, for hypothesis fifteen reported in Table 18, the observed and expected frequencies for the males are the same and the females show a negative association between the independent variables and the predicted low GPA. This negative association is very low however, and non-signficant. TABLE 17 X2 test for association between low SCAA plus low ISIS plus high AEL with high consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. Mod.+Lo ll 10 6 6 High 4 5 3 3 n—l5 n—9 x2 = 0.30 x2 = 0.00 def. = 1 def. = 1 n.s. n.s. 36 TABLE 18 X2 test for association between low SCAA plus low ISIS plus high AEL with low consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. Low 6 6 5 7 High+Mod. 11 ll 15 13 n-l? n := 20 x2 = 0.00 x2 = 0.88 dofo = 1 def. -' l n.s. n.s. Hypothesis sixteen asks for a moderate to low GPA given the situation existing. Table 19 shows that although the chi-square values for both the male and female samples are in the hypothesized direction, the results are non significant and therefore the hypothesis will have to be rejected. TABLE 19 X2 test for association between high SCAA plus low ISIS plus low AEL with a high consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. Mod.-Low 10 9 14 12 High 4 5 4 6 n=1u n=18 X2 = 0.31 x2 = 0,99 dof. = l dgf. 1' l n.s. n.s. 37 Hypothesis seventeen predicts a moderate to low GPA as does hypothesis sixteen. The only difference is that in this present situation the low expectations are not supported by high consensus as in hypothesis sixteen. Therefore, the probability of positive association should be somewhat lower. This is indeed the case for the females where we find a change from a positive association with a X2 value of 0.99 in hyp- othesis sixteen to a negative association with a chi-square value of 3.25 in hypothesis seventeen. Even though this expectation was upheld, nevertheless the hypothesis has to be rejected on the basis of a negative association with a non- significant chi-square for the females. The results for the male sample were, as can be seen in Table 20, in the expected direction, but non-significant. TABLE 20 X2 test for association between high SCAA plus low ISIS plus low AEL with a low consensus and GPA Males Females GPA OBS. EXP. OBS. EXP. Mod.-Low ll 10 6 9 High 4 5 7 u n=15 n=13 X2 = 0.30 X2 = 3.25 def. = 1 def. = l n.s. n.s. Hypothesis 15 and 16, both of which predict low achieve- ment, have both been upheld with very high chi-square values 38 and very high levels of significance in both the male and female samples. of these hypotheses reSpectively. Tables 21 and 22 report the results for both As anticipated, the sit- uations which these hypotheses describe, have a large number of occurrences. r‘ GPA Low High+Hod. GPA Low High+fiod. TABLE 21 X2 test for association between low SCAA plus low ISIS plus low AEL with a high consensus and GPA Males Females OBS. EXP. OBS. EXP. A2 29 66 39 45 58 50 77 n=87 n=ll6 def. = 1 def. = l sig. at 0.005 TABLE 22 2 sig. at 0.001 X test for association between low SCAA plus low ISIS plus low AEL with low consensus and GPA Males Females 083. X . OBS. EXP. 52 29 50 27 34 57 31 54 n=86 n§81 x2 = 27.52 X = 29.39 def. = l d.f. = 1' sig. at 0.001 sig. at 0.001 DISCUSSION OF RESULTS AND CONCLUSION A discussion of the results of the analysis was re- served until this section so that an overall picture of the results could be obtained. Questions arose such as what hyp- otheses were upheld for both males and females and what hyp- otheses were rejected for both males and females. More inter- esting however is the question of whether any partial combina- tions of the independent variables were consistently found to be associated with success or failure of hypotheses. Table 22 shows those hypotheses which were accepted for both the male and female sample. Acceptance of the hypotheses was at the minimum of a significance level of 0.05. The hypotheses are arranged from highest chi-square value to the lowest chi-square value. Similarly Table 23 shows those hypotheses which were rejected for the male and female samples. Here the hypotheses are also arranged in descending order from the highest non-signi- ficant chi-square to the lowest non-significant chi-square value. There are two very striking things to note in contrasting these tables. The first is that for both the male and female samples, those hypotheses whose results were significant had the most combinations of independent variables which might 39 #0 TABLE 23 Hypotheses accepted, with n's, four independent variables and GPAZpredicted - ordered from highest X value to lowest Males N SCAA ISIS AEL OCOh. GPA FRED. Hyp. 5: 82 High High High Low High Hyp. 19: 86 Low Low Low Low Low HYD. Us 87 High High High High High Hyp. 8: 19 High High Low High hod. Hyp. 10: 32 Low High Low High Low Hyp. 18: 87 Low Low Low High Low Hyp. 12: 28 High Low High High High-Mod. Females N SCAA ISIS AEL OCon. GPA PHED. HYP. U: 9? High High High High High Hyp. 5: 103 High High High Low High Hyp. 19: 81 Low Low Low Low Low Hyp. 18: 116 Low Low Low High Low Hyp. 12: 29 High Low High High High-Mod. Hyp. ll: #1 Low High Low Low Mod-Low Hyp. 8: 17 High High Low High Hod 41 Hypotheses rejected, with n's, four independent variables and GPA predicted - ordered from highest X2 value to lowest Males N SCAA ISIS AEL 0Com. GPA PHED. Hyp. 11: 27 Low High Low Low Mod.