DRIFrS IN LOCUS or CONTROL AS A . FUNCTION OF moss - CULTURAL EXPERIENCES - ' Thesis for the Degree of M. A ' MICHIGAN STATE UNIVERSrrY ' HILTON T. THOMAS 1975 mmwnfl0mnw’ WW I IWWWHWHHW , 3 1293 01094 3920_ " “m4t1\uesm-~umv~’e I ”ES‘S L I n “a; a "\ t; ' “‘2'? ,-. 'h'. ‘.-. > §~ ,3. g P j: ‘ I; . . ._ 1" _ ‘i’ #32: ' ' ‘ ‘ g»; 43:1." . Amt a ‘2 iowa '73 3% @753 i052 ABSTRACT DRIFTS IN LOCUS OF CONTROL AS A FUNCTION OF CROSS-CULTURAL EXPERIENCES BY Hilton T. Thomas A number of researchers have studied the Internal-External Locus of Control construct. Many of these researchers have studied how the construct relates to Blacks and lower class individuals, only to report these groups to be more external than Whites or middle class individuals. For the most part, these researchers have not been attuned to the experiential differences in poweranuiexpectancy for Black and White popu- lations. Invariably the internal perspective, with the exception of the extreme internal viewpoint, arises as the most intelligent, competent and desirable locus of control. Inasmuch as externality has a high frequency of occurence in minority p0pulations, these negative implications of exter- nality may be inappropriately transferred to minority popula- tions. This externality is a socially learned construct and not an inherent racial distinction. One's past experiences are of considerable importance in formulating a locus of control. The present study is based on the contention that there is a direct relationship between locus of control and past Hilton T. Thomas experiences. Therefore, altering experiences should also alter locus of control, in experiences where Blacks as a group have some expectancy for success they operate in an internal, achievement oriented manner. The purpose of this study was to set in perspective the internal-external locus of control construct as it relates to Blacks as a group. It is not enough to simply diSpel the stigma of Blacks as inferior by reason of externality but also to offer, study and explain alternate interpretations and motivations for externality. The experiences of subjects were experimentally altered through the use of a simulation game (SIMSOC). By modifying the prescribed SIMSOC procedure slightly, it was possible to elicit the desired effects ethically, economically and safely. One modification of SIMSOC was the delegation of power posi- tions by race rather than by chance. Two separate games were administered, one with Blacks having the power positions and the other with Whites having the power positions. In each game there was a minority of the subordinate population who were delegated to power positions. In addition the three regions in which players "lived" were a function of their power positions. The first region was completely composed of the dominant race who held most of the power. The second region was racially integrated and was moderately powerful. The third region was totally composed of the subordinate race and was void of any power, The Black dominated game repre- sented the reversal situation while the White dominated game Hilton T. Thomas represented.thestatus quo. The effects of the experimental game treatment were measured by pre- and post-test scores on the Rotter I-E Scale, in addition to situational, observational, behavioral rates. The results indicate that Blacks in power positions in the Black dominated game scored significantly more internal on the I-E post-test (p .025). Blacks not having power positions in the Black game apparently identified with Blacks in power to a small degree as there was a trend toward internality (E .09). Whites not in power in the Black game did not move toward externality as a function of their subordinate game experience. Though there was no significant movement in this group, the slight movement that did occur was in the internal direction. As predicted there was no significant movement in the White (status quo) dominated game. Though no change was predicted for the White dominated game it is difficult to assert the cause of the null hypothesis. The behavioral ratings were apparently measuring a different construct than the I-E Scale, or at least a different aspect of the phenomenon as their correlation was near zero. Implications of the above findings and their generality to real life situations were discussed. Recommendations for further research were suggested in the forms of training programs and the development of Black models. Hilton T. Thomas Approved: Date: Thesis Committee: J. Gordon Williams, Chairman Robert Calsyn Stan Kaplowitz DRIFTS IN LOCUS OF CONTROL AS A FUNCTION OF CROSS-CULTURAL EXPERIENCES BY Hilton T. Thomas A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1975 To my Parents p. g... ACKNOWLEDGMENTS I sincerely wish to thank my chairman, Dr. J. Gordon Williams, for his invaluable assistance in the planning and preparation of this thesis. I am deeply indebted for the tremendous amount of time he devoted in the final stages of this research. His generosity and patient indulgence through the many stages and aspects of this research is greatly appreciated. I would also like to gratefully acknowledge the other members of my committee, Dr. Robert Calsyn and Dr. Stan Kaplowitz. They offered many suggestions and constructive criticisms in the design and analysis phases of this study. Further thanks is due to Dr. Kaplowitz for the utilization of his sociology class in the training of the experimenters. Special thanks and sincere appreCiation is expressed to Dr. Lawrence Messé for his unsurpassed statistical guidance. Although he was not a member of the committee and initially did not know me, he took time out from a very busy schedule for numerous statistical consultations. He is truly one of the most dedicated psychologists it has been my pleasure to encounter. I would like to thank Dallas Williams for his helpful assistance in the early ANOVA analyses. iii Tribute is given to Caryl, my extended family, and many friends who throughout the years have been invaluable in their support, understanding and warm wishes. Finally, my deepest appreciation is expressed to my parents. To my mother, whose faith, prayers and encouragement has sustained me throughout this research and throughout my life. To my late father, though deceased, his spirit continues to be one of the most motivating forces in my life. iv LIST OF TABLES TABLE OF CONTENTS LIST OF APPENDICES O O O O O O O O O 0 CHAPTER I II III IV INTRODUCTION . . . . . . . Black Experiences . . . Implications of Black Experiences for Internal-External Control overView O O O O O O O 0 Operational Hypotheses . Experimental Hypotheses. METH OD O O O O O O O I O 0 Subjects . . . . . . . . Experimenters . . . . . Instruments . . . . . . Procedure . . . . . . . RESULTS 0 O O O O O O O I O Biases in Interpreting the Treatment Effects . . . Collapsed Treatment Effect Summary of Results . . . DISCUSSION . . . . . . . . Hypothesis Related Findings Comparison with I-E Literature . Additional Findings--ANOVA Analyses Rotter I-E