322% WNHWWWWWIIIUHIHW‘WHI‘WWI ”THS ‘— IHESIB ._ .‘w __:_,‘, LIBR A R Y ' Michigan State University This is to certify that the thesis entitled THE EFFECTS OF TYPE OF INFORMATION, DOGMATISM, AND SEX OF DYAD ON PREDICTIVE ACCURACY presented by Robert William Bundens has been accepted towards fulfillment of the requirements for MA. Jggree in Communication flaw @3771/@ Major professor DateJannaxy 7 , 1980 0-7639 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return to remove charge from circulation records THE EFFECTS OF TYPE OF INFORMATION, DOGMATISM, AND SEX OF DYAD 0N PREDICTIVE ACCURACY BY Robert William Bundens A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1980 ABSTRACT THE EFFECTS OF TYPE OF INFORMATION, DOGMATISM, AND SEX OF DYAD ON PREDICTIVE ACCURACY BY Robert William Bundens Miller and Steinberg's (1975) theoretical perspective of interpersonal communication differentiates noninterper- sonal and interpersonal communication on the basis of the type of information individuals utilize to make predictions. These scholars argue that the type of information solicited by an individual and his cognitive style will affect his ability to make accurate predictions about a message source. The present study focused on empirically testing Miller and Steinberg's conceptualization by examining the effects of sociological and psychological information on predictive accuracy. In addition, the effects of dogmatism and the sex of dyad on predictive accuracy were assessed. Students from undergraduate classes at Michigan State University were exposed to a videotape of a male or female stimulus person presenting sociological or psychologi- cal information about themself. Troldahl and Powell's Robert William Bundens (1965) short-form dogmatism scale was administered to assess subjects' dogmatism. The results indicated that subjects exposed to psychological information are more accurate predictors of a stimulus person's responses than subjects exposed to sociological data. No significant effect were obtained for dogmatism or the sex of dyad. These findings are discussed in terms of their implications for future research. Accepted by the faculty of the Department of Com- munication, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Master of Arts degree. flM/éQ f W ’1 1,431 Directof of Thesis ’ 47 fZ Guidance Committee: a4%%2%N1}/firfl?/&%¢' 3?, Chairman v'gmmm/ f JJL Belg m. Hag 31w W ii Dedicated to Mom and Dad iii ACKNOWLEDGMENTS I wish to express my sincere gratitude to the members of my committee: To Dr. Gerald R. Miller, chairman of my committee, I extend my deep appreciation for his intellectual stimulation and guidance. I thank him for his encouragement, patience, and treating me with fair- ness. To Dr. Norman E. Fontes, I extend my sincere grati- tude for instilling in me a desire for scholarship and intellectual rigor. I thank him for his friend- ship and understanding. He had faith in me when I had very little in myself. I am truly indebted to him. To Dr. Hal Hepler, I extend my thanks for his help and understanding. I would like to also thank Jennifer Shelby for her constant interest, advice and friendship. Finally, I would like to express a special thanks to Kathy Ayres for her love and encouragement. iv TABLE OF CONTENTS Page LIST OF TABLES O I O O O O O O O O O O O Vii Chapter I. INTRODUCTION . . . . . . . . . . . 1 II 0 PROCEDURES 0 O O O O O O O O O O O 19 Definitions . . . . . . . . . . . 19 Design. . . . . . . . . . . 24 Preparing the Stimuli. . . . . . . 24 Instructions to the Subjects . . . . . 25 Description of Subjects . . . . . . . 25 III 0 RESULTS 0 O O O O O O O O C O O O 26 Iv. DISCUSSION 0 O O O O O O O O O O 0 3S APPENDICES Appendix A. Transcript of Interview Soliciting Sociological Information from a Female Stimulus Person . 41 B. Transcript of Interview Soliciting Psychological Information from a Female Stimulus Person . 42 C. Transcript of Interview Soliticing Sociological Information from a Male Stimulus Person. . 44 D. Transcript of Interview Soliciting Psychological Information from a Male Stimulus Person. . 45 E. Predictive Accuracy Scale for Female Stimulus Person. . o o . O O O O I O O I 48 Chapter Page F. Predictive Accuracy Scale for Male Stimulus Person 0 O O O O O O O I O O O O 55 G. Troldahl and Powell Short-Form Dogmatism Scale . . . . . . . . . . . . . 62 FOOTNOTES 0 O O O O O O O I O O O O O O 65 BIBLIOGRAPHY O O O O O O O O O O O I O O 66 vi Table 1. LIST OF TABLES Zero-Order Correlations Among Independent variables 0 I O O O O O O O O O O Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, and Coefficients of Determination of Predictive Accuracy Regressed on Type of Information, Dogmatism and Sex of Dyad (N=160) . . . . Zero-Order Correlations of Main Effect Variables and Interaction Terms Using Dummy Coding. . Zero-Order Correlations of Main Effect Variables and Interaction Terms Using Orthogonal Coding . . . . . . . . . . . . . Prediction of Individual's Total Accuracy of Prediction Using Type of Information, Dogmatism, Sex of Dyad, Type of Information X Doqmatism, and Type of Information X Sex of Dyad (N=160) o o o o o o o o 0 vii Page 28 28 30 32 33 CHAPTER I INTRODUCTION Several theoretical frameworks have been developed to analyze various types of communication endemic to rela— tional development. Social penetration theory (Altman and Taylor, 1973) assumes that "interpersonal exchange gradually progresses from superficial, nonintimate areas to more intimate, deeper layers of the selves of the social actors" (p. 6). Duck's (1973) theroetical analysis focuses on the acquaintance process and how people utilize various types of information to assess positive and negative aspects of relationship formation. Berger and Calabrese (1975) state that as relationships develop, information which is exchanged moves sequentially from nonintimate to intimate tOpics. Wright (1978) contends that "friendship develops from an opportunity in interpersonal contact through a prelude in friendly relations to varying levels of voluntary inter- dependence" (p. 201). He goes on to argue that self dis- closure increases as relationships progress from initial contact to friendship. Miller and Steinberg (1975) distin- guish noninterpersonal and interpersonal communication on the basis of the type of information exchanged by inter- actants. At the noninterpersonal level of communication, information about one's cultural and sociological membership in groups is exchanged. At the interpersonal level of communication, psychological data specific to an individual's past experiences is exchanged. All of these conceptual frameworks are predicated on at least three assumptions: (1) Relationships have different levels of intimacy; (2) in- formation concerning oneself and others is exchanged in relationships; and (3) different information is exchanged at different levels of intimacy. Miller and Steinberg (1975) delineate three types of information which may be solicited from an individual and which differentiates noninterpersonal from interpersonal communication. The information solicited may include cultural, sociological and/or psychological data about a person. Cultural information is data based on the norms, values and customs peculiar to a particular culture. "I sleep on a 'tatami' mat placed on the floor," is an example of information based upon customary Japanese sleeping habits. Sociological information refers to data about a person's membership in certain social groups. "I attend regular services at a Catholic Church," is an example of sociologi- cal information indicating membership in the Catholic Church. Both cultural and sociological information are not idiosyn- cratic to the individual who reveals such information but is generalizable to other members of the individual's cultural and/or sociological group. Both cultural and sociological information are characteristic of noninterpersonal communi- cation. Psychological information, however, is data con- cerning the unique individual learning experiences of a person. "I have an irrational fear that I will die from cancer," is an example of psychological information. Psy- chological information exchange typifies interpersonal com- munication. Miller and Steinberg (1975) contend that in relational developments, cultural information generally precedes sociological information, which precedes psychologi- cal information. The sequential receipt of information in relation- ship formation has been suggested by several researchers (Berger, Gardner, Clatterbuck, and Schulman, 1976; Berger and Larimer, 1974; Larimer and Berger, 1974; Taylor and Altman, 1966), where informational content was rated in terms of nonintimate versus intimate or biographic—demographic versus Opinions and values, rather than cultural, sociologi- cal or psychological in nature. While not directly con- firming the sequential solicitation of various types of information in relationship deve10pment suggested by Miller and Steinberg, these studies do lend some support to their framework since biographic-demographic (nonintimate) infor- mation is typically more cultural and sociological in nature, and Opinions and values (intimate information) are character- istically more psychological. There has been increasing criticism (e.g., Bochner, 1978) that the developmental framework of interpersonal com— munication offered by Miller and Steinberg is conceptually limited. One of the general arguments Bochner presents is the lack of explicitness to be found in developmental models of communication: "The Miller and Steinberg approach does not identify the mechanisms of change which lead to deve10p- ment. Nor does it explain many important relational pro- cesses" (Bochner, 1978, pp. 186-187). While this inquiry is not a rebuttal to the criticisms offered, until a con- ceptual framework is empiricially tested it is difficult to conceptually evaluate it. To date there has been virtually no empirical research on the deveIOpmental process of inter- personal communication proposed by Miller and Steinberg. The purpose of this inquiry is to empirically test and evaluate Miller and Steinberg's framework by examining the relationship between types of information and the accuracy of predicting an individual's behavior. Since pre-test data by Fontes (1978) indicates that individuals are not able to clearly distinguish between cultural and sociological information, this inquiry focuses on sociologi- cal and psychological types of information and predictive accuracy. The effects of coqnitive style, specifically dogmatism, on type of information and its subsequent effect on the accuracy of predicting a message source's future behavior is examined. That the similarity of sexes facilitates predictive accuracy in addition to the type of information is also explored. The assumption that "when people communicate, they make predictions about the effects, or outcomes, of their communication behavior" is crucial to the conceptual frame— work of interpersonal communication offered by Miller and Steinberg (1975, p. 7). The exchange of psychological in- formation about the unique characteristics, beliefs and behaviors of an individual is necessary to move from a non- interpersonal to an interpersonal relationship. Such infor- mation provides a basis for making predictions about the responses to our communicative behavior and relationships. In addition, this information is necessary in selecting an apprOpriate communication strategy from the alternative strategies available to us for any given transaction. The importance of predictions in relationships has been noted by several researchers. For example, the impor- tance of obtaining and utilizing information to predict future communicative transactions and outcomes is stressed by Altman and Taylor (1973) when they state that forecasts involve estimations of potential