-Low Hyp. 6: 15 Low High High High Mod. Hyp. 16: 13' High Low Low High Mod.- Low Hyp. 14: 15 Low Low High High Mod.- Low Hyp. 17: 15 High Low Low Low Mod.- Low Hyp. 7: 18' Low High High Low Mod. Hyp. 13: 35* High Low High Low Mod. Hyp. 15: 17 Low Low High Low Low Hyp. 9: 17 High High Low Low Mod. Females N SCAA ISIS AEL OcOn. GPA FRED. Hyp. 7: 13* Low High High Low hod. Hyp. 6: 19 Low High High High Nod. Hyp. 16: 18* High Low Low High hod.- Low Hyp. 10: MH' Low High Low High Low Hypo 9: 1h High High Low Low Mod. Hyp. 13: 62 High Low High Low Mod. Hyp. in: 9 Low Low High High hod.- Low Hyp. 15: 20 Low Low High Low Low Hyp. 17: 13 High Low Low Low Mod.— Low TABLE 2a 42 be called "theoretically congruent." That is, the combina- tions of independent variables in these hypotheses would seem to belong together more logically than the ones which were rejected. It is also interesting to note that the samples for these hypotheses were for the most part among the largest. The exceptions to this rule were hypotheses eight for the males and eight for the females. These as you will note, had relatively lower n's and were not as theoretically congruent as the others. Similarly, if we look at Table 23 for those hypothe- ses which were rejected we will note that there is little theoretical congruence and that the n's are comparatively small, with a few exceptions in the samples of both male and female. The other important result to note is that for those hypotheses which were not upheld, the GPA predictions were all either moderate or moderate to low. The exceptions were hypo- thesis 15 for the males and hypotheses 10 and 15 for the females. Similarly, the hypotheses which were held predicted either high or low achievement with the exceptions being hypotheses H and 12 for the males and 8,—11, and 12 for the females. These hypotheses which called for high or low GPA are more logical and are also the polar extremes in four out of the seven situations. The general conclusion which must be drawn is that although we obtained results which were in the hypothesized direction for practically all those where a moderate or mod- erate to low GPA was called for, the lack of logical consistency among the combinations of the four independent variables weak- ened the influence which each exerted upon the student, and 43 allowed for the influence of other variables to be exerted, in what at present appears to be a non-predictable fashion. The attempt of this thesis which was to find a formula by which school achievement could be predicted was partially successful. I don't feel however that the attempt to improve this formula should end with this present research. In the first place, the operationalizing of the concepts of ISIS, degree of structuring, and degree of structural integration can be improved upon as mentioned earlier. A second area of explanation is that of determining what other variables are operating in those cells which predict moderate achievement, to counteract the influence of the four independent variables used here. QUESTIONNAIRE Bureau of Educational Research, Michigan State University High School Study Introduction: The Bureau of Educational Research at Michigan State University has been doing a study to find out what high school students think about themselves, their school work and their future plans. Many of you have helped in one part of the study. The informatiOL that you gave us has been very useful and we thank you for your help. Now we would like you to help in this part of the study by answering the following questions as honestly as you can. Please read carefully the directions on each part of the questionnaire before you answer. If you have any ques- tions, raise your hand and someone will help you. The answers you give will not be shown to your teachers or anyone else, and will in no way affect your grades. Noone will see the answers you give except the research staff. YOUR HELP IN THIS STUDY IS GREATLY APPRECIATED: 4h 45 PLEASE PRINT THE FOLLOWING INFORMATION Name 8 g 9 Birthdate: , , Sex: M F Name of Present School : What School Did you Attend Last Year? Circle the letter in front of the statement which best answers each question. 1. How do you rate yourself in school ability compared with your close friends? A. I am the best. B. I am above average. C. I am average. D. I am below average E. I am the poorest. 2. How do you rate yourself in school ability compared with those in your class at school? A. I am among the best. B. I am above average. C. I am average. D. I am below average. E. I am among the poorest. 