Scale and Behavioral Ratings . . . . . . . . Methodological Considerations Implications for Further Research LIST OF REFERENCES . . . . . . . . . . APPENDICES Page vi -vii 11 15 18 18 20 20 21 21 23 27 27 29 36 38 40 41 45 45 47 50 54 58 Table Appendix A1 A2 LIST OF TABLES Pre-Test Mean Differences on Rotter I—E Scale for Black and White Power Groups . . . Group Movements Toward Internality on the Rotter I-E Scale . . . . . . . . . . . . Summary of ANOVA for Rotter I-E Scale . . . Summary of ANOVA for Behavioral Ratings . . Means of Race X Power (Ratings) . . . . . . Means of Race X Game (Rotter) Over Pre-Test and Post-Test Scores . . . . . . . . . . . . Summary of ANOVA for Rotter I-E Scale (With the Exclusion of Game) . . . . . . . . Summary of ANOVA for Behavioral Ratings (With the Exclusion of Game) . . . . . . . . Simple Effects of Race X Power (Ratings) . . Simple Effects of Race X Game (Rotter) . . . vi Page 28 30 33 34 35 36 37 38 68 68 Appendix A B LIST OF APPENDICES Demographic Questionnaire Social Reaction Inventory SIMSOC Instructions . . . Rating Guidelines . . . . Analysis of Simple Effects vii Page 58 59 63 66 68 CHAPTER I INTRODUCTION The present research is designed to examine some of the drifts or shifts in locus of control as a function of cross- cultural experiences. Internal locus of control refers to individuals who believe that reinforcements are contingent upon their own behavior, capacities or attributes. External control refers to individuals who believe reinforcements are not under their personal control but rather under the control of powerful others', luck, chance, fate, etc., (Rotter, 1966). An individual who attributes life's outcomes to skill will foster an internal perspective. His counterpart, who regards skill as incidental to his progress will form an external perspective of the world (James and Rotter, 1958; Crandall, 1963; wychoff and Sedowsky, 1955; Phares, 1962). These perceptions of controlling forces in one's environment have led to the development of the present Internal-External (I-E) Locus of Control construct. The measurement of this con- struct has been undertaken by various researchers (Phares, 1957; James, 1957; Rotter, Liverant and Crowne, 1961). The most recognized instrument has been developed by Rotter (1966). In Rotter's instrument the "items deal exclusively with the subject's belief about the nature of the world" (p. 10). 1 The items are constructed to measure generalized expectancy but none of the items are directly addressed to the preference for internal or external control. Internal-External Locus of Control is determined by the effects of reinforcement upon expectancy (Rotter, 1954, 1955, 1960). A person's expectancies about reinforcement in the future are greatly due to the value of the reinforcements available and to his prior experiences with reinforcement.. For example, the value and meaning of reinforcements and rewards vary across events and across individuals. What might be reinforcing for one individual in a particular setting may be totally worthless to someone else. This reinforcer there- fore may have different effects on the expectancies of various individuals. Consequently these differences have great implications for the cultural differences highlighted in this study. In addition to the value of a reinforcement, the frequency of the reinforcement may also effect expectancy. Thus, an expectancy for reinforcement in the future is diminished when the reinforcement schedule has been low. When applied to this study, it would indicate that individuals in a culture such as the Black culture, who have had relatively little relevant or positive reinforcement for social and economic achievement, would have a low expectancy for reinforcement of subsequent achievement. This is supported in the work by Goodnow and Pettigrew (1955), where differential reinforcement experiences had marked differences on ensuing expectancies. When placed under extinction conditions, subjects who had high reinforcement schedules adapted signi- ficantly quicker when reinforcement was reinstated than did subjects with low reinforcement schedules. Goodnow and Pettigrew stated the underlying variable of their finding as an expectancy for success based on prior successful experi- ences. While the View of expectancy presented by Goodnow and Pettigrew has been well documented in the literature (Feather, 1963; Aronson and Carlsmith, 1962; Lowin and Epstein, 1965; Brock, Edelman, Edwards and Schuck, 1965), a second form of expectancy should be distinguished in order to broaden the concept of expectancy. The distinction is made between: 1) an expectancy for success versus failure and 2) an expectation or belief that one's success and failure is, or is not, due to his own actions or lack of actions (Crandall, 1974).1 Although an individual may have had a history of successful experiences, if he does not attribute the outcome of these experiences to his artistry, his per- ception may be quite external. There is some indication that these expectancies for success and/or failure may be culturally influenced. Inasmuch as certain cultures in America have not had free access to the social and economic reinforcers, their expectancies for success have been rather low. Researchers (Battle and Rotter, 1963; Lefcourt and Ladwig, 1965a, 1965b; Crandall, Katkovsky, and Crandall, 1965; 1V. C. Crandall, personal communication, February 7, 1974. Rotter, 1966) have reported that Blacks and lower class individuals typically score higher in external locus of control than do Whites and middle class individuals. This means that Blacks and lower class individuals as groups perceive "others" or external forces as being the controlling factors of their lives, while Whites and middle class indi- viduals as groups see themselves as the controlling forces in their lives. In reviewing the above literature invariably the inter- nal perspective, with the exception of the extreme internal viewpoint, arises as the most intelligent, competent and desirable locus of control. Inasmuch as externality has a high frequency of occurrence in minority populations, one must be careful not to transfer inappropriately the implicit negative connotations of externality to these minority populations. The reader should be reminded that externality is a socially learned construct and not an inherent racial distinction. Furthermore, this external construct has been learned as a function of the subordinate role minorities have played in America. It is the experiences and feelings associated with this subordinate role which has brought about this externality. As indicated by Gurin, Gurin, Lao, and Beattie (1969) events may correctly be perceived by Blacks as external but have nothing to do with randomness or luck. Focusing on external factors may be motivationally healthy for Blacks instead of damaging when it concerns assessing one's probabilities for success against systematic and real external obstacles rather than exigencies of fate. Thus an internal response reflecting