outcomes in areas of more intimate exchange, a fact which results in the relationship moving forward to new and potentially more satisfying interactions. . . . Continuous exchange at the same level of intimacy is also part of a "working through" process, in which each person increases his knowledge of the other and builds a more accurate cognitive picture of the other person. This facilitates more accurate forecasting to unexplored areas of interaction (p. 39). Berger and Calabrese (1975) state that during the course of interactions various information is exchanged which facilitates inferences about attitudes and beliefs not yet explicitly revealed. Berger and Calabrese, in a theory of initial interaction, posit the fundamental assump- tion that "when strangers meet, their primary concern is one of uncertainty reduction or increasing predictability about the behavior of both themselves and others in the interaction" (p. 100). These scholars use the concept of uncertainty reduction in a dual sense. First, it refers to a "proactive" process of developing predictions about a person's behavior. That is, uncertainty reduction is con- ceived as an analysis of possible ways in which each inter- actant might behave in response to one another. This analy- sis facilitates the selection and prediction of the other person's behavior. The second way Berger and Calabrese utilize the concept of uncertainty reduction concerns the "retroactive explanation" of a message source's behavior. Specifically, a message recipient attempts to reduce the number of plausible alternative explanations for the message source's past behavior. There is a fair amount of uncertainty present when individuals engage in initial interactions. To reduce this uncertainty, information is exchanged. However, since the consequences of exchanging intimate information during initial interactions may prove costly, it would be reasonable to expect nonintimate information to be exchanged. Altman and Taylor (1973) contend that "because of uncertainty of outcomes, it is expected that most people adOpt a conserva- tive approach, moving into areas of exchange where outcomes are better known" (p. 41). With the acquisition of infor- mation about a specific person's beliefs, attitudes and per- sonality, it would be reasonable to assume an increase in accurately predicting the behavior of that individual in future interactions. Empirical support for this position comes from research executed by Taft (1966) who found that personality trait predictions of casual acquaintances are more accurate than those of nonacquaintances. Research by Hjelle (1968) also supports this position. He found that intimate acquaintances significantly predicted their partner's social and personality responses more accurately than either casual acquaintances or nonacquaintances. Miller and Steinberg's (1975) conceptual framework is congruent with other researchers in that it focuses on different types of information used to make predictions. The exchange of sociological or psychological information during interactions may decrease uncertainty but will have different consequences on predictive accuracy because of the nature of the information. With the receipt of socio- logical information, the likelihood of engaging in social categorization occurs in which the individual is not con- sidered, and not reacted to, as a unique person different from other individuals with the same sociological background. This categorization process is a basis for stereotyping and is a fundamental basis of Miller and Steinberg's conception of stimulus generalization in relational interactions. Stimulus generalization occurs when a message recipient receives information about a group of people and selects traits common to them all while ignoring the differences between persons. The message recipient then ascribes to each member of the group the common traits. Miller and Steinberg point out that predictions based on stimulus gen- eralization are subject to a high error rate. Miller and Steinberg contend that when individuals receive psychological information, the person making pre- dictions will be capable of making stimulus discriminations, seeking out relevant differences, and recognizing the unique characteristics of the communicator. Thus, Miller and Stein- berg argue that with psychological information, the predictor will be able to make more accurate predictions about a person because he will engage in stimulus discrimination rather than stimulus generalization. Based upon this reasoning the following hypothesis was generated: H1: Individuals exposed to psychological information about a person w111 pred1ct that person's behav1or more accurately than individuals exposed to socio- logical information about a person. Given the various types of information which are available to predict another person's behavior, an empirical assessment of cognitive style upon the utilization of in- formation available and its consequences on the accuracy of predicting a message source's behavior becomes relevant. Cognitive style refers to "characteristic ways in which an individual structures his/her beliefs and attitudes about the world, and to the ways he processes and reSponds to incoming information" (Miller and Steinberg, 1975, p. 144). A central premise which Miller and Steinberg discuss is that movement to the interpersonal level of communication is facilitated by certain cognitive styles and inhibited by other styles. Congruent with this position are studies by Ehrlich (1973) and Dornbusch, Hastorf, Richardson, Muzzy, and Vreeland (1965) which strongly indicate that some indi- viduals enter relationships with stereotypes and expecta- tions about peOple which leads them to classify individuals into social categories. Ehrlich (1973) argues that the classification of social objects into groups reflects societal and individual responses to the complexity of the environment. He titles these shared classifications "social categories." Social categories consist of persons or objects in which the indi- vidual differences are minimized and the categorical charac- teristics are maximized. Ehrlich contends that this minimax process facilitates the storage and recall of information in order for a person to act. Research by Dornbusch gt El: (1965) found that past eXperience and a person's cognitive structure affect the selection of information. These researchers suggested that a perceiver actively engages in selecting and organizing information about the characteristics of people he encounters 10 into categories which provide some utility to him/her. As Wegner and Vallacher (1977) point out in regard to the study by Dornbusch gt 31. (1965), Each perceiver selects a consistent set or categories of information in each person he attempts to understand. Every time he meets someone new, instead of just noticing the particularly outstanding features or behaviors of that person, he also notices whether the person has a certain set of qualities that he looks for in everyone (p. 92). If the interpersonal level of communication is characterized by psychological information with which dis- criminating features or behaviors among individuals can be identified, it is probable that certain cognitive styles will facilitate the use of psychological information while others will not. One form of cognitive style expected to influence the utilization of psychological information to make accurate predictions is dogmatism. Dogmatism is defined by Rokeach (1956) as a relatively closed cognitive organization of beliefs and disbeliefs about reality, organized around a central set of beliefs about absolute authority which, in turn, provide a framework of intolerance and quali- fied tolerance toward others (p. 3). Rokeach states that in general, a person must process or code information in such a way that the information is either rejected or fitted into his/her belief-disbelief system. Because of the rigidity of their cognitive style, high dogmatics may have difficulty processing and utilizing psychological information when it is presented to them. In such a case, it would be anticipated that this inability to ll process and use psychological information would influence the accuracy of their predictions. Research by Plant, Telford, and Thomas (1965) suggests that high dogmatics are impulsive, conventional, and stereotyped in their thinking. That high dogmatics engage in coding peOple into socially polarized categories has been demonstrated by Rokeach (1960), and Kirtley and Harkless (1969). White, Alter, and Rardin (1965) also found evidence that high and low dogmatics differ in their categorizing behavior. These researchers had high and low dogmatics freely sort occupational titles in terms of pres- tige. The two groups were also given a list of undesirable social acts and asked to sort the items into piles repre- senting undesirability, creating as many categories as they wished. The results indicated that high dogmatics utilized fewer and broader categories in their sorting behavior. These findings suggest that highly dogmatic indi- viduals have difficulty dealing with novel psychological information, and may rely on stimulus generalization rather than stimulus discrimination in relational interactions. If this is the case, highly dogmatic individuals will not make very accurate individual predictions at the interper- sonal level. It may be the case that for high dogmatic persons, psychological information will be either ignored or made to conform to stereotypes elicited by sociological information. 12 In addition, at the noninterpersonal level of com- munication, high dogmatics would have more difficulty pro- cessing and utilizing sociological information to make accurate predictions than their low dogmatic counterparts. High dogmatics with sociological information would resort to classifying individuals into broad and fewer social cate- gories. Low dogmatics with sociological information, how- ever, would utilize more discriminating categories. Thus, at the psychological and sociological levels of information, low dogmatics would be able to engage in stimulus discimina- tion while high dogmatics would rely upon stimulus generali- zation. Based upon this reasoning the following hypothesis was posited: H2: Regardless of the type of information individuals are exposed to, the h1gher the dogmat1sm the lower their accuracy of prediction. In considering the joint effects of type of infor- mation and dogmatism it is anticipated that low dogmatics with psychological information will make more accurate predictions than high dogmatics with psychological informa- tion. High dogmatics exposed to psychological information will utilize the information to a limited extent and will consequently make more accurate predictions than low dogmatics exposed to socioloqical information. Note that it is assumed that type of information contributes significantly more to the prediction of predictive accuracy than dogmatism. Low dogmatics exposed to sociological data however will make 13 more accurate predictions than high dogmatics exposed to sociological data. This latter assertion is based on prior reasoning that low dogmatics with sociological information will refrain from invoking stereotypes to predict a person's behavior while a high dOgmatic with sociological information will tend to classify the individual into broad and rigid social categories and base future predictions using those categories. Based upon hypotheses one and two, and the rationale presented above the following hypothesis positing an ordinal interaction between level of information and dogmatism was generated: H3: Individuals with a lesser degree of dogmatism exposed to psychological information will make more accurate predictions than individuals with a greater degree of dogmatism exposed to psycho- logical information, who will make more accurate predictions than individuals with a lesser degree of dogmatism exposed to sociological data, while individuals with a greater degree of dogmatism exposed to sociological information will make the least accurate predictions. The predicted pattern of means hypothesized to emerge is: X1>X2>X3>X4 where i1 = mean of individuals with a lesser degree of dogmatism exposed to psychological information. xl ll 2 mean of individuals with a greater degree of doqmatism exposed to psychological information. xl u 3 mean of individuals with a lesser degree of dogmatism exposed to sociological information. xl ll 4 mean of individuals with a greater degree of dogmatism exposed to sociological information. 