3. Where do you think you would rank in your high school graduating class? A. Among the best. B. Above average. C. Average. D. Below average. E. Among the poorest 1:6 Do you think you have the ability to complete college? A. Yes, definitely. B. Yes, probably. C. Not sure either way. D. Probably not. E. No. Where do you think you would rank in your class in college? A. Among the best. B. Above average. C. Average. D. Below average. E. Among the poorest. In order to become a doctor, lawyer, or university prof- essor, work beyond four years of college is necessary. How likely do you think it is that you could complete such advanced work? A. Very likely. B. Somewhat likely. C. Not sure either way. Do Unlikely. E. Host unlikely. Forget for a moment how others grade your work. In your own opinion how good do you think your work is? A. My work is excellent. B. My work is good. C. My work is average. D. My work is below average. E. My work is much below average. What kind of grades do you think you are capable of getting? Mostly A's. Mostly B's. Mostly C's. Mostly D's. Mostly E's. MUOCUi> 10. 11. 12. 13. la. How 47 important to you are the grades you get in school? A. Very important. B. Important. C. Not particularly important. D. Grades don't matter to me at all. How important is it to you to be high in your class in grades? A. Very important. B. Important C. Not particularly important. D. Doesn't matter to me at all. How do you feel if you don't do as well in school as you know you can? A. Feel very badly. l. Feel badly. C. Don't feel particularly badly. D. Doesn't bother me at all. How important is it to you to do better than others in school? A. Very important. B. Important. C. Not particularly important. D. Doesn't matter to me at all. Which statement best describes you? A. B. C. D. like to get better grades than everyone else. like to get better grades than almost everyone else. like to get about the same grades as everyone else. don't care about any particular grades. HHHH In your school work do you try to do better than others? A. B. C. D. All of the time. Most of the time. Occassionally. Never. #8 15. How important to you are good grades compared with other aspects of school? A. B. C. D. Good grades are the most important thing. Good grades are among the important things in school. Some other things in school are more important. Good grades don't matter to me at all. Please answer the following questions as you think your PARENTS would answer them. If you are not living with your parents answer for the family with whom you are living. Circle the letter in front of the statement that best answers Each Question. 16. How do you think your parents would rate your school ability compared with other students your age? tilOOCUI> Among the best. Above average. Average. Below average. Among the poorest. 17. Where do you think your parents would say you would rank in your high school graduating class? A. B. C. D. E. Among the best. Above average. Average. Below average. Among the poorest. 19. Do you think that your parents would say you have the abil- ity to complete college? A. B. C. D. E. Yes, definitely. Yes, probably. Not sure either way. Probably not. Definitely not. 19. 20. 21. 49 In order to become a doctor, lawyer, or university prof- essor, work beyond four years of college is necessary. How likely do you think your parents would say it is that you could complete such advanced work? A. Very likely. B. Somewhat likely. C. Not sure either way. D. Somewhat unlikely. E. Very unlikely. What kind of grades do you think your parents would say you are capable of getting in general? A. Mostly A's. B. Mostly B's. C. Mostly C's. D. Mostly D's. E. Mostly E's. How far do you think your parents expect you to go in school? A. They expect me to quit as soon as I can. B. They expect me to continue in high school for a while. C. They expect me to graduate from high school. D. They expect me to go to secretarial or trade school. E. They expect me to go to college for a while. F. They expect me to graduate from college. G. They expect me to do graduate work beyond college. For your parents to be most pleased with you, what kind of grades should you get in school in general? A. Mostly A's. B. A's and B's. C. Mostly B's. D. B's and C's. E. Mostly C's. C's and D's. Mostly D's. D's and E's. Mostly E's. My grades do not make any difference to my parents. MFHUIQ*U 50 Please answer the following questions as you think your closest friend would answer them. Circle the letter in front of the statement that best answers each question. 23. How do you think your closest friend would rate your school ability compared with other students your age? A. Among the best. B. Above average. C. Average. D. Below average. E. Among the poorest. 2Q. Where do you think your closest friend would say you would rank in your high school graduating class? A. Among the best. B. Above average. C. Average. D. Below average. E. Among the poorest. 