TELEVISION AND CHILDREN’S IMAGES OF OCCUPATIONAL ROLES A Dissertation for Hm Degree of DI... D. MICHIGAN STATE UNIVERSITY Gary Robert HeaId I976 LL; LIBRARY Michigan State - a; University " r“. _ This is to certify that the thesis entitled Television and Children's Images of Occupational Roles presented by Gary Robert Heald has been accepted towards fulfillment of the requirements for Ph. D. Communication degree in Major rofessor DMe August 20, 1976 0-7639 m M. W s ABSTRACT TELEVISION AND CHILDREN'S IMAGES OF OCCUPATIONAL ROLES BY Gary Robert Heald There is much still unknown about the processes of socialization in this culture. The nature of learning about occupational roles, in particular, is a relatively uncharted area due to the complexity of the societal work roles, as well as the multiplicity of information sources now available. This study examines the occupational socialization phenomenon by focusing on organizing principles that lead to consistency in the ways that individuals perceive work roles, and by ex- ploring some of the effects of learning about occupational roles from "primary" as opposed to "secondary" information sources. Primary information sources are those sources that are shared by large societal aggregates (e.g., mass media), with the primary sources providing largely undifferentiated messages. In contrast, secondary sources (e.g. family, friends) are more idiosyncratic and individualized thus allow— ing diverse, specialized information to be learned. The essay begins with a discussion of "incidental learn- ing" as it relates to children and adolescents. Learning in this paradigm centers on the acquisition of images, not iso- lated facts, about occupational roles. Special attention is given the part that the self—concept plays in image perception. Gary Robert Heald Ten hypotheses are drawn out of this as to specific personal, relational and material characteristics that individuals use to distinguish occupational roles. The theoretic section then addresses the consequences of learning about work roles from a shared, primary source such as television. Hypotheses are offered as to the effects of this primary information source on cultural homogenization, occupational stereotyping, status conferral and individual perceptions of work role dis- tributions. The research hypotheses were tested through a survey of 210 fourth, sixth and eighth grade students. Using a combina- tion of paired-comparison and unidimensional measures, fifteen occupational roles were studied. Seven occupations were chosen due to their emphasis in the television medium; the remaining eight, while not emphasized in TV programming, were chosen owing to their use in previous research. Multivariate analyses revealed that eight of the ten hy- pothesized attributes were as predicted in contributing to occupational images held by children. Counter to expectations, examination of the occupational images held by male versus female, and lower as opposed to upper socioeconomic children revealed no consistent patterns. Also contrary to hypotheses are the results indicating that learning about occupational roles from primary as compared to secondary information sources does not lead to greater cultural homogenization. The hypothesis of a stereotyping effect received mixed, Gary Robert Heald inconclusive support. There is evidence, however, that per- sons receiving a proportionally larger part of their informa- tion from primary sources tend to have more stereotypic views of occupational roles. The hypothesis of a status conferral effect traceable to receiver dependence on primary information sources was not sustained. Predictions that primary sources can influence perception of occupation role distributions were confirmed. Conditional analyses further demonstrated partial support for predictions that this effect is greatest where secondary information about work roles is relatively absent. The relationships found between exposure to primary ver- sus secondary information sources and subsequent perceptions of occupational roles are intriguing. Investigators of social role imagery, however, would do well to seek additional ante- cedent and intervening variables as the amount of variance now accounted for in the dependent variables is still quite small. Equally important, considerable attention should be given to methodological issues surrounding the multidimension- al scaling of cultural perceptions, and especially the problem Of ascertaining the amount of variance surrounding points in conceptual spaces. TELEVISION AND CHILDREN'S IMAGES OF OCCUPATIONAL ROLES BY Gary Robert Heald A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1976 This one is for Mary ii ACKNOWLEDGMENTS Numerous individuals aided me in my doctoral program. Joseph Woelfel chaired my Guidance Committee. He showed in- comparable patience while providing intellectual encourage- ment when needed most. I am equally indebted to Donald Cushman for his unfaltering direction during my course studies as well as throughout the writing of this disserta- tion. His insightful questioning and genuine friendship greatly enriched these experiences. Special gratitude is also extended to