14 Another variable which may affect predictive accuracy during social interaction is the sex of the interactants. Paralinguistic, facial affect, and kinesmorphic cues are largely determined by sex (e.g., Birdwhistell, 1974; Buck, Savin, Miller, and Caul, 1974) and provide valuable initial information about an individual. Knowledge about the sex of an individual may determine how subsequent information is processed and influence expectations concerning future interactions. According to the Miller and Steinberg paradigm, the sex of a person is classified as sociological information which individuals can utilize to generalize sex- role stereotypes to that person. Numerous researchers have demonstrated the existence of sex-role stereotypes (Broverman, Vogel, Broverman, Clark- son, and Rosenkrantz, 1972; Fernberger, 1948; Komarovsky, 1950; Lunneborg, 1970; McKee and Sherriffs, 1959; Reece, 1964; Rosenkrantz, Voegl, Bee, Broverman, and Broverman, 1968; Sappenfield, Kaplan, and Balogh, 1966; Sherriffs, and McKee, 1957). These findings viewed within the context of Miller and Steinberg's conceptual framework indicate that when individuals receive information about the sex of an individual, predictions about the subsequent behaviors of that individual will be based on global sexual stereotypes. While peOple may possess global stereotypes about sex roles, whether accurate or inaccurate, research on the truth validity of stereotypes (Cauthen, Robinson, and Krauss, 1971; Cline and Richards, 1960; Wegner and Vallacher, 1977) 15 suggest that individuals will utilize stereotypes differently depending upon whether he/she will be making attributions or predictions about a cross-sex or same-sex person. Cauthen £3 31. (1971), for example, point out that the traits con- tained in sex-role stereotypes may be true or false but their truth or falsity has little to do with their inclusion in a stereotype. Campbell (1967) suggests that stereotypes initially originate from an observer's experience with various groups of people. A person, for example, may observe the individual member of a group and consistently perceive similar behavior across individuals. Consequently, the person may infer that certain attributes are applicable to all members of that group. Prom Campbell's viewpoint, stereotypes contain a "grain of truth." Research by Wegner, Benel, and Riley (1976) indicate that people do develop certain expectancies about groups and make inferences that correspond with their experience. Chapman and Chapman (1969), however, have pointed out that individuals sometimes view things as related when they are not. They referred to this perceptual tendency as illusory correlation. A study by Hamilton and Gifford (1976) suggests that stereotyping may be partially attri- butable to illusory correlation. Cline and Richards (1960), in a study of the generality of ability to accurately judge others, found that while this global ability can be meaning- fully measured, the accuracy of a judge is dependent upon 16 whether he/she has an accurate stereotype or whether he/she is able to predict specific differences between individuals. These findings suggest an amendment to Miller and Steinberg's conceptual framework. When members of the opposite sex are engaged in a communicative exchange, we would expect sexual stereotyping of a global nature to be used as one means of inferring attitudes and beliefs about the other. Moreover, these stereotypes may be used to make predictions about other types of behaviors as well. How— ever, the research discussed indicates that when same-sex persons interact, the individuals involved will be able to discriminate between the truth and falsity of the stereo- typical generalizations. Predictions based upon inferred attitudes and beliefs about a person of the same sex would be more accurate than predictions made by an individual of the opposite sex. For example, while both males and females may possess global stereotypes about females, males would be incapable of distinguishing aspects of the global stereo- type which are true. Females, however would be capable of differentiating which attributions in the global stereotype are true and which are false. A second possibility is that people may be able to read subtle cues of the same sex better, thus permitting a more rapid acquisition of infor- mation which facilitates predictive accuracy. Thus, in this example, females would be more accurate predictors than males. This additional conceptualization indicates that a person's accuracy of predicting an individual's 17 subsequent behavior is not only contingent upon the type of information solicited but also on the similarity of sexes between the interactants. Therefore, the following hypothe- sis was posited: H4: Individuals exposed to information about a person of the same sex w111 pred1ct that person's behaV1or more accurately than individuals exposed to infor- mation about a person of the opposite sex. In considering the joint effects of type of infor- mation and the sex of dyad, it is anticipated that a person of the same sex exposed to psychological information will make more accurate predictions than a person of the opposite sex exposed to psychological information. Since it is posited that type of information will make a more signifi- cant contribution to the prediction of predictive accuracy than the sex of dyad, a person of the opposite sex exposed to psychological information will make more accurate pre- dictions than a person of the same sex exposed to sociologi- cal information. Finally, persons of the opposite sex ex- posed to sociological information will not only invoke global sex-role stereotypes to predict a person's behavior but will also be restricted to infer future behavior from social information about that person. Consequently, they should make the least accurate predictions. Based upon hypotheses one and four, and the preceding reasoning, the following hypothesis reflecting an ordinal interaction between type of information and the sex of dyad was posited: H5: 18 Individuals exposed to psychological information about a person Of the same sex will make more accurate predictions than individuals exposed to psychological information about a person of the Opposite sex, who will make more accurate pre- dictions than individuals exposed to sociological information about a person Of the same sex. Indi- viduals exposed to sociological information about a person Of the Opposite sex will make the least accurate predictions. The predicted pattern of means hypothesized to emerge is: where E xl XI XI Xl > X2 > mean Of individuals information about a mean of individuals information about a mean Of individuals information about a mean of individuals information about a x3>x4 exposed to psychological person of the same sex. exposed to psychological person of the Opposite sex. exposed to sociological person of the same sex. exposed to sociological person Of the Opposite sex. CHAPTER II PROCEDURES Definitions In this section conceptual and Operational definitions for the following variables will be presented: (1) type of information; (2) accuracy of prediction; (3) dogmatism; and (4) sex Of dyad. (1) Type Of information was conceptually defined as sociological or psychological information exchanged by inter- actants during communication transactions which provides a basis for predicting behavior. Sociological information was defined as data about the characteristics Of an individual which identifies his/ her membership in social groups. Some examples of sociologi- cal information would include his/her sex, political affili- ation, religious affiliation, socio-economic status, and occupation. Psychological information was conceptually defined as data about an individual concerning his/her unique learning experiences. This type of information allows inter- actants to identify characteristics which are unique to a person. "My father always respected my decisions and 19 20 instilled in me a great deal of confidence," is illustrative of psychological information. Type of information was Operationalized in the following manner. One female and one male, 25 years and 31 years Of age respectively, were interviewed and videotaped. These stimulus persons were videotaped sitting behind a table responding to questions develOped by Fontes (1978) designed to solicit sociological and psychological informa- tion. The questions from interview protocol one designed to elicit sociological information were: 1. How Old are you? 2. What do you perceive your ethnic affiliation to be? 3. How much education have you had? 4. Approximately how much money did you spend last year? 5. DO you have any children? 6. What is your marital status? 7. What do you do for a living? The questions from interview protocol two designed to elicit psychological information were: 1. How do you think love and sex differ? 2. If you or your sex partner became unintentionally pregnant, and the child poses an economic hardship, what would you do? 3. If you were pregnant or your sex partner were pregnant and you wanted the child, but medical tests administered during the initial stages Of the 5. 6. 21 pregnancy indicated that the fetus was severely malformed, what would you do? What kinds of interpersonal experiences do you find extremely depressing? What do you fear the most? What kinds of interpersonal experiences do you find extremely rewarding? If you discover that someone you have an inter- personal relationship with has intentionally deceived you, what would you do? (2) Accuracy of prediction was conceptually defined as the congruency between a person's predicted behavior and his/her actual behavior. The higher the congruency between predicted behavior and actual behavior, the higher the accuracy of prediction. This conceptualization was Operationalized using the following procedure. The stimulus persons were interviewed using the following set of questions from interview protocol three developed by Fontes (1978). 1. Which if any charitable organizations do you belong to? At your place Of employment, if you were given the Opportunity to decide whether or not to join a union, what would your decision be? Who would you like to see elected President in the next election? ' 22 4. Would you agree to terminate life support systems for a parent? 