acceptance of blame for one's failures which might be considered "normal" in the typical middle class experience, may be extreme and intrapunitive for a Black person having grown up in poverty, in the ghetto. This study attempts to examine the effects of a situ- ational modification of these experiences. It is hypothesized that Blacks, placed in and reinforced for a dominant, decision- making role, will experience a shift toward internality. If, of course, this dominant decision-making role is not permanent or congruent with other roles in the individual's repertoire, the shift will only be a situational one. This expectancy is supported by Williams and Stack (1972), who found situa- tional factors to be of primary concern in predicting and interpreting the behavior of Black subjects. They found that given appropriate attention to expectancies and reinforce— ment value, Blacks performed in an internal, achievement- oriented manner. Their results also concur with the present study's premise that Blacks do not perform internally in subordinate settings as a result of their role in a White middle class society. The contention that internality can be learned in the appropriate environment is supported and expanded by Hunt and Hardt (1969) in their comparison of Black and White students in the Upward Bound Program. In their study both Blacks and Whites increased in motivation for college, interpersonal flexibility and orientation for the future, though all these variables may naturally increase with age. Increases in measures of self-esteem and internal control were observed only for the culturally disadvantaged high school students, and these variables do not typically increase with age. The fact that self-esteem and internal control measures increased only for "culturally disadvantaged students" suggest that the potential for increasing inter- nality may be greater with minority groups than with majority groups. This.isnot at all surprising, if one considers that most majority groups have had prior successful experiences with expectancy and reinforcement. It may be somewhat naive therefore to imagine that more training should produce a significant increase for majority groups. If we accept the findings of Hunt and Hardt (1969), it appears that internality may be acquired through training. Additionally, it appears that such training programs were most effective with minorities. Gore and Rotter's work (1963) may serve as a starting point for developing a training program to achieve this increase in internality. They found that means on the I-E test followed closely the degree of social-action taking. Those individuals who were more inclined to view themselves as determiners of their own fate tended to commit themselves to more personal and decisive social action. If we may assume their premise functions conversely, individuals who assume more personal and decisive social action are more inclined to regard themselves as determiners of their own fate. Thus we arrive at the central theme of this study which may be simply stated: those who have the opportunity to control their own fate will become more internal. In order to develop a program which will attempt to situationally alter the experiences, reinforcements and sub- sequently the expectancy for locus of control, a further exploration of the concomitants of the existing situation for Blacks is needed. Hence to fully understand the theme that those who have the opportunity to control their own fate will become more internal, it may be helpful to grasp how this opportunity was blocked. In order to alleviate a situation, it is helpful if one is cognizant of the situation. The terse account of the Black Experience provided by the following section will attempt to sensitize the reader to the purported blocked opportunities. Black Experience The focus here will be an attempt to give a minimal account of the Black experiences with which this study is concerned. It is by no means an attempt to give a history of struggles by the Black race. In the history of Negroes--Blacks--Afro Americans, certain victimization has occurred. Some see this victimiza- tion as a hardship they have endured at the hands of the White race. Others may see the situation as one of Black's allowing themselves to be victimized by the White race. The former point of view of powerlessness and marginality is one that is instilled in the main of the Black population. Ralph Ellison illustrates this feeling in the prologue to his Invisible Man, (Ellison, 1953, p. 7). I am an invisible man. . .I am invisible under- stand simply because people refuse to see me like the bodiless heads you see sometime in the circus sideshows, it is as though I have been surrounded by mirrors of bad distorting glass. When they approach me they see only my surroundings, themselves or figments of their imagination. . .indeed, every- thing and anything except me. In their discussion of the roots of racism, Knowles and Prewitt (1969, p. l) have quoted St. Clair Drake in his discussion of this victimization: Negroes in America have been subject to victimiza- tion in the sense that a system of social relations operates in such a way as to deprive them of a chance to share in the more desirable material and nonmaterial products. . . .They are 'victimized', also, because they do not have the same degree of access. . . Fanon (1967, p. 7) has also tapped into this feeling in his introduction as he quotes Aime Sesaire: I am talking of millions of men who have been skillfully injected with fear, inferiority complexes, trepidation, servility, despair, abasement. As we find the Blackman in this marginal status he has two alternatives that marginal men often choose (Fairweather, 1967). He may become hostile and resort to revolutionary tactics or he may become apathetic assuming no responsibility at all. Despite which course is chosen by the Blackman there is a type of powerlessness with which he is faced. This powerlessness is discussed by the National Advisory Commission on Civil Disorders (1968, p. 205). In their reports they point out how the Blackman "lacks the channels of communication, influence and appeal that traditionally have been available to ethnic minorities within the city which enable them-~unburdened by color--to scale the walls of the White ghettos in an earlier era." Researchers in powerlessness and social learning (Seeman, 1963, 1966, 1967; Seeman and Evans, 1962; Nettler, 1957) have been concerned with the effects of powerlessness upon action. As stated in Jordan (1973) a feeling of powerlessness is inversely related to attempts of people to control the environmental factors in their life situations. That is, a high feeling of powerlessness is related to a low level of effort toward manipulating situations. The lack of effort exerted to manipulate unfair situations has resulted in marginal citizenry for the main of the Black populace. The Blackman in his place of marginality must be raised to his rightful position of prominence. As Fairweather (1967, pp. 7—8) states "the problem of our society involves 10 changing both attitudes and behaviors inextricably associated with a marginal status, formarginal man is the product of his society." The writer would like to clarify the meanings that will be taken of "our society" and "his society" in Fairweather's statement in this particular study. The "our society" mentioned must be the Black society for no other will assume such a task. The "his society" mentioned for the most part must mean the White society for this is where much of the control has been maintained. However this is not an attempt to exonerate the Black society completely, for some control of the Blackman's marginality is found in the Black population. Blacks must take some responsibility, because Black- White interactions, as all interactions, have required both parties to "play the game" (Pettigrew, 1964). For "the game" to work Blacks have had to defer and submit to the inferior role prescribed for them by White supremacists. However some Blacks chose not to play the game, a description of one who chose not to play the game is taken from the New York Tribune of April 24, 1889: . . .Before the torch was applied to the pyre the Negro was deprived of his ears, fingers, and other portions of his body with surprising fortitude. Before the body was cool, it was cut to pieces, the bones were crushed into small bits and even the tree upon which the wretch met his fate was torn up and disposed of as souvenirs. The Negro's heart was cut in several pieces, as was also his liver. Those unable to obtain the ghastly relics directly, paid more fortunate possessors extravagant sums for them. Small pieces of bone went for 25 cents and a bit of liver, crisply cooked, for 10 cents. 11 Implications <3f Black Experiences for Internal-External Control Being cognizant of the brief account of Black experi- ences given in the preceding section, consider the impact of experiences on expectancy. If we concur that expectancy is a function of past experiences (Phares, 1957; Crandall et al., 1965) "the game" was anything but conducive to an expectancy for success. Increasing one's objective proba- bility for success does not automatically increase his sub- jective perception of probability for success. One's subjective expectancy may not even be increased in circum- stances where opportunities are rising. Some motivational relearning may be necessary to change expectancies so they conform to the present rather than past realities (Jordan, 1973). In situations where Blacks have had past successful experiences, Blacks exhibit a greater expectancy for success. Lefcourt and Ladwig (1965a) had three groups of Black reformatory inmates compete against White stooges. The subjects were given a task to perform which was falsely represented to involve musical skills. Black inmates were stratified by their jazz experiences into three groups. The experiences ranged: 1) from jazz musicians, 2) to those who were or had been interested in jazz, 3) to those having no history of jazz interests. Individuals in these groups competed against White stooges, and the outcome was 12 contrived so that the White stooges consistantly won. All factors in the study were held constant except prior experi- ences and perceived personal competancy in the task. The jazz musicians, having the greatest experience in the perceived task, persisted in competition longer than the other two groups. In a situation in which one has some basis on which to expect to be successful as in the study presented above, he may be observed to act in a competitive and achievement oriented fashion. However when an individual's expectancies are limited by situational variables, his competitive behaviors and achievement-orientations are altered. Epps (1969) suggests that situational factors for Black and White college students are quite different. The most striking differences in Black and White students as groups are in family income and in occupational and educational levels. The difference on the whole suggest that as groups, (Black) and White students express somewhat differ- ing sets of modal experiences, differing modes of expression and communication and differing ways of viewing the world. These results imply that, in addition to problems of prejudice and discrimination, (Black) students at major colleges may simply find many aspects of the academic community tailored for someone else. (p. 10). As was pointed out by Epps there are some drastic differences in Blacks and Whites, yet these differences have not been widely considered in assessing the expectancies and locus of control for Blacks. It appears that many 13 researchers, some of whom have been more in tune with these Black distinctions, have found I-E control not to be the unitary concept it was presumed to be in earlier research. In the context of children's beliefs, Crandall et a1. (1965) noted the importance of distinguishing different types of external environmental forces. In their View, control by impersonal forces should be separated from control by other people since certain successes and failures may have little to do with chance or luck, but still be subject to external control. Hersch and Scheibe postulated that control should be separated further as they observed that people who score highly externally often exhibit greater variance in behavior than people who score strongly internally. Consequently they concluded that the meaning of externally should be further differentiated. They also stress the need to assess how realistic it is for a person to perceive that events are beyond his control and whether he considers external forces to be benevolent or malevolent. Gurin et a1. (1969) support that possible advantages of external control for Blacks as they point out that the literature has neglected the fact that an internal orientation may have negative implications. An unjustified internal orientation may lead to inappropriate self-derogation and self-blame. For example, in Blacks who have encountered social constraints associated not only with race but also with low-income and lower-class status, an internal orientation based on responsibility for their failures 14 may be more reflective of intrapunitiveness than of efficacy. Gurin et al. (1969) also states that low—income groups experience many external obstacles that have nothing to do with chance. These are class-tied obstacles which may be viewed appropriately by the low-income person as external but not as a matter of randomness or luck. Gurin et a1. (1969), in assessing the attitude structure of Black college students, found the I-E construct not to be a unidimensional concept. By factor analysis, the Gurin study developed a distinction between self and other in one's perception of control. Items with a first person referent were grouped under self or personal control, which is very close to the conceptual definition of internal control given by Rotter (1966). Items with a third-person referent were grouped under ideological control (Protestant Ethic