Eugene Jacobson for his meticulous scrutiny of this dissertation effort. Charles Atkin, a final member of my Guidance Committee, generously granted me the oppor- tunity to pursue this study as a portion of a funded research project. This dissertation was partially supported by Grant No. 90-C-63S from the Office of Child Development; Department of Health, Education and Welfare. Richard Jones, Eugene Golanda and Richard Johnson, Principals in the Grand Ledge School District, were truly helpful and cooperative during the data collection at their schools. Mark Miller, Jack Wakshlag and Mary Heald assisted in the administration of the survey questionnaires. Beyond this, the critical abilities of Mark Miller and Jack Wakshlag, along with those of Robert McPhee and Mike Cody, focused my iii efforts and enlarged my understanding in this communication study. To these scholars I express my deep appreciation. And then there is Mary, my wife. Mary asked that I not extol her love and devotion--her wish is granted. I do, however, want to thank Mary for her willingness to exper- ience financial hardship, endure loneliness and forego un- told pleasures, especially during these trying dissertation times. iv TABLE OF CONTENTS Chapter Page I INTRODUCTION . . . . . . . . 1 Problem Area . . . 1 Primary Versus Secondary Information Sources . . . . . . . 2 II THEORETIC PARADIGM . . . . . . 5 Television: A Primary Information Souce For Children . . . . . 5 Image Perception . . . . . . 8 Self— -Perception . . . . 10 Personal Characteristics of Self- Perception . . . . l3 Relational Characteristics of Self- Perception . . . . . . 16 Material Characteristics of Self- Perception . . . . . . . 20 Primary Sources and Cultural Homogenization . . . . . . 24 Primary Sources and Stereotypes . . 27 Primary Sources and Status Conferral . 31 Primary Sources and Perceived Role Distributions . . . . . . 34 Summary . . . . . . . . 36 III METHODOLOGY . . . . . . . . 41 Survey Questionnaire . . . . . 41 Sample . . . . . . . . 45 Survey Procedures . . . . . 47 Statistical Analyses . . . . . 49 IV RESULTS . . . . . . . . . 57 Summary . . . . . . . . 99 Chapter Page V DISCUSSION . . . . . . . . . 102 Perceived Occupational Attributes . . 102 Cultural Homogenization . . . . 103 Stereotypes . . . . . . . 105 Status Conferral . . . . . . 107 Perceived Role Distributions . . . 107 Conclusions and Suggestions . . . 108 APPENDICES . . . . . . . . . . . 112 A Survey Questions . . . . . . . 112 B Permission Form Example . . . . 152 C Statement of School Policy Compliance . . 153 BIBLIOGRAPHY . . . . . . . . . . 154 vi Table LIST OF TABLES Standardized Regression Weights for Six Orthogonal Dimensions Used to Predict Mean Occupation Ratings on Personal, Relational and Material Attributes for Total N . . Standardized Regression Weights for Six Orthogonal Dimensions Used to Predict Mean Occupation Ratings on Unidimensional Attributes Among Male vs. Female Subjects Standardized Regression Weights for Six Orthogonal Dimensions Used to Predict Mean Occupation Ratings on Unidimensional Attributes Among Lower vs. Upper SES Subjects . . . . . . . . . Differences Between Occupational Images Held by Male and Female Subjects in Single Source vs. Mixed Sources Conditions . . . . Zero-Order Correlations of the First Four Corresponding Orthogonal Dimensions of the Occupational Spaces Held by Male and Female Subjects in Single Source vs. Mixed Sources Conditions. . . . . . . . . Differences Between Occupational Images Held by Lower and Upper SES Subjects in Single Source vs. Mixed Sources Conditions . . Zero-Order Correlations of the First Five Corresponding Orthogonal Dimensions of the Occupational Spaces Held by Lower and Upper SES Subjects in Single Source vs. Mixed Sources Conditions . . . . . . Vii Page 59 65 67 70 73 75 78 Table Page 8. Average Variances for the Paired-Comparison Measures of Seven Television Emphasized Occupations Among Subjects with Single vs. Mixed Sources of Information . . . . . 8O 9. F-tests Comparing Variances Around the Mean Attribute Ratings of Police Officers by Subjects with Single vs. Mixed Sources of Information . . . . . . . . . 81 10. F-tests Comparing Variances Around the Mean Attribute Ratings of Doctors by Subjects with Single vs. Mixed Sources of Information . 82 ll. F-tests Comparing Variances Around the Mean Attribute Ratings of Nurses by Subjects with Single vs. Mixed Sources of Information . . 83 12. F-tests Comparing Variances Around the Mean Attribute Ratings of Private Detectives by Subjects with Single vs. Mixed Sources of Information . . . . . . . . . 84 13. F-tests Comparing Variances Around the Mean Attribute Ratings of Paramedics by Subjects with Single vs. Mixed Sources of Information . 85 14. F-tests Comparing Variances Around the Mean Attribute Ratings of Truck Drivers by Subjects with Single vs. Mixed Sources of Information . . . . . . . . . 86 15. F-tests Comparing Variances Around the Mean Attribute Ratings of Lawyers by Subjects with Single vs. Mixed Sources of Information . . 87 16. Zero-order Correlations Between Frequency of Viewing Television Programs Emphasizing Particular Occupational Roles and the Perceived Importance of the Occupations . . 