5. You and your fiance have agreed to a monogomous relationship. You become aware that your fiance has gone to bed with another person. What would you do? 6. You discover that your spouse has had an extra- marital affair and when you confront him/her about it, he/she lies and denies it. How would you react? Note that the first three questions solicited sociological infOrmation while the latter three questions solicited psy- chological information. This information was Obtained to serve as baseline data for the stimulus person's actual behavior in order to develop questions to measure predictive accuracy. Each question in the predictive accuracy scale contained three foils and the correct response. After exposure to sociologi- cal Or psychological information Obtained from interview protocol one or two, the predictive accuracy scale was administered to subjects who predicted how the stimulus per- son reSponded. Each item was then assigned a value Of zero if the subject chose an incorrect response or a value of one if he/she chose the response the stimulus person gave. A total accuracy score for each subject could then be computed by summing across the prediction items. The scores could range from 0 indicating that the subject could not correctly predict any of the stimulus person's reSponses to 6 23 indicating that the subject could correctly predict all of the stimulus person's responses. (3) Dogmatism was conceptually defined as "a rela- tively closed cognitive organization of beliefs and dis- beliefs about reality, organized around a central set of beliefs about absolute authority which, in turn, provide a framework of intolerance and qualified tolerance toward others" (Rokeach, 1956, p. 3). y This conceptualization was Operationalized using Troldahl and Powell's (1965) Short-Form Dogmatism Scale. A dogmatism score was created by summing across the scale items. The scores could range from 20 indicating that a subject is low dogmatic to 120 indicating that a subject is high dogmatic. (4) Sex Of dyad was conceptualized in accordance with the sex Of the interactants. Specifically, a cross-sex dyad consisted Of male subjects making predictions about the responses of a female stimulus person or Of a female subject making predictions about the responses of a male stimulus person. A same-sex dyad consisted of male or female subjects making predictions about the responses of a male or female stimulus person respectively. This concept was Operationalized by matching the sex of the stimulus person (male, female) with the sex of the subject (male, female) to Obtain a same-sex dyad (melee male, female-female) or a cross-sex dyad (male-female, female-male). 24 Design A multiple regression design was used in this experi- ment in which type of information (sociological, psychologi- cal) and sex of dyad (cross-sex, same-sex) were manipulated categorical variables and dogmatism was treated as an independent continuous variable. Subjects were randomly assigned to one Of four experimental treatment groups defined by type of information and sex Of stimulus person. Equal numbers Of male and female subjects were randomly assigned to treatment groups to facilitate manipulation of sex Of dyad. Random assignment was executed until a sample size of forty subjects per treatment group was attained. TO assess participant's dogmatism, Troldahl and Powell's (1965) Short- Form Dogmatism Scale was administered in all four conditions. Preparing the Stimuli A male and female were solicited through friends Of the researcher to serve as stimulus persons. Both stimulus persons were interviewed and reSponded to interview proto- cols one and two discussed earlier. These interviews were videotaped in color using a frontal camera shot which recorded the seated stimulus person from the waist up. Four different stimulus videotapes were developed from these interviews. One stimulus tape contained the male's response to sociological type questions (see Appendix C). The second stimulus tape contained the male's response to psychological questions (see Appendix D). The 25 third and fourth stimulus tapes contained the female's response to the sociological and psychological questions respectively (see Appendix A and B). Instructions to the Subjects Participants were informed that Dr. Norman E. Fontes and his research team in the Department Of Communication, Michigan State University, were conducting interpersonal communication research. They were told they would view a videotape interview Of a person providing some information about him/herself. After viewing a videotape subjects were additionally instructed that they would complete a question- naire containing questions concerning the person who appeared on the tape and some questions that solicited information about themselves. After viewing the videotape the parti- cipants completed the questionnaire and were told that they would receive a debriefing letter from their instructors at a later date. Description Of Subjects Subjects for this study were recruited from under- graduate communication courses at Michigan State University. One hundred sixty students participated of which 80 were male and 80 were female. Their average age was 19.14 years. None Of the subjects knew the stimulus person who appeared in the videotaped interview. CHAPTER III RESULTS The following predictive equation was generated to test the hypotheses posited in Chapter One. A Y = Blel + Byzxz + BY3x3 + BY4x4 + Bysxs where: g = an individual's predictive accuracy X1 = type Of information X2 = an individual's level of dogmatism X3 = sex of dyad X4 = type Of information X dogmatism interaction X5 = type of information X sex of dyad interaction An assumption of the E-test in the multiple regres- sion model is that the dependent variable is measured at the interval level (McNemar, 1969, pp. 430-431; Namboodiri, Carter,~and Blalock, 1975, p. 87). An arc-sine transforma- tion was executed on total predictive accuracy prior to data analysis (Cohen and Cohen, 1975, pp. 256-257; Neter and Wasserman, 1974, pp. 507-508; Winer, 1962, pp. 397-400) since this measure was nominally scaled, i.e., either the 26 27 subject predicted a stimulus persons response correctly or incorrectly for each item.1 Prior to testing the hypotheses, potential problems concerning multicollinearity were assessed. Multicolline- arity is defined as the condition where some or all Of the independent variables are highly intercorrelated. When explanatory variables are highly correlated, estimates of the unique contribution Of independent variables in a regression analysis may be misleading (Cohen and Cohen, 1975, p. 116). Three multicollinearity assessments were utilized in this study. The first assessment consisted of an exami- nation Of the zero-order correlations among the independent variables. Substantial correlations among the independent variables would indicate multicollinearity problems (Hanushek and Jackson, 1977, p. 90). An examination of Table 1 reveals that the correlations among independent variables were not significant suggesting that no problems with multicolline- arity existed. A second check on multicollinearity is the examina- tion Of the significance of the multiple R2 and the regres- sion coefficients when the dependent variable is regressed on the predictor variables. A significant multiple R2 and nonsignificant regression coefficients would be symptomatic Of multicollinearity (Hanushek and Jackson, 1977, p. 91; Kmenta, 1971, p. 390). Table 2 indicates that the multiple R2 for the regression equation and the regression 28 Table 1.--Zero-Order Correlations Among Independent Vari- ables. X1 x2 x3 x1 -—- x2 -.14 --- x3 .00 -.02 --- *p < .05 Table 2.--Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, and Coeffi- cients Of Determination of Predictive Accuracy Regressed on Type of Information, Dogmatism and Sex Of Dyad (N=160). A Y = Blel + BYZX2 + By3X3 Multiple R = .32 d.f. = 3, 156 p < .05 R Square = .10 F = 5.99 Betas Type Of Information 8 = .29 F = 14.13 d.f. = 1, 156 p < .05 Dogmatism B — -.03 F — < l d.f. = l, 156 p > .05 Sex Of Dyad B = -.13 F = 3.04 d.f. = l, 156 p > .05 29 coefficient for type Of information is significant. This suggests the absence Of problematic multicollinearity. A final check for the threat Of multicollinearity is an examination Of the R2 increase when correlated vari- ables are entered into a regression equation. An appreciable increase in the R2 (when independent variables are corre- lated) would suggest that the correlated independent vari- ables may lay claim to largely the same portion of the systematic variance in the dependent measure (Cohen and Cohen, 1975, p. 116). If there is no increase in the R2 with the addition Of an independent variable which is corre- lated to a minimal degree with other independent variable(s) in the equation, multicollinearity would not pose a problem. 2 The R for predictive accuracy regressed against type Of information is .09. The addition Of dogmatism and sex Of dyad to the regression equation increases the R2 to .10. If the dogmatism or sex Of dyad variable posed threats to multicollinearity, the R2 increase would be substantially higher.2 Based upon these criteria, it was concluded that multicollinearity did not pose a threat to the estimation and interpretation Of the regression coefficients. TO assess the significance of the interactions posited in Hypotheses 3 and 5, interaction terms (X4,X5) were created from the product of type of information x dogmatism and type Of information X sex of dyad, and entered into the simple regression equation. However, this inter- action testing procedure has received a considerable amount 30 Of criticism (Althauser, 1971; Southwood, 1978). Althauser (1971) and Southwood (1978) suggest that since main effect variables and interaction terms will be highly correlated in a regression equation, estimates of regression coeffi- cients for independent variables will be unstable. Dummy coding categorical variables (type Of information, sex of dyad) and retaining dogmatism as a continuous variable, zero-order correlations were computed between main effect variables and interaction terms. The results presented in Table 3 indicate high correlations between the main effect variables and the interaction terms and between interaction terms themselves. This suggests that when utilizing dummy coding, estimates Of the proportion of variance of the dependent variable uniquely accounted for by each main effect and interaction effect will be confounded. Table 3.--Zero-Order Correlations of Main Effect Variables and Interaction Terms Using Dummy Coding. x1 x2 x3 x4 x5 x1 --- x2 -.14 --- x3 .00 -.02 --- x4 .97 .01 .01 --- x5 .58 -.06 .58 .57 -_- 31 TO assess the unique contribution Of each main and interaction effect, the correlations among independent vari- ables may be minimized by orthogonally coding variables. When the sum of products of two variables is zero, the respective variables are orthogonal and their correlation is zero (Cohen and Cohen, 1975, pp. 195-197; Kerlinger and Pedhauzer, 1973, pp. 131-132). Consequently, the categori- cal variables were orthogonally coded and a linear transfor- mation was executed on the dogmatism scores by subtracting the mean Of dogmatism from each score. This transformation was utilized to attain a symmetry of dogmatism scores around a mean of zero. As a result of this procedure, the sum of products of each interaction term with type Of information is zero. Additionally, the sum of products of the type of information X sex of dyad interaction with sex Of dyad is zero. Thus, X1 is orthogonal to X4 and X5, and X3 is ortho- gonal to x5. The results of orthogonally coding the cate- gorical variables and linearly transforming dogmatism scores are reflected in the zero-order correlations Of Table 4. Note that the correlations between main effect variables (X1,X2,X3) did not change while correlations between main and interaction effect variables (X4,X5), and between inter- action effect variables, were substantially reduced. The low correlations of Table 4 indicated that the unique con- tribution of the main effects and the interaction effects could be adequately tested in a simple regression equation 32 Table 4.