Ideology), which is one's general beliefs about the role of internal and external forces in determining success and failure in the culture at large. The Gurin et a1. identify the separation of personal and ideological levels as vital in differentiating between Black and White populations. They found that Blacks feel as strongly as Whites that people in general control their destinies. However, a difference does appear in responses to questions using personal referent. Blacks are less internal than Whites in answering questions about their own life experiences. Though Blacks adopt general cultural beliefs about internality, their experiences tell them that 15 those beliefs cannot be applied in their own situations. Whites are less likely to perceive this inconsistency between cultural beliefs and what works for them, since they have not had the experiences Blacks have had with discrimination and racial prejudice. Mex The focus of the preceding review has been to set in perspective the internal-external locus of control construct as it relates to Black people as a group. It is not enough to simply dispel the stigma of Blacks as inferior by reason of externality, but it becomes incumbent on this author and other researchers to offer, study, and explain alternate interpretations and motivations for externality (Crandall et al., 1965; Gurin et al., 1969; Hersch and Scheibe, 1969). Furthermore it seems that what Rotter (1966) termed internal or external may not be such a unidimensional concept (Crandall et al., 1965; Gurin et al., 1969; Lao, 1970). This implies that the high incidence of external scores by Blacks on the Rotter Scale may not represent a true reading of externality as it is most commonly construed. In support of this premise, Blacks have been observed to perform internally in situations in which they have some expectancy for success (Lefcourt and Ladwig, 1965; Williams and Stack, 1972). However, when expectancy for success is low, Blacks have Operated in an appropriate external manner (Gurin et al., 1969). This external behavior in Blacks as a group when expectancies for VJ 1 EC Th 513 16 success are low, has been labeled the blocked opportunity theory which is defined as ". . .the prolonged exclusion of (Blacks) from American economy and social life" (Caplan and Paige, 1968, p. 15). Forward and Williams (1970) supported the blocked opportunity theory in their study of Detroit rioters. They found young Black militants to be those who had developed some confidence in their ability to shape events in their own lives if given the chance. From this, it appears that if given the chance to control their own destinies, Blacks behave in an internal fashion. Experimentally, the present study attempts to give Blacks the chance to control their own fate. Involving Black students in a game situation which simulates a reversal of their typical life circumstances presents them with the opportunity to control their situational fate. A simulation game was employed because of its "economy, visibility, reproducibility and safety" (Raser, 1969, p. 41). Economically it is understandable that a model would be cheaper to produce than the event in the natural setting. The simulation aids in highlighting the expectancy phenomenon as well as allowing a construction and reversal of the situ- ation as it exists in the natural milieu. Reversal of a natural situation is much safer and more ethical in a simulation than in the real world. The game allows the Opportunity to "pluck out of social life generally a circum— scribed arena, and attempt to reconstruct the principal ’ 17 rules by which behavior in this arena is governed and the principal rewards that it holds for the participants" (Cole- man, 1966, p. 4). Coleman believes that simulation games and most other games constitute a kind of "caricature of social life." Drawing from Piaget's research, Coleman notes ". . .that for children games are more than a caricature of life; they are an introduction to life--an introduction to the idea of rules. . ." (p. 3). If we are indeed introduced to life and rules through games as Coleman proclaims it would be quite appropriate to return to games when relearning and restructuring is attempted. The game "Simulated Society"(SIMSOC, Gamson, 1972) in particular is valuable in the simulation proposed because it does not constrain the environment and make the player's choices more apparent than real. The environment in SIMSOC is minimally programmed to insure many alternate ways of playing that may work equally well. Gamson (1972) reports that a major design challenge has been to keep the forces in balance so no single course of action appears to be the best for all players. By modifying the SIMSOC procedure slightly, it was possible to elicit the desired effects ethically, economically and safely. One modification of SIMSOC was the delegation of power positions by race rather than by chance. Two separate games were administered, one with Blacks having the power positions and the other with Whites having the power positions. 18 The Black dominated game represented the reversal situation while the White dominated game was the status quo. Operational Hypotheses Game experiences that more closely simulate the natural milieu are expected to have the least effect upon shifts in locus of control. The more diametrically opposed the game experience is to the natural milieu the more effect it has on the shift in locus of control, provided the game experi- ence is not so removed from the natural experience that it cannot be absorbed. A greater shift in Black externals toward internality is expected than White internals toward externality. In addition Blacks who are not awarded personal power in the Black dominated game experience, will shift toward internality through identification with their racial peers. Life experiences in the natural situation may be dif- ficult to counteract with a brief reversal game experience. Therefore situational measures may indicate more of a change than generalized measures. Experimental Hypotheses 1. Blacks in power positions in the Black game should have more movement toward internality than any other treatment group. 2. Blacks without power positions, in the Black game should have a shift toward internality through identification with Blacks in power. 19 3. Whites without power positions in the Black game may experience a slight shift toward externality. 