91 17. Conditional Zero-Order Correlations Between Frequency of Viewing Television Programs Emphasizing Particular Occupational Roles and the Perceived Importance of the Occupa- tion by Subjects in Single vs. Mixed Sources Conditions . . . . . . . . . 92 18. Zero-order Correlations Between Frequency of Viewing Television Programs Showing Repeat- edly Emphasized Occupational Roles and the Perceived Number of Persons Having the Occupations . . . . . . . . . 95 viii Table 19. 20. 21. Conditional Zero-order Correlations Between Frequency of Viewing Television Programs Showing Repeatedly Emphasized Occupational Roles and the Perceived Number of Persons Having the Occupations by Subjects in Single vs. Mixed Sources Conditions . . . . Zero-order Correlations Between Frequency of Viewing Television Programs Showing Females in Non-Standard Occupational Roles and the Perceived Number of Females Having the Occupations . . . . . . . Conditional Zero-order Correlations Between Frequency of Viewing Television Programs Showing Females in Non-Standard Occupational Roles and the Perceived Number of Females Having the Occupations by Subjects in Single vs. Mixed Sources Conditions . . . . ix Page 96 97 98 Figure 1. LIST OF FIGURES Graphical Representation of Three Attributes on Dimension One and Dimension Two of the Conceptual Space for the Total Sample . . Graphical Representation of Six Attributes on Dimension One and Dimension Three of the Conceptual Space for the Total Sample . . Graphical Representation of One Attribute on Dimension Two and Dimension Three of the Conceptual Space for the Total Sample . . Graphical Representation of the Mean Differences Between Male and Female Occupa- tional Perceptions in the Single Source Condition . . . . . . . . . Graphical Representation of the Mean Differences Between Male and Female Occupa- tional Perceptions in the Mixed Sources Condition . . . . . . . . . Graphical Representation of the Mean Differences Between Lower and Upper SES Subjects‘ Occupational Perceptions in the Single Source Condition . . . . . Graphical Representation of the Mean Differences Between Lower and Upper SES Subjects' Occupational Perceptions in the Mixed Sources Condition . . . . . Page 62 63 64 71 72 76 77 CHAPTER I INTRODUCTION Problem Area Learning about diverse occupational roles represents an important aspect of children's socialization. Theories of child development buttressed by empirical research argue the significance of work role information for children be— ginning around the third and fourth grades (Van Hoose and Leonard, 1966; Smith, 1968; Thompson, 1969). Children's occupational socialization, however, is not as straight- forward as might be expected. The complexity and sheer num- ber of different occupations are part of the problem; there are in excess of 20,000 separate occupational roles (Hollander and Parker, 1969). Moreover, unionization and child labor laws limit children's learning by barring them from work settings (Ausubel, 1954; Borow, 1973). Compound- ing this are contemporary living patterns that separate home and work environments. Children do not fare much better in learning about dif- ferent occupations from authoritative sources at school. Goodson (1968) reports that occupational information in school libraries is often dated and narrowly focused on a few jobs (e.g., "policeman," "post office worker,‘ "secret l 2 serviceman"). Amos and Grambs (1968) join Goodson in con— cluding that a large portion of the occupational material available in schools will not, or cannot, be read by the children that need the information most. Educational cur— riculums, in addition, infrequently contain classes design— ed to inform children about occupational roles; less than half of the U.S. high schools provide vocational guidance of any form (Lathrop, 1974). Finally, surprisingly limited communication about oc- cupational roles occurs between children and their parents. In numerous studies adolescents report talking to each other about occupations and careers far more often than they con- fer with older, more knowledgeable sources (Woelfel, 1972; Woelfel, 1976). DeFleur (1964) plus Hollander and Parker (1969) have suggested that much of children's occupational learning occurs through mediated communication. In particular, DeFleur (p. 57) has proposed that occupational socialization "takes place largely through accidental or haphazard ex— posure to a variety of learning sources. Among these, the mass media appear to play a major role." The following essay weighs the impact of learning about occupatiOnal roles from these common, shared information sources. Primary Versus Secondary Information Sources Of concern are the effects of social learning in "pri— mary" as opposed to "secondary communication systems." Following the lead of Woelfel (1976), primary systems will here entail sources of information that are shared by large social aggregates, with the primary sources providing undif— ferentiated messages. In contrast, secondary systems will refer to tailored information sources which allow diverse, specialized learning. For this discussion, information gained from television will serve as