25. In order to become a doctor, lawyer, or university prof- essor, work beyond four years of college is necessary. How likely do you think your closest friend would say it is that you could complete such advanced work? A. Very likely. Somewhat likely. Not sure either way. Somewhat unlikely. Very unlikely. LTJDOW 26. What kind of grades do you think your closest friend would say you are capable of getting in general? A Mostly A's. B. Mostly B's. C. Mostly C's. D. Mostly D's. E. DIOStly E's. 27. How far do you go in school? A. B. C. D. C)WEU He He He He He He He expects expects expects expects expects expects expects think me me me me me me me to to to to to to to 51 your closest friend expects you to quit as soon as I can. continue in high school for a while. graduate from high school. go to secretarial or trade school. go to college for a while. graduate from college. do graduate work beyond college. ZR. For your closest friend to be most pleased with you, what kind of grades should you get in general? A. B. C. D. E. F. G. H. I. J. r'IOStly A. S o A's and B's. MOStly B's. B's and C's. MOStly C's. 0'3 and D's. Mostly D's. D's and E's. Mostly E's. My grades do not make any difference to my closest friend. Please answer the following questions as you think your favorite teacher would answer them. like best; work. This teacher should be the one you the one you feel is most concerned about your school Circle the letter in front of the statement that best answers each question. 29. How do you think your favorite teacher would rate your school ability compared with other students your age? A. B. C. D. E. Among the best. Above average. Average. Below average. Among the poorest 30. 31. 32. 33. 34- 52 Where do you think your favorite teacher would say you would rank in your high school graduating class? A. B. C. D. E. Among the best. Above average. Average. Below average. Among the poorest. In order to become a doctor, lawyer, or university prof- essor, work beyond four years of college is necessary. How likely do you think your favorite teacher would say it is that you could complete such advanced work? A. B. C. D. E. Very likely. Somewhat likely. Not sure either way. Somewhat unlikely. Very unlikely. What kind of grades do you think your favorite teacher would say you are capable of getting in general? A. B. C. D. E. Mostly A's. Mostly B's. Mostly C's. MOStly D. S 0 Mostly E's. How far do you think to go in school? A. B. C. D. E. F. G. He expects He expects He expects He expects He expects He expects He expects me me me me me me me to to to to to to to your favorite teacher expects you 9/ quit as soon as I can. continue in high school for a while. graduate from high school. go to secretarial or trade school. go to college for a while. graduate from college. do graduate work beyond college. Would you rather be a good student or good in sports? A. B. C. D. Good student. Good in sports. Can't decide. Both. 35. 36. 37. 38. NOTE: 53 Would you rather be a good student or well liked by others of your sex. A. Good student. B. Well liked by others of my sex. C. Can't decide D. Both Would you rather be a good student or popular with those of the opposite sex? A. Good student. B. Popular with those of the opposite sex. C. Can't decide. D. Both Would you rather be a good student or a leader in school activities? A. Good student. B. Leader in school activities. C. Can't decide. D. Both If you were free to go as far as you wanted in school, how far would you like to go? A. I'd like to quit right now. B. I'd like to go to high school for a while. C. I'd like to graduate from high school. D. I'd like to go to business or technical training school. E. I'd like to go to college for a while. F. I'd like to graduate from college. G. I'd like to do graduate work beyond college. (This is an abbreviated version of the original ques- tionnaire) LIST OF REFERENCES Backman, Carl W. and Paul F. Secord, A Social Psychological View of Education, New York: Harcourt, Brace & World, Inc., 1968. Brookover, Wilbur IL, et. al., Self Concept 9: Ability and School Achievement, III, Cooperative Research Project, No. 2831, Michigan State University, 1967. Faunce, William, A., Problems of an Industrial Society, New York: McGraw-Hill Book Company, 1968. Herriott, Robert E. and Nancy Hoyt St. John, Social Class And The Urban School, New York: John Wiley & Sons, Inc. 1967. Hobson, Julius W. v. Carl Hansen, Supt. of Schools and The Board of Education for the District of Columbia. United States District Court for District of Columbia Civil Action No. 82-66, 1967. fqead, George Herbert, Mind, Self, and Society, Chicago, The University of Chicago Press, 1967. Rosenthal, Robert, and Lenore Jacobson, "Teacher's Expectan- ciesx Determinants of Pupil's I.Q. gains," Psycholggical Reports, Vol. 19, 1966. U. S. Commission on Civil Rights, Racial Isolation In Public Schools, Washington, D. c.: U. S. Government Printing Office. HICHIGRN STATE UNIV. LIBRRRIES 31293100003049