4. No substantial shifts are eXpected in the White game. CHAPTER I I METHOD This study is based on the premise that Internal- External Control is learned construct and is a function of one's past experiences. Therefore, if one is given the appropriate experiences, it is quite possible to increase one's internality. ‘In illustration of this point, the plan of this study was to situationally induce internality in Afro-American by experimentally altering their experiences. These new experiences were conferred through participation in a simulated society game (SIMSOC). Shifts in generalized expectancy as a result of these experiences were measured by the Rotter I-E Scale. Observational ratings were employed as a potential measure of the effects of the experiences on situational expectancy. Subjects The participants in the study were drawn from members of introductory psychology courses at Michigan State University. Students were issued a brief demographic questionnaire and the Rotter I-E Scale in their psychOlogy classes. From this pool of 700 students, 71 subjects were randomly selected with the restriction of predetermined sex and race ratios. All subjects 20 21 were paid two dollars and were offered five extra credit points. The compensation received by subjects was independent of their roles or degree of participation in the game. Experimenters There were six experimenters involved in the research. Two experimenters served as the coordinators of SIMSOC as outlined by Gamson (1972). The other four experimenters rated the situational behaviors of the players. The experimenters were selected to be as heterogeneous as possible in respect to race and sex. This heterogeneous selection was in keeping with that of the subject population and was a control for the extraneous variables which may have been introduced by a homogeneous sexual or racial hierarchy. Therefore the experimenters were composed of two Black males, one White male, one Black female, and two White females. Five of the experimenters were undergraduate students at Michigan State University. The author served as one of the coordinators but had no input in the observational ratings. Instruments The nature of the present study, being situational and counterbalanced, called for little demographic material. Information deemed necessary for this study included age, sex, and race. Other information such as name, address, phone number and convenient times to be called were included to aid in administration but were not included in any analysis (Appendix A). 22 Rotter Internal-External Locus of Control Scale The Internal-External Locus of Control Scale (Rotter, 1966) referred to as the I-E Scale is a 29 item forced-choice test. Included in the 29 items are six filler items intended to make the purpose of the test ambiguous. The test is scored by the total number of external choices, and is con- sidered to be a measure of generalized expectancy. "The items are constructed to deal exclusively with the subject's belief about the nature of the world" (p. 10). The internal consistency of the test has been shown to be stable in various samplings. The test-retest reliability for combined males and females over a one month period was found to be .72. The test shows reasonable homogenity considering that items are sampling broadly generalized characteristics over a number of different situations. The test appears to be more valuable in the investigation of group differences than for individual prediction (Appendix B). SIMSOC SIMSOC (Simulated Society) is a game developed by Gamson (1972) to focus on the establishment and maintenance of social order. It creates a situation in which the partici- pant must actively question the nature of social order and examine the processes of social conflict and social control. SIMSOC does not attempt to emulate a real society in every respect, but characteristics are included to highlight certain issues and problems. 23 SIMSOC is characterized by a mixture of mutual depen- dence and conflict, of partnership and competition among the players. It requires participants to confront certain central problems of collective decision-making as an organizational problem. SIMSOC offers at least three central focuses: l) the processes of large-scale conflict, protest, social control and social change; 2) the exploration of interpersonal feelings, communication trust and other aspects of face-to- face interaction; 3) the challenge of creating utopia. Despite the variability of forms taken by SIMSOC, all SIMSOC games generally pass through three phases: 1) problems of scarcity, 2) problems of power and authority, and 3) pro- blems of prosperity. Though the visibility of these phases may vary for different administrations, it usually is possible to observe all three in some form. Procedure Each subject had previously completed a demographic questionnaire and the Rotter I-E Scale which had been dis- tributed in their psychology courses about 6 weeks earlier. From this information subjects were randomly assigned to one of two groups to represent the greatest heterogenity with respect to sex and race. Both of the groups discussed below engaged in a modified version of SIMSOC. A set of instructions (Appendix C) continaing the modifications was distributed and discussed prior to beginning the game. 24 One modification of SIMSOC which was not discussed with the subjects was the delegation of power position by race rather than by chance. The first game, which somewhat represented the status quo, engaged Whites in the power positions while Blacks held the nonpower positions. The second game, representing the reversal experience, employed Blacks in the power positions with White subjects in the nonpower positions. In each game there was a minority of the subordinate population who were delegated to power positions. Power was defined as being head of one of the basic groups (Basin, Pop, Empin, Masmed, and Judco, see Appendix C), or possessing a travel or subsistence agency. In addition the three regions in which players "lived" were a function of their power positions. The green region was completely composed of the dominant race who held most of the power. The yellow region was racially integrated and was moderately powerful. The red region was totally cOmposed of the sub- ordinate race and was void of any power. Each game was composed of four half—hour sessions with a five minute break between sessions. One of the purposes of the break was to allow the coordinators time to tabulate game statistics (national indicators). The game statistics were necessary for the operation of the game for each sub- sequent session; however, there was no analysis of nor inferences drawn from this data. To control for the apathy that is generated when 25 players believe the society is coming to an end (Gamson, 1972), participants were told there would be five sessions rather than four. The time allowed for the fifth session was used by the coordinators to re-administer the Rotter I-E Scale, to pay subjects and to sign extra-credit cards. Before any formal training of coordinators and raters was undertaken, the experimenters were instructed to read Gamson's (1972) participant's manual (coordinators read the instructor's manual as well as the participant's manual). Subsequently the experimenters were participants in a SIMSOC game administered as part of the curriculum of an advanced undergraduate sociology class at Michigan State University. Following these initial experiences, raters received extensive training on a video tape of one region's responses in a previous SIMSOC. During this segment the experimenters were monitored for