an example of learning from a primary source. Learning from varied interpersonal discussions and direct experiences will constitute instances of learning from secondary sources. Secondary sources will be considered mainly for purposes of comparison. Chapter II begins by introducing a paradigm of inci- dental learning. This paradigm centers around children and young adolescents. It is proposed that children's learning from the primary source television is largely incidental. The processes of perception and learning by contiguity are discussed. Learning in this paradigm concerns the acquisi- tion of images, not isolated facts, about occupational roles. Attention then focuses on the occupational attributes that conceivably are central to incidental learning. In this section the importance of the individual's self-concept is addressed. Finally, consideration is given the nature of the effects that are linked to learning about occupations from a shared primary information source. Chapter III details an empirical study conducted to test hypotheses that emerge in the theoretic discussions of the 4 first chapter. The survey questionnaire, the sample and administration procedures are described. Ethical issues as- sociated with the survey are addressed. This chapter con- cludes with an overview of the statistical analyses. The results of the empirical study are contained in Chapter IV. Each hypothesis is restated, empirical findings are presented as tests of the predictions. The results are interpreted, with decisions then made as to the tenability of the hypotheses. A final section, Chapter V, summarizes and discusses the empirical results. Conclusions are drawn as to what this investigation reveals about children and the effects of their social learning from primary as opposed to secondary informa- tion sources. Here, also, future research is considered. CHAPTER II THEORETIC PARADIGM Television: A Primary Information Source For Children Three things concerning children and television are well documented by Maccoby (1954); Himmelweit, Oppenheim and Vince (1958); Schramm, Lyle and Parker (1961); Roberts and Schramm (1971); Liebert, Neale and Davidson (1973). First, children watch a lot of TV. Close to one—sixth of their waking hours are spent in this manner. Second, children watch the breadth of programming offered. In addition to child-oriented programs, situational comedies, drama, varie- ty and action-adventure shows are common to their television diets. Third, children watch TV to be entertained. They attend to this medium because they are amused by its content. These three factors spark the observation that children have extensive contact with the adult world as depicted on this primary source. Furthermore, it is the combination of these factors that gives rise to a particular form of learning. Specifically, much of children's learning from TV appears to be "incidental" in nature. It is relatively uncommon for children to seek information in television; most learning from TV takes place when the child "goes to television for entertainment and stores up certain items of information 5 6 without seeking them" (Schramm, et 31., 1961:75). Krugman and Hartley (1970) describe this as "passive" learning, as it is "typically effortless . . . and characterized by an absence of resistance" to the material presented. Other re— searchers have posited incidental learning to involve recall of material nonessential and often irrelevant to the plot of the program (cf., Hale, Miller and Stevenson, 1968; Collins, 1970). Thus there are behavioral and content qualities of the phenomenon collectively indicating that incidental learn- ing entails the unintentional learning of peripheral informa- tion in a media presentation. Empirical studies have isolated conditions where inci- dental learning is most likely to occur. Schramm,et 31., (1961) report that television absorbs the attention of young children and it is this capacity to capture attention, com- bined with the presentation of novel materials, that results in incidental learning. As the initially novel tOpic becomes "old stuff," however, children cease to pay close attention (Brodbeck, 1955). Maccoby (1963) similarly suggests that children's learning from television portrayals will be great- est in areas where interpersonal information is lacking and in areas where there are few "real-life experiences." Learn- ing from this common, shared information source is thus thought to be most pronounced where other sources are limited. A second group of researchers have concluded that there is a curvilinear relationship between age and incidental 7 learning from mass media presentations (cf., Hale, 35 al., 1968; Collins, 1970). Holding exposure constant, incidental learning has been shown to steadily increase for children from age eight through age twelve. Among children in the seventh and eighth grades there is a noticeable decline, how- ever. This nonmonotonic relationship has been attributed to increases in older children's learning abilities, with early adolescents additionally beginning to focus attention on the central elements of media programs. These two sets of studies shed light on the domain