--Zero-Order Correlations Of Main Effect Variables and Interaction Terms Using Orthogonal Coding. X X X 1 2 3 4 5 x1 -—- X2 -.14 —-_ x3 .00 -.02 --- x4 .00 -.10 .06 --- x5 .00 .06 .oo -.02 --- by orthogonally coding type of information and sex Of dyad and linearly transtrming dogmatism scores. The .05 level Of significance was used to test each of the hypotheses.3 Hypothesis 1 was significant. The results reported in Table 5 reveal that the beta weight for type of information was significantly different from zero indicating it significantly influenced the subjects' accuracy in predicting the stimulus person's behavior. The mean arc- sine transformed predictive accuracy score for subjects exposed to psychological information was 1.48 (s.d. = .46) whereas subjects exposed to sociological information had a mean predictive accuracy score Of 1.20 (s.d. = .49). Translated into percentages, subjects exposed tO psychologi- cal information had a mean predictive accuracy score Of 46 percent, while subjects exposed to sociological informa- tion correctly predicted the stimulus person's responses 33 Table 5.--Prediction of Individual's Total Accuracy of Pre- diction Using Type of Information, Dogmatism, Sex Of Dyad, Type of Information X Dogmatism, and Type Of Information X Sex Of Dyad (N=160). Y = BYIX1 + Bszz + By3x3 + aux4 + BYSXS Multiple R = .32 d.f. = 5, 154 p < .05 R Square = .11 F = 3.64 Betas Type Of Information 8 = .29 F = 14.05 d.f. = l, 154 p < .05 Dogmatism B = -.02 F = < l d.f. = l, 154 p > .05 Sex Of Dyad B = -.13 F = 3.06 d f. = l, 154 p > .05 Type Of Information X B = .02 F = < l Dogmatism Interaction d.f. = l, 154 p > .05 Type of Information X B = -.04 F = < 1 Sex of Dyad Interaction d.f. = l, 154 p > .05 32 percent Of the time on the average. Thus, subjects exposed to psychological information made more accurate predictions than subjects exposed to sociological informa- tion. The relationship posited in Hypothesis 2 between an individual's dogmatism and his predictive accuracy was not supported. The beta weight for subjects dogmatism was not significantly different from zero. A Spearman-Brown split- half reliability computed on the dogmatism measure yielded a coefficient of .65. 34 Hypothesis 3 which predicted an ordinal interaction between type Of information and dogmatism was not supported. The beta weight for the type of information X dOgmatism variable was not significantly different from zero. Hypothesis 4 which posited a relationship between sex of dyad and subject's accuracy of prediction was not significant. While the beta weight for sex Of dyad did approach significance (p = .08), the relationship between sex of dyad and predictive accuracy was not in the direction hypothesized. The mean arc-sine transformed predictive accuracy score for subjects exposed to information about a cross-sex stimulus person was 1.40 (s.d. = .46) while sub- jects exposed to information about a same-sex stimulus person had a mean score of 1.27 (s.d. = .52).4 Hypothesis 5 which predicted an ordinal interaction between type of information and sex Of dyad was not supported either. The beta weight for the type Of information X sex Of dyad interaction was not significantly different from zero. CHAPTER IV DISCUSSION The relationship between type of information and predictive accuracy specified in Hypothesis 1 was signifi- cant. The data from this study indicated that individuals exposed to psychological information about a stimulus person predicted that person's behavior more accurately than indi- viduals exposed to sociological information about the same individual. This result supports Miller and Steinberg's conceptualization that predictions based on psychological data will be more accurate than predictions based on socio- logical data. It should be noted that the total variance accounted for in predictive accuracy by type of information was 9 percent. The prediction that the higher the dogmatism of subjects, the lower their predictive accuracy was not supported. However, given a reliability Of .65 on the dogmatism measure it is difficult to arrive at a definitive conclusion concerning the effects of dogmatism on predictive accuracy. While it is possible that dogmatism has no effect on predictive accuracy, such a conclusion may entail 35 36 committing a Type II error, accepting the null hypothesis when it is false. The probability of committing a Type II error can be minimized by increasing the power Of a statisti- cal test. Power is defined as the probability Of rejecting the null hypothesis when an alternative hypothesis is true, and is determined largely by the effect size (Cohen, 1977; Keppell, 1973).5 The larger the effect size, other factors (significance level, sample size) being equal, the greater the power of the statistical test. Unreliability tends to decrease the effect size and consequently decreases the power Of the test (Cleary, Linn, and Walster, 1970). Given the low reliability Of the dogmatism measure in this study, the power Of the statistical test for Hypothesis II was diminished.6 The unreliability of the measuring instrument may be attributable to the types of items included in the instrument. Troldahl and Powell (1965) assessed the reliability Of their measuring instrument using a sample Of adult subjects from Boston and Lansing. It is quite possible that these subjects were attitudinally and socioeconomically different from the subjects who participated in this study. The items included in the instrument may measure dogmatism in adult samples more effectively than younger student samples. Future research should take this potential problem into account. The pro- blem of the unreliability of the measuring instrument was compounded by the nature of the distribution of dogmatism scores. The mean, mode and median were 60-20, 62.00, and 37 60.30 respectively. Though the distribution was relatively symmetrical, it was leptokurtic and minimally positively skewed. This indicated that the majority of the scores were clustered around a point Of the dogmatism scale sug- gesting that most of the subjects were not very dogmatic. Past researchers utilizing the dogmatism variable have executed quartile splits to facilitate assessing the effects Of high versus low dogmatics on a dependent variable.7 This study, however, utilized all dogmatism scores. Consequently, the truncated distribution of dogmatism scores with very few scores on the high and/or low extremes did not facilitate an adequate test Of the hypothesis. This factor combined with the relatively low reliability Of the dogmatism measure itself significantly reduced the possibility Of detecting a significant relationship between predictive accuracy and level of dogmatism if one exists. Therefore conclusions concerning the relationship between dogmatism and predictive accuracy based upon this study must be approached with caution. Hypothesis 3 positing an ordinal interaction between type of information and dogmatism was not supported. Again, however, the unreliability of the dogmatism instrument and the resulting distribution of scores did not allow for a rigorous test Of the hypothesis. Further research is needed to assess the joint effects Of type of information and dogmatism on predictive accuracy. 38 The relationship between sex of dyad and predictive accuracy specified in Hypothesis 4 was not supported. The beta weight for sex Of dyad did, however, approach signifi- cance although in the direction Opposite that hypothesized. The finding that cross-sex predictions were somewhat more accurate than same-sex predictions merits some considera- tion. This researcher suspects that when individuals receive information about the same-sex stimulus person, predictions are based on inferences about perceived simila- rity Of sex. For example, females predicting other females' behavior may base their predictions upon what they would do given similar circumstances with minimal consideration given to the information provided concerning the individual whose behavior is being predicted. Predictions of this nature would be less accurate because they would be based on the assumption Of a considerable degree of similarity between the individual making predictions and the target of those predictions. Alternatively, when individuals receive infor- mation about persons Of the Opposite sex, they may not assume the existence of a higher degree of similarity and utilize more fully information provided in the development Of predictions about another individual's behavior. If this is indeed the case, cross-sex predictions would be some- what more accurate than same-sex predictions. This speculative explanation is not incompatible with Miller and Steinberg's conceptual position concerning the role of stimulus generalization and discrimination 39 in the prediction process. While the unanticipated relation- ship between sex Of dyad and predictive accuracy did not reach statistical significance, further research is needed to explore the influence Of sex of dyad on predictive accuracy. Finally, the ordinal interaction posited in Hypothe- sis 5 was not supported. The joint effect of type of infor- mation and sex Of dyad simply did not appear to have any systematic effect on predictive accuracy. Four implications from this study should be con- sidered in future research. First, given the scarcity Of studies empirically testing Miller and Steinberg's conceptual approach to interpersonal communication, future research efforts should attempt to replicate the present study and include more reliable measures Of dogmatism. Second, the role Of amount of information in relationship to predictive accuracy is an important area that merits empirical atten- tion. Recall that type of information accounted for 9 per- cent Of the total variance in predictive accuracy. While not a small percentage, by increasing the amount of infor- mation presented to individuals we might expect a corre- sponding increase in variance accounted for. Third, since this study was exploratory, future research should also identify other cognitive, personality, and communication variables which may influence interactants' abilities to make accurate predictions about other individuals' behavior. We are currently in the process of replicating the study 40 presented here and of extending our research to examine the relationship between type Of information, certainty of pre- diction, homOphily, and predictive accuracy. Finally, researchers may wish to consider testing Miller and Stein- berg's conceptual framework in a natural interactive setting. One major limitation Of this study was that information was presented to subjects rather than having subjects solicit information themselves in an interactive setting. Communi- cators in a natural interactive setting may solicit specific information providing them with greater predictive accuracy. APPENDICES APPENDIX A TRANSCRIPT OF INTERVIEW SOLICITING SOCIOLOGICAL INFORMATION FROM A FEMALE STIMULUS PERSON APPENDIX A TRANSCRIPT OF INTERVIEW SOLICITING SOCIOLOGICAL INFORMATION FROM A FEMALE STIMULUS PERSON Interviewer: Interviewee: Interviewer: Interviewee: Interviewer: Interviewee: Interviewer: Interviewee: Interviewer: Interviewee: Interviewer: Interviewee: Interviewer: Interviewee: How Old are you? 25. What do you perceive your ethnic affiliation to be? . White, Caucasian. How much education have you had? I've had um . . . a college education. I've graduated from Michigan State. Approximately how much money did you spend last year? . I haven't got the slightest idea, um . I really don't know. DO you have any children? NO. What is your marital status? Single. What do you do for a living? Um . . . right now I'm employed at a fast food restaurant. 