improper scoring and instructed in correct responses. Improper scoring consisted of either choosing the incorrect locus of control or choosing an unrelated guideline statement (Appendix D). After experimenters became clear on the task they were to perform, the format shifted to an examination of rater agreement. Raters then scored new, five minute segments of the video tape to establish inter- rater reliability. Raters had to agree on all three of the following criteria in order to demonstrate agreement: 1) the specific taped statement which was rated, 2) whether the statement was internal or external, 3) and which 26 guideline statement was the model. The raters reached an average of 73 percent agreement on the last of 8 days of training. Subsequent to training, yet prior to the experi- ment, the experimenters assumed in a pilot game the roles they would perform later in the actual experiment. In the pilot as well as the actual study, the coordina- tors were positioned in the corridor so that they were centrally located between the three classrooms which housed the three regions. The location of this station was important to insure that a travelling player would pass the coordinators before entering another region. The function of the coor- dinator was to collect travel tickets, subsistence tickets, tabulate national indicators, serve as the broker for Basin and answer procedual questions. The coordinators were guided by the SIMSOC instructions (Appendix C), participants manual and instructors manual (Gamson, 1972). The raters were stationed in the individual regions, one rater to each region (with the extra rater designated as an alternate). The raters identified and scored subjects' statements as internal or external. The statements had to be directly related to a statement on the "Rating Guidelines" (Appendix D) before they could be scored. Statements that were perceived as internal or external, but could not be related to a specific model statement in the guidelines, were not scored. The raters operated on alternating on-off five minute segments. Any statement made during an off segment was not scored regardless of its content. CHAPTER III RESULTS Biases in Interpreting the Results Ineqpality of Groups Randomization was employed in assigning individuals to groups within sex and race. This consequently meant that power positions were randomly distributed among Whites (and a few Blacks) in the White game, and randomly distributed among Blacks (and a few Whites) in the Black game. Despite the randomization there were observable pre-test differences between groups. Initial differences were observed between White and Black subjects on the Rotter I-E Scale: as expected the mean score of White subjects was more internal (lower)than the mean score of Black subjects, (Table 1). In addition to the expected differences in racial groups, all power groups were uniformally more internal from the beginning than nonpower groups, though this difference did not reach significance (Table 1). This phenomenon was unanimously upheld across groups, which obviously presents problems in inter- preting post-test data. It is quite apparent that one cannot conclude internal control is a function of power by inspecting post-test data alone. However, it is possible to consider the amount of movement from pre-test to post-test. A procedure 27 28 of this nature controls for the initial starting point of a group by measuring the amount of movement toward internality rather than the final or post-test score of internality. Mathematically this procedure may be sound, however theoreti- cally it is confounding. If one hypothesizes that internality covaries with power such that an increase in power increase internality, it would follow that an internal individual has had prior exposure to power. Thus, a treatment of power upon an individual accustomed to power may not move him along the gradient toward internality at the same rate as an individual unaccustomed to power. This resounds the principle that the value of a reinforcer is dependent upon the indi- vidual's past experiences. In addition, internal groups have less variance of movement available to them than external groups. These factors may be suppressing the effects of power on this study since the power groups were initially more internal. TABLE 1 PRE-TEST MEAN DIFFERENCES ON ROTTER I-E SCALE FOR BLACK AND WHITE POWER GROUPS Power Power-Position Non-Power Position 5 = 1.57 p = NS Blacks 12.33 13.7 Overall Blacks 13.05 Race Whites 9.69 12.19 Overall Whites 11.07 Overall Power 10.77 Overall Non-Power 12.77 29 Incongpuent Measures The main measures of the study, the Rotter I-E Scale and the observational, behavioral ratings seemed to be mea- suring two distinctly different phenomena. The correlations of these two measures were not significantly different from zero. The correlation coefficient of behavioral ratings and Rotter pre-test scores was 1 = —0.0323; Of behavioral ratings and Rotter post-test scores, I = -0.0972. It is quite apparent that these correlation coefficients do not represent any relationship between measures; however, it is not clear what implication should be drawn from these coefficients. Possibly the phenomena being measured were totally different, or perhaps there is quite a discrepancy in what people are observed to do, compared to what people report. Treatment Effects Hypothesis Related Analysis As will become evident later the main and interaction effects of the ANOVA analyses were not strikingly significant. However, movement toward internality on the Rotter Scale was Observed in certain key groups which were directly supportive of the hypotheses. Since these Observations were felt to be of prime importance they were made despite the outcome of the corresponding overall F test (Winer, 1962). Student's 3 distribution was employed to evaluate the significance Of movement in the individual comparisons indicated by hypotheses 30 one and two. An orthogonal comparison was used to test hypothesis four. Hypothesis I: The first hypothesis states: Blacks in power positions in the Black game should have more movement toward internality than any other treatment group. Hypothesis one is clearly supported as the movement toward internality of Blacks with Power in the Black Game is significant p < .025. Furthermore this is the only group which has significant movement toward internality (lower scores) on the Rotter Scale (Table 2). EMHE 2 GHDUPDDVEMEIVI'S'IWARDINIEWAII'I'Y ONTHEROI'I‘ERI-ESCAIE anps Marara MamrPafl: Mowment Blacks with Power-Black Game 12.63 9.88 p = 2.815 df = B < .025 Blacks without Power-Black Game 14.6 12.0 p = 3.0 df = p:<.09 Whites without Power-Black Game 10.91 9.36 Wrong by inSpection Whites*with Pcwer‘White Game 10.6 11.0 WhiteS‘withoutiPowerHWhite Game 15.0 16.4 BlackS‘without.PowerHWhite Game 13.29 12.0 31 Hypothesis II: The second hypothesis states: Blacks without power positions in the Black game should have a shift toward internality through identification with Blacks in power. Hypothesis