of incidental learning, but they do not reveal the logic of the phenomenon. For example, little attention is afforded the fact that implicit in the proposition of incidental learning is the premise that learning can occur in the absence of l . . . . This merits cons1derat1on. "detectable reinforcement." Debates abound between factions as to the necessary con- ditions and processes of learning. In one particular camp are the theories by Mowrer (1960) and Sheffield (1961) which argue for learning by contiguity. The central position is that repetitive, contiguous presentations of stimuli result in cognitive associations. Mowrer's paradigm is especially relevant due to his notion that "an image is a conditioned sensation . . ." (p. 171). An individual responds to a new stimulus as it "acquires an image" of another stimulus with which the individual is familiar. Consistent with this are recent essays by Roberts (1971), Schramm (1973) and Danes (1975) 8 which posit that mass communications linking environmental objects principally affect audience "beliefs," "pictures," "images" of reality. But what is the process leading to image perception? Image Perception Several scholars have hypothesized that the act of per- ception is essentially a matter of differentiation in which an individual notes similarities (or dissimilarities) in objects with regard to underlying attributes (cf. Bruner, 1964; Woelfel, 1974). The process is in large part the same whether the objects of attention are concrete (e.g., people) or abstract (e.g., social roles). Ultimately, it is the aggregate of the discerned similarities between environmental objects along the chosen attributes that constitutes an over- all perception of the objects--an image of reality. An example will clarify this. Occupational roles can be differentiated along various attributes. The overall simi- larity between the occupations along the attributes yields a picture, an image of the work world. Consider the following matrix of hypothetical data.2 Self Determi- Formal Work Income Prestige nation Training Conditions Fireman 3 4 3 3 4 Mail Carrier 4 3 4 4 3 Physician 1 ' l 1 1.5 2 Professor 2 2 2 1.5 1 9 This matrix represents a rank ordering of four occupations along the five different attributes. The rank-orderings dem- onstrate that physicians and university professors are appre- ciably more alike than are medical doctors and mail carriers. Clearly, the similarity of occupations can be evaluated along a single attribute. But looking across both a range of occupations and attributes, the intricacy of the work world emerges. There is a wide variety of attributes that can be used to discriminate a single object or class of objects. The previous studies reveal that the attributes noted are at times as much a function of the perceiver as the perceived (Yarrow and Campbell, 1963; Dornbusch, Hastorf, Richardson, Muzzy and Vreeland, 1965). There furthermore appears to be consistency in the attributes a perceiver uses to differen— tiate objects (Hastorf, Richardson and Dornbusch, 1958). In light of this and owing to the focus of this essay, it be- comes crucial to address three questions. (1) What is the organizing principle that leads to consistency in the attri- butes an individual uses to differentiate occupational roles? (2) What are the common attributes that different persons use to distinguish occupational roles? (3) What attributes are conceivably stable across groups? The answer to the first question is alluded to in Schilder's (1942) treatise on perception and thought. Schil- der contends that consciousness of external objects is 10 "necessarily connected to self—observation." In the case of children perceiving occupational roles, the reasons why this is true are compelling. Children are paramountly concerned with themselves. Possibly the clearest evidence of this appears in their lan- guage behavior--much of young children's speech is "egocen- tric" (Piaget, 1955; Vygotsky, 1962). Beginning around age eight egocentric speech declines. The decline, however, does not mark an end to children's interests in themselves. As Vygotsky describes it, this is where "individualized" activity begins; with the development of vocabulary, chil- dren's abilities to differentiate themselves increase, plus their thoughts become qualitatively more precise. This notion of individualization has considerable overlap with social-psychological theories of self-concept and self-per— ception. Self-Perception It is difficult to pinpoint a conceptual definition of "self—concept" and a best discussion of its development. For some theorists, the self is a structure, for others it is a process (Woelfel, 1967). One school of thought offers that the self is stable, a second contends that it is situational and still a third posits that it is both (Kuhn, 1964; Gordon, 1968; Mahoney, 1973). One of the few things for which there appears to be near unanimity is the belief that the self evolves from the individual's information about himself. 11 This information, moreover, is thought to relate the individ— ual to objects in his environemnt. Self-conception, then, is a special case of perception. Woelfel (1967:43—44