41 APPENDIX B TRANSCRIPT OF INTERVIEW SOLICITING PSYCHOLOGICAL INFORMATION FROM A FEMALE STIMULUS PERSON APPENDIX B TRANSCRIPT OF INTERVIEW SOLICITING PSYCHOLOGICAL INFORMATION FROM A FEMALE STIMULUS PERSON Interviewer: How do you think love and sex differ? Interviewee: . . . Hmmm . . . Okay, love uh . . . can be a relationship between two peOple and I think . . . Um . . . that relationship can come through sex as well . . . um. . . . But then on the other hand, sex can be . . .separated from love when its, its just . . . an act . . . where as love, is uh . . . something that develOps between two peOple . . . from love when its, its just . . . an act . . . where as love, is uh . . . something that develOps between two peOple . . . over a period Of time. Interviewer: If you became unintentionally pregnant and the child poses an economic hardship, what would you do? Interviewee: . . . I would put the child up for adOption, if the, if the, . . . if it was gonna be that hard on my economic situation. Interviewer: If you were pregnant and you wanted the child but medical tests administered during the initial stages Of the pregnancy indicated that the fetus was severely malformed, what would you do? Interviewee: I would probably have the pregnancy terminated. Interviewer: What kinds Of interpersonal experiences do you find extremely depressing? Interviewee: . . . Hmmm . . . I think the most depressing . . . experience would be just any conflict between myself and somebody else . . . that I cared for. 42 Interviewer: Interviewee: Interviewer: Interviewee: Interviewer: Interviewee: 43 What do you fear the most? . . . Um . . . My greatest fear would be . . . um . . . if by some chance I was left without any friends or family, that would be my greatest. What kinds Of interpersonal experiences do you find extremely rewarding? . . . Hmmm . . . just um . . . everyday con- tact with my friends. If you discover that someone you had an inter- personal relationship with has intentionally deceived you, what would you do? . . . I would confront that person and . . . I would hOpe that we would be able to decide whether to terminate our friendship or not. APPENDIX D TRANSCRIPT OF INTERVIEW SOLICITING PSYCHOLOGICAL INFORMATION FROM A MALE STIMULUS PERSON Interviewer: How do you think love and sex differ? Interviewee: When you love somebody . . . you have, I think ‘ you take on responsibilities and Obligations for the person you love. Ah, with sex I don't think that's necessarily true. Sex doesn't have to . . . cause long term responsibilities. . . . I think it should but it doesn't . . . love does. Interviewer: If your sex partner became unintentionally pregnant and the child poses an economic hard- ship, what would you do? Interviewee: Well I guess . . . it would probably depend on whether this, ah, for me it would depend on whether it was in or out of marriage. That would have a bearing on it. I guess one con- sideration would be abortion. Ah . . . one would Obviously be keep the child. We'd have to work it out between the two Of us. . . . Obviously, if she's, she's carrying the child I can't, I couldn't make the decision, say, concerning abortion; I couldn't make that for her. But we could agree on that I think. Interviewer: If your sex partner were pregnant and you wanted the child, but medical tests administerd during the initial stages Of the pregnancy indicated that the fetus was severely malformed, what would you do? Interviewee: Umm . . . that, that would pose the same, same kinds Of problems and raise the same kinds of questions as to the economic hardship. With economic hardship it depends on, on how badly are you going to be totally incapable of 45 Interviewer: Interviewee: Interviewer: Interviewee: Interviewer: Interviewee: Interviewer: 46 supporting the child. With, with the ah, with the physiological problems do they indicate that the child is just going to be a vegetable or something. . . . Here again we'd have the same range Of choices and . . . between the two Of us we would have to consider abortion an adOption, or having and raising the child. What kind Of interpersonal experiences do you find extremely depressing? Between peOple. Between peOple . . . I guess to find out that someone that I had really, really had faith in, that I really trust, hasn't been honest with me. (Lying . . .) or keeping something from me . . . that . . . that I should have known or should have been a part Of. What do you fear the most? Fear the most? . . . That sounds like something out of 1984 where they, they went right to the heart Of what each person feared. Ah . . . I guess losing ah, losing uh my family. If any- thing were to happen, ah, to my wife or kids; as opposed to something happening personally to me that, would be worse than, to me, than something physically happening to me. What kinds of interpersonal experiences do you find extremely rewarding? . . . Well there are times when you're . . . you're working with people when everything seems to click. Ah, you're, you're all think- ing along the same track and you're just . . . you're meshing perfectly. And this can happen. And maybe this is, this is, ah, why I married Pat because it happens so frequently, but it also happens with other people, ah, that I work with. . . . It just . . . you're all on the same frequency and you can, you can, can sense what each other needs. And ah . . . you just kind of get a unity of thought and action. That's . . . both in ah . . . in friendships, and in working situations I think that's very rewarding. O.K. If you discover that someone you have an interpersonal relationship with has intention- ally deceived you, what would you do? Interviewee: 47 GO to 'em and find out why. (Direct confron- tation?) Right. And it, it wouldn't be a, an angry kind Of thing. Just simply be trying to find out whats ah; I know such and such happened and you told me it didn't, or something like that. Why? I don't understand. . . . It just, it would . . . those kinds of things would, would hurt me more than they would raise anger. APPENDIX E PREDICTIVE ACCURACY SCALE FOR FEMALE STIMULUS PERSON APPENDIX E PREDICTIVE ACCURACY SCALE FOR FEMALE STIMULUS PERSON Dear Participant: Dr. Norman E. Fontes and his research team are conduct- ing a series of studies in Interpersonal Communication. Today you are participating in a study that is part Of this research. You have just received information in the form Of an interview from a person whom has told you something about himself/herself. At this time we would like to ask you to complete the attached questionnaire giving careful consider- ation to each question. Complete instructions for each section are included directly before each of the sections. If you have any questions concerning the instructions or any part of the questionnaire, please raise your hand and the person administering the questionnaire will be happy to answer them. The purpose of this study will be explained to you after everyone participating has completed their question- naire. Your assistance is extremely important to us and your cooperation is greatly appreciated. 48 49 INSTRUCTIONS: The following are questions which were asked Of the person who was just interviewed. These questions were not included in the preceding interview. Following each question are four multiple-choice answers, one Of which was actually given by the person interviewed. Please choose the answer which you think the interviewee gave. Please choose only one answer for each question. Following each set of responses are questions concerning how certain you are Of your answer, and how truthful you feel the interviewee was in responding. Please do not skip any questions, and place a check mark (/) by the answer you have chosen. Please read your choices carefully because the response choices for some of the questions are ordered differently. 1. Which, if any, charitable organizations do you belong to? I don't belong to any I'm a volunteer at a hospital I work with the United Way Fund I contribute to the Sierra Club How confident are you Of your answer? not very confident at all not too confident confident very confident The person interviewed provided a truthful response to the question. strongly agree agree disagree strongly disagree 50 Would you agree to terminate life support systems for a parent? Absolutely not. If they were ever in a situation where they would be under a life support system, there is always a chance for recovery since they're still living. I couldn't make the decision on my own. If they were ever under a life support system, a decision like that would have to be made by the entire family. Yes, if beforehand my parents had indicated they would like that action to be taken--if they were ever in a situation where they would be under a life support system. I don't know. If they were ever in a situation where they would be under a life support system-- I couldn't say-~never having been in that situation. How confident are you Of your answer? very confident confident not too confident not very confident at all The person interviewed provided a truthful reSponse to the question. strongly disagree disagree agree strongly agree 51 You and your fiance have agreed to a monogamous relation- ship. You become aware that your fiance has gone to bed with another person after you agreed to the monogamous relationship. What would you do? I would confront him with the situation and accord- ing to his response, I would try and work out some- thing that would be agreeable to both Of us whether we break up or stay together. Chances are we probably wouldn't stay together. I would confront him with the situation and listen to what he had to say but it probably wouldn't make any difference. I would ask him to leave. I could never forgive him for doing something like that to me. I would confront him with the situation and listen to his reasons for why he broke our agreement. We could probably reach an understanding about his affair. Chances are I would probably forgive him and forget it. I would confront him with the situation and listen to his reasons. We would have to try to reach a new agreement which would be fair to the both Of us--where neither of us would be confined to a monogamous relationship. That would probably work out. How confident are you Of your answer? very confident confident not tOO confident not very confident at all The person interviewed provided a truthful response to the question. strongly agree agree disagree strongly disagree 52 At your place of employment if you were given the Oppor- tunity to decide whether or not to join a union what would your decision be? At my present place Of employment, I probably would join a union because I've worked there for a while now and the conditions aren't that good. I'll be at that particular place for awhile so I think it would be good to join a union if it were available» Definitely not at my present place of employment. As long as I've been employed there the conditions have been pretty good. The management has been fair and I've never had any problems. Joining a union there could create problems. I don't feel the need for joining a union there. At my present place Of employment I probably would not join a union, but I would study all the facts and see what advantage a union would be in my parti- cular place because I feel a union is good in some instances and in some instances it isn't. At my present place of employment I've never really considered joining a union because I only work part- time and part-time workers don't have the Opportunity --aren't eligible to join a union. I plan to con- tinue on a part-time basis so the thought Of joining a union hasn't really crossed my mind. How confident are you Of your answer? not very confident at all not too confident confident very confident The person interviewed provided a truthful response to the question. strongly disagree disagree agree strongly agree 53 You discover that your spouse has had an extramarital affair and when you confront him about it, he lies and denies it. How would you react? I'd probably be extremely depressed at first but maybe it would be an indication that something must be really wrong with our relationship and that we should sit down and talk to each other. I would definitely feel that if he did something like that there must be very little left in our relationship, eSpecially if he lies to me. I would probably consider seeking a divorce. I'd probably be extremely depressed but I would con- tinue to try and talk to him and reason with him until it would be impossible to go any further. I would continue to encourage him to speak up or tell me what the problem is in our relationship that would make him lie and that would make him go out- side Of the marriage for an extramarital affair. I'd probably be extremely depressed but I'd probably confront him about the affair and try to find out why he was lying to me and if he continued tO deny it, then I think that our relationship had somehow changed although I think I could still love him. I don't think I could trust him as I did. His affair and his lying to me would always bother me and I'm afraid that would probably show. I'd probably be extremely depressed but I definitely would still love him. We'd have to discuss it and he'd have to consider my feelings too. I would try to explain to him that I did know about his affair and try to get him to admit it. After that I would hOpe he would agree to give up the affair if I for- gave him and forget it ever happened. How confident are you Of your answer? very confident confident not too confident not very confident at all The person interviewed provided a truthful response to the question. strongly disagree disagree agree strongly agree 54 Who would you like to see elected President in the next election? I think if John Conally runs I would vote for him because he comes from Texas which is becoming an important energy producing state and I'm pretty concerned about the energy problem. On that I have no Opinion. I'm not at all familiar with the potential candidates for the next election. I really can't say. I'd probably vote for Carter because he seems to be more concerned with the working middle-class than other potential candidates. It seems to me that inflation is an important issue. It affects everybody. I'd vote for someone who showed a lot of concern for inflation. How confident are you Of your answer? not very confident at all not too confident confident very confident The person interviewed provided a truthful response tO the question. strongly agree agree _ disagree strongly disagree APPENDIX F PREDICTIVE ACCURACY SCALE FOR MALE STIMULUS PERSON APPENDIX F PREDICTIVE ACCURACY SCALE FOR MALE STIMULUS PERSON Dear Participant: Dr. Norman E. Fontes and his research team are conduct- ing a series Of studies in Interpersonal Communication. To- day you are participating in a study that is part Of this research. You have just received information in the form of an interview from a person whom has told you something about himself/herself. At this time we would like to ask you to complete the attached questionnaire giving careful consider- ation to each question. Complete instructions for each section are included directly before each of the sections. If you have any questions concerning the instructions or any part of the questionnaire, please raise your hand and the person administering the questionnaire will be happy to answer them. The purpose Of this study will be explained tO you after everyone participating has completed their question- naire. Your assistance is extremely important to us and your cooperation is greatly appreciated. 55 56 INSTRUCTIONS: The following are questions which were asked of the person who was just interviewed. These questions were not 1ncluded in the preceding interview. Following each quest1on are . four multiple-choice answers, one of which was actually g1ven by the person interviewed. Please choose the answer wh1ch you think the interviewee gave. Please choose only one answer for each question. Following each set of responses are questions concerning how certain you are of your answer, and how truthful you feel the interviewee was in responding. Please do not skip any questions, and place a check mark (/) by the answer you have chosen. Please read your ch01ces carefully because the response choices for some Of the questions are ordered differently. 1. Which, if any, charitable organizations do you belong to? Charitable organizations? Well, you mean that I really participate in? NO, I guess that I don't belong to any. Charitable organizations? Well, I give and work for the Muscular DystrOphy Association every Labor Day weekend. I guess that would be the only one. Charitable organizations? Well, the Red Cross I guess although that's not really a membership. It's a participation or activity--I guess that would be the only one. I Charitable organizations? Well, I support a child through the Christian Children's Fund although that's not really a membership. I guess that would be the only one. I How confident are you Of your answer? very confident confident not too confident not very confident at all The person interviewed provided a truthful response to the question. strongly agree agree disagree Ill strongly disagree 57 Would you agree to terminate life support parent? For one of my parents? That would, depend upon the circumstances--what were in, what the prognosis was for recovery. Under some circumstances For one Of my parents? Absolutely, stances were such that there was no recovery. decide to end their lives. I guess systems for a I suppose, condition they any kind of I might. if the circum- prognosis for For one Of my parents? NO, I wouldn't. I couldn't I could justify my decision but my emotions wouldn't let me. For one Of my parents? I can't really answer that question, I've never been in those circumstances. How confident are you Of your answer? very confident confident not too confident not very confident at all The person interviewed provided a truthful response to the question. strongly disagree disagree agree strongly agree 58 You and your fiancee have agreed to a monogamous rela- tionship. You become aware that your fiancee has gone to bed with another person after you agreed tO the monogamous relationship. What would you do? My first reaction would probably be some mixture Of being hurt and angry. I would probably ask her what was going on but I would love her so much I wouldn't break it Off. My first reaction would probably be some mixture Of being hurt and angry. I would probably ask her what was going on--find out what had happened, and see what there was about the other guy that had broken up the agreement we'd made. My first reaction would probably be some mixture Of being hurt and angry. I would probably ask her what was going on and probably break it Off. My first reaction would probably be some mixture of being hurt and angry. I would ask her what was going on and would probably gO out and see other peOple without her approval. How confident are you Of your answer? not very confident at all not too confident confident very confident The person interviewed provided a truthful response to the question. strongly agree agree disagree strongly disagree 59 At your place of employment if you were given the Oppor- tunity to decide whether or not to join a union what would your decision be? Of course I can either join or not join and keep the same job? Hate to qualify all the time but it would depend on the working conditions. By and large I guess I would have a tendency not to, but if the conditions were pretty rank and it looked like it was advantageous, I might. Of course I can either join or not join and keep the same job? Yes, by and large it is always advantageous tO join a union. Unions have a tendency tO guarantee better working conditions and benefits. Of course I can either join or not join and keep the same job? Well, I don't know--given the nature Of my profession it would be difficult for me to answer that question. Of course I can either join or not join and keep the same job? NO, by and large, unions, while solving other problems, have a tendency to create new ones. I don't think they're a good thing. How confident are you Of your answer? very confident confident not tOO confident not very confident at all The person interviewed provided a truthful response to the question. strongly disagree disagree agree strongly agree 60 You discover that your spouse has had an extramarital affair, and when you confront her about it, she lies and denies it. HOw would you react? I know for a fact and she says no? I suppose I could continue our relationship, but I wouldn't feel I could trust her any longer which probably wouldn't be right because marriage is based on trust. I know for a fact and she says no? I suppose I wouldn't pursue the issue anymore, but I would probably go out on her although I don't know if I'd feel right about that. I know for a fact and she says no? I suppose I would probably be partly responsible for her affair. It takes two people to make a relationship and I must have failed her somehow. I know for a fact and she says no? I suppose I could make it as plain as I could that I did in fact know what was going on despite what she said. If she continued to deny it, I don't know in a situation like that--I don't know. How confident are you Of your answer? very confident confident not too confident not very confident at all The person interviewed provided a truthful response to the question. strongly disagree disagree agree strongly agree 61 Who would you like to see elected President in the next election? DO I have a choice Of candidates here or can I pick anybody? Well, taxes seem pretty important to everybody and I think it would depend on the candi- dates' stand on the tax reduction issue. DO I have a choice Of candidates here or can I pick anybody? Well, it all depends upon who gets the Democratic nomination. If it was Carter, I'd vote Republican. If it was Kennedy, I might vote for him. DO I have a choice Of candidates here or can I pick anybody? Well, if Carter runs again, I don't know who's liable to Oppose him. I voted for Carter last time and if Ronald Reagan ran again I'd vote for Carter again--probably Carter. DO I have a choice Of candidates here or can I pick anybody? Well, I don't know who's liable to run, but I'd want tO vote for a Democrat--maybe Kennedy or Brown. How confident are you Of your answer? very confident confident not too confident not very confident at all The person interviewed provided a truthful response to the question. strongly disagree disagree agree strongly agree APPENDIX G TROLDAHL AND POWELL SHORT-FORM DOGMATISM SCALE APPENDIX G TROLDAHL AND POWELL SHORT-FORM DOGMATISM SCALE Appearing below are some statements about which people have different Opinions or beliefs. Please indicate the extent Of your agreement or disagreement with each statement by circling the apprOpriate choice number for each statement. Please read your choices carefully because the response choices for some Of the questions are ordered differently. 1. In this complicated world Of ours the only way we can know what's going on is to rely on leaders or experts who can be trusted. (l) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. My blood boils whenever a person stubbornly refuses tO admit he's wrong. (1) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. There are two kinds of people in this world: those who are for the truth and those who are against the truth. (1) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) Mildly Agree; (5) Agree; (6) Strongly Agree. Most peOple just don't know what's good for them. (1) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. Of all the different philOSOphies which exist in this world there is probably only one which is correct. (1) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) Mildly Agree; (5) Agree; (6) Strongly Agree. 62 10. 11. 12. 13. 14. 63 The highest form of government is a democracy and the highest form Of democracy is a government run by those who are most intelligent. (l) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) Mildly Agree; (5) Agree; (6) Strongly Agree. The main thing in life is for a person to want to do something important. (1) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) Mildly Agree; (5) Agree; (6) Strongly Agree. I'd like it if I could find someone who would tell me how tO solve my personal problems. (1) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. Most of the ideas which get printed nowadays aren't worth the paper they are printed on. (1) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. Man on his own is a helpless and miserable creature. (l) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. It is only when a person devotes himself to an ideal or cause that life becomes meaningful. (1) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) Mildly Agree; (5) Agree; (6) Strongly Agree. Most peOple just don't give a "damn" for others. (1) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. TO compromise with our political Opponents is dangerous because it usually leads tO the betrayal of our side. (1) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) Mildly Agree; (5) Agree; (6) Strongly Agree. It is Often desirable to reserve judgment about what's going on until one has had a chance to hear the Opinions of those who one respects. (1) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. 15. 16. 17. 18. 19. 20. 64 The present is all too Often full of unhappiness. It is only the future that counts. (1) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) Mildly Agree; (5) Agree; (6) Strongly Agree. The United States and Russia have just about nothing in common. (1) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. In a discussion I Often find it necessary to repeat myself several times to make sure that I am being under- stood. (l) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) Mildly Agree: (5) Agree; (6) Strongly Agree. While I don't like to admit this even to myself, my secret ambition is to become a great man, like Einstein, or Beethoven, or Shakespeare. (1) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) Mildly Agree; (5) Agree; (6) Strongly Agree. Even though freedom Of speech for all groups is a worth- while goal, it is unfortunately necessary to restrict the freedom Of certain political groups. (1) Strongly Disagree; (2) Disagree; (3) Mildly Disagree; (4) M11dly Agree; (5) Agree; (6) Strongly Agree. It is better to be a dead hero than a live coward. (l) Strongly Agree, (2) Agree, (3) Mildly Agree, (4) Mildly Disagree, (5) Disagree, (6) Strongly Disagree. FOOTNOTE S FOOTNOTES 1Two separate data analyses were conducted, one utilizing untransformed predictive accuracy scores and the second using arc-sine transformed predictive accuracy scores. The analyses yielded almost identical results. The results reported in this study are based on transformed scores. Results from the analysis on untransformed predictive accuracy scores are available upon request. 2Note however, that a large R2 increase with the addition Of a variable does not indicate multicollinearity in the absence Of high correlations between independent variables. 3Hypotheses concerning the effects Of subjects dogmatism X sex of dyad interaction and type Of information X subjects dogmatism X sex Of dyad interaction were also tested and found to be non-significant. Results from this analysis will be made available upon request. 4This indicated that cross-sex predictions had a mean accuracy score of 42 percent whereas same-sex predictions had a mean predictive score Of 35 percent correct. 5Other factors affecting the power of a test are the alpha level and sample size. This discussion focuses on the effect size since the unreliability Of a measure directly affects the magnitude Of effect. 6The power Of the statistical test for the beta weight of dogmatism was less than .10. Computations are available upon request. 7For a review of past research exploring the effects Of dogmatism see Ehrlich and Lee (1969) and Vacchianco, Strauss, and Hochman (1969). 65 BIBLIOGRAPHY BIBLIOGRAPHY Althauser, R. P. Multicollinearity and non-additive regres- sion models. In H. M. Blalock, Jr. (Ed.), Causal Models in the Social Sciences. Chicago: Aldine, 1971, pp. 453-472. Altman, I., and Taylor, D. A. Social Penetration: The Development of Interpersonal Relationships. New York: Holt, Rinehart and Winston, 1973. Berger, C. R., and Calabrese, R. J. Some explorations in initial interaction and beyond: Toward a develOp- mental theory Of interpersonal communication. Human Communication Research, 1975, 1, 99-112. Berger, C. R., Gardner, R. R., Clatterbuck, G. W., and Schulman, L. S. Perceptions Of information sequencing in relationship development. Human Communication Research, 1976, 1, 29-46. Berger, C. R., and Larimer, M. W. When beauty is only skin deep: The effects Of physical attractiveness, sex and time on initial interaction. Paper presented at the annual convention of the International Com— munication Association, New Orleans, April, 1974. Birdwhistell, R. L. Masculinity and femininity as display. In S. Weitz (ed.), Nonverbal Communication. New York: Oxford University Press, 1974, pp. 144-149. Bochner, A. On taking ourselves seriously: An analysis Of some persistent problems and promising directions in interpersonal research. Human Communication Research, 1978, 1, 179-191. Broverman, I. K., Vogel, S. R., Broverman, D. M., Clarkson, F. E., and Rosenkrantz, P. S. Sex-role stereotypes: A current appraisal. Journal Of Social Issues, 1972, 28, 59-78. 66 67 Buck, R. W., Savin, V. J., Miller, R. E., and Caul, W. F. Communication of affect through facial expressions in humans. In S. Weitz (ed.), Nonverbal Communication. New York: Oxford University Press, 1974, pp. 51-64. Campbell, D. T. Stereotypes and the perception of group differences. American Psychologist, 1967, 21, 817-829. Cauthen, N. R., Robinson, I. R., and Krauss, H. H. Stereo- types: A review Of the literature 1926-1968. The Journal of Social Psychology, 1971, 81, 103-125. Chapman, L. J., and Chapman, J. P. Illusory correlations as an obstacle to the use Of valid psychodiagnostic signs. Journal Of Abnormal Psychology, 1969, 14, 271-280. Cleary, T. A., Linn, R. L., and Walster, G. W. Effect Of reliability and validity on power Of statistical tests. In E. F. Borgatta and G. W. Bohrnstedt (eds.), Sociological Methodology 1970. San Francisco: Jossey-Bass, 1970. Cline, V. B., and Richards, J. M. Accuracy of interpersonal perception--A general trait? Journal Of Abnormal and Social ngchology, 1960, 69, 1-7. Cohen, J. Statistical Power Analysis for the Behavioral Sciences. New York: Academic Press, 1977. Cohen, J., and Cohen, P. App1ied Multiple Regression/ Correlation Ana1ysis for the Behavioral Sciences. New York: John Wiley & Sons, 1975. Dornbusch, S. M., Hastorf, A. H., Richardson, S. A., Muzzy, R. E., and Vreeland, R. S. The perceiver and the perceived: Their relative influence on the categories Of interpersonal perception. Journal of Personality and Social Psychology, 1965, 1, 434-440. Duck, S. W. Personal Relationships and Personal Constructs. London: John Wiley and Sons, 1973. Ehrlich, H. J. The Social Psychology Of Prejudice. New York: John Wiley and Sons, 1973. Ehrlich, H. J., and Lee, D. Dogmatism, learning, and resistance to change: A review and a new paradigm. Psychological Bulletin, 1969, 21, 249-260. 68 Fernberger, S. W. Persistence of stereotypes concerning sex differences. Journal Of Abnormal and Social Psy- chology, 1948, 41, 97-101. Fontes, N. E. Interpersonal Research Protocol. Unpublished research, Michigan State University, 1978. Hamilton, D. L., and Gifford, R. K. Illusory correlation in interpersonal perception: A cognitive basis of stereotypic judgments. Journal of Experimental Social Psychology, 1976, 13, 392-407. Hanushek, E. A., and Jackson, J. E. Statistical Methodsifor Social Scientists. New York: Academic Press, 1977. Hjelle, L. A. Accuracy Of personality and social judgments as functions Of familiarity. Psychological Reports, 1968, 11, 311-319. Keppel, G. Design and Analysis: A Researcher's Handbook. Englewood Cliffs, N.J.: Prentice-Hall, 1973. Kerlinger, F. N., and Pedhauzer, E. J. Multiple Regression in Behaviorg1 Research. New York: Holt, Rinehart, & Winston, 1973. Kirtley, D., and Harkless, R. Some personality and atti- tudinal correlates Of dogmatism. Psychological Kmenta, J. Elements Of Econometrics. New York: Macmillan, 1971. Komarovsky, M. Functional analysis of sex roles. American Sociological Review, 1950, 15, 508-516. Larimer, M. W., and Berger, C. R. Interpersonal compatability and the development Of interpersonal relationships. Paper presented at the annual convention Of the Speech Communication Association, Chicago, December, 1974. Lunneborg, P. W. Stereotypic aspects in masculinity- femininity measurement. Journal Of Consulting and Clinical Psychology, 1970, 11, 113-118. McKee, J. P., and Sherriffs, A. C. Men's and women's beliefs, 1deals, and self-concept. American Journal of Sociology, 1959, pg, 356-363. McNemar, Q. Psychological Statistics (4th ed.). New York: John Wiley & Sons, 1969. 69 Miller, G. R., and Steinberg, M. Between People: A New Analysis Of Interpersonal Communication. PalO Alto: Science Research Associates, 1975. Namboodiri, N. K., Carter, L. F., and Blalock, H. M., Jr. Applied Multivariate Ana1ysis and Experimental Desigp . New York: McGraw-Hill, 1975. Neter, J., and Wasserman, W. Applied Linear Statistical Models. Homewood, I11.: Richard D. Irwin, 1974. Plant, W. T., Telford, C. W., and Thomas, J. A. Some per- sonality differences between dogmatic and nondogma- tic groups. The Journal Of Social Psychology, 1965, 67, 67-75. Reece, M. M. Masculinity and femininity: A factor analytic study. Psychological Reports, 1964, 11, 123-139. Rokeach, M. Political and religious dogmatism: An alterna- tive to the authoritarian personality. Psychologi- cal Monographs, 1956, Vol. 70, NO. 18. Rokeach, M. The Open and Closed Mind. New York: Basic Books, 1960. Rosenkrantz, P. S., VOgel, S. R., Bee, H., Broverman, I. K., and Broverman, D. M. Sex-role stereotypes and self- concepts in college students. Journal of Consulting and Clinical Psychology, 1968, 11, 287-295. Sappenfield, B. R., Kaplan, B. B., and Balogh, B. Perceptual correlates Of stereotypical masculinity-femininity. Journal of Personality and Social Psyghology, 1966, 1, 585-590. Sherriffs, A. C., and McKee, J. P. Qualitative aspects Of beliefs about men and women. Journal Of Personality, 1957, 11, 451-464. Southwood, K. E. Substantive theory and statistical inter- action: Five models. American Journal Of Sociology, 1978, 11, 1154-1203. Taft, R. Accuracy Of emphatic judgments of acquaintances and strangers. Journal of Personality and Social Psy- chology, 1966, 1, 600-604. Taylor, D. A., and Altman, I. Intimacy-scaled stimuli for use in studies Of interpersonal relations. Psycho- logical Reports, 1966, 11, 729-730. 70 Troldahl, V., and Powell, F. A short-form dogmatism scale for use in field studies. Social Forces, 1965, 11, 211-215. Vacchiano, R. B., Strauss, P. S., and Hochman, L. The Open and closed mind: A review of dogmatism. Psychologi- cal Bulletin, 1969, 11, 261-273. Wegner, D. M., Benel, D. C., and Riley, E. N. Changes in perceived intertrait correlations as a function Of experience with persons. Paper presented at the Southwestern Psychological Association, Albuquerque, April, 1976. Wegner, D. M., and Vallacher, R. R. Implicit Ppychology: An Introduction to Social Cognition. New York: Oxford University Press, 1977. White, B. J., Alter, R. D., and Rardin, M. Authoritarianism, dOgmatism, and usuage Of conceptual categories. Journal Of Personality and Social Psychology, 1965, 1, 293-295. Winer, B. J. Statistical Principles in Experimental Design. New York: McGraw-Hill, 1962. Wright, P. H. Toward a theory of friendship based on a conception Of self. Human Communication Research, 1978, 1, 196-207. TATE UN HICHIGRN 5 IV. LIBRARIES lllllllllllllflll (ll Ill)!I!“IIIIIIMHIWIVIHWNll 31 2 1 0 7371 90 930