two was not clearly supported on the Rotter Scale since Blacks without power in the Black Game did not move significantly toward internality (lower scores). How- ever there was an observable trend in the predicted direction p < .09. Though significance was not reached the data was conSistent with the direction of prediction (Table 2). Hypothesis III: The third hypothesis states: Whites without power positions in the Black game may experience a slight shift toward externality. Hypothesis three was not supported by movement on the Rotter Scale. Though the movement that did occur was not significant it was in the internal direction (lower) while the hypothesis was in the external direction. The movement data of this group was not even consistent with the predicted direction (Table 2). Hypothesis IV: The fourth hypothesis states: No substantial shifts are expected in the White game. Hypothesis four was tested and tentatively supported by the orthogonal comparison of movement in pre-test--post- test means on the Rotter I-E Scale. The group means tested were the groups which participated in the White Game. The orthogonal comparison was made because the hypothesis was not specific to any particular group but an overall prediction 32 of all groups in the White Game. As predicted there was no significant movement in the White Game (Table 2). ANOVA Analysis: Rotter I-E Scale and Behavioral Ratings Main effects. The design of the analysis was essentially a 2 X 2 X 2 X 2 factorial on the Rotter I-E Scale and a 2 X 2 X 2 factorial on the behavioral ratings. Given a near zero correlation between the instruments, the analyses were handled separately. The four main effects of the Rotter Scale ANOVA were: (Table 3) 1) Race (A) 2) Power (B) Blacks (Al) Power Position (B ) Whites (A2) Non-Power Position (B2) 3) Game (C) 4) Time (D) Pre-Test (D Black Game (C ) Post-Test (D2) ) White Game (Cl 2) The main effects of the behavioral ratings took the same dimensions as the Rotter with the exception of Time. Time was excluded due to the situational nature of the ratings (Table 4). An inspection of the Behavioral Rating ANOVA summary table (Table 4) will not indicate any significnat main effects in the rating analysis. The Rotter analysis indicates a significant effect of Power (p < .05, Table 3). Power appears to be the most potent single treatment. subjects in power position tested more internally (lower) on the Rotter Scale than subjects in nonpower positions, across the other three treatment variables. However, as previously stated the subjects who were placed in power SUMMARY OF ANOVA FOR ROTTER I-E SCALE 33 TABLE 3 Source df Ms F Race (A) 1 7.6723 <1 Power (B) 1 152.3253 4.43* Game (C) 1 41.4070 1.20 A X B 1 6.2336 <1 A X C 1 121.9227 3.55** B X C 1 7.5682 <1 A X B X C 1 0.5535 <1 Error I 40 34.3779 Time (D) 1 15.2135 2.1393*** A X D 1 11.3130 1.5908*** B X D 1 0.0011 <1 C X D 1 11.3762 l.5997*** A X B X D 1 0.0504 <1 A X C X D 1 0.1793 <1 B X C X D 1 0.4988 <1 A X B X C X D 1 1.1152 <1 Error II 40 7.1114 *p < .05 **p < .10 ***E < .25 34 positions were initially more internal than those in nonpower positions. Therefore, the true significance of power may not be represented by this analysis. TABLE 4 SUMMARY OF ANOVA FOR BEHAVIORAL RATINGS Source df MS F Race (A) 1 .011875 <1 Power (B) 1 .002157 <1 Game (C) 1 .075806 <1 A X B 1 .397610 5.234259* A X C 1 .197642 2.601819** B X C 1 .00070 <1 A X B X C 1 .177850 2.341271** Error 39 .075963 *p < .05 **p < .25 Interaction effects. Although the power variable on the Rotter was the only main effect to reach significance, there were no interactions with power. A Race X Power interaction did gain significance on the behavioral ratings (p.< .05), yet a test of simple effects could not attribute the variance to a specific effect, but attributed the variance to the overall effect of the interaction (Table A1, Appendix E). Though the simple effects test failed significance, the 35 table of means indicates that Blacks in power and Whites not in power are more internal than Blacks not in power and Whites in power (Table 5). TABLE 5 MEANS OF RACE X POWER (RATINGS) Power (B) Power Non-Power Position (B1) Position (B2) Blacks (Al) .338295 .557489 Race (A) Whites (A2) .533974 .279852 The Race by Game interaction presents an interesting theoretical contradiction in group assignment (p < .10). The data of the Rotter Scale would indicate that the setting or game controlled by the other race would be more conducive to internal.control(Tab1e 6). However, since this effect does not involve an interaction with time it is difficult to assume that the data represents a change. Therefore an excessive amount of internals were assigned to the game controlled by the other race. The greatest amount of variance was found to be between Whites in the White Game and Whites in the Black Game (p < .01, Table A2, Appendix E). Some degree of variance is also attributed to the difference between Blacks in the Black Game and Whites in the Black Game (p < .05, Table A2, Appendix E). 36 TABLE 6 MEANS OF RACE X GAME (ROTTER) OVER PRE-TEST AND POST-TEST SCORES Game (C) Black Game (Cl) White Game (C2) Blacks (A1) 12.3 11.07 Race (A) Whites (A2) 8.63 13.27 Collgpsed Treatment Effects It was evident that certain treatment groups had small sample sizes which may have had an adverse effect upon signi- ficance. These small treatment cells were eliminated by._ ignoring the game effect which would have two sample popula- tions in each remaining condition. The cell having the largest population was retained while its parallel cell was eliminated. The groups retained from the Black Game were Blacks with Power and Whites without Power, those retained from the White Game were Blacks without Power and Whites with Power. The design was subsequently altered to a 2 X 2 X 2 factorial in the Rotter Scale analysis, and a 2 X 2 factorial in the behavioral rating analysis. Since the game effect was included in the original design these proposed analyses would be posteriori considerations. An experimenter may make priori and posteriori comparisons in the same experiment, after the experimenter makes the planned 37 comparisons (Kirk, 1968). The focus of these posteriori comparisons is on the Race and Power dimensions on both the Rotter Scale and Behavioral Rating measures. The Time dimension is also included on the Rotter measurement. Con- sequently there is no consideration of the effect Of Game, but only the effect of Race, Power and Time. Analysis of the posteriori comparisons (Table 7 and 8) failed to exhibit greater significance on the H test than the priori comparisons. The overall H test of the posteriori comparisons displays considerably less significance than the priori comparisons. However there was an interesting finding. The main effect of Time increased from nonsigni- ficance (p < .25, Table 3) to a trend in support of the predictions (p < .10, Table 7). TABLE 7 SUMMARY OF ANOVA FOR ROTTER I-E SCALE (WITH THE EXCLUSION OF GAME) Source df MS F Race (A) 1 37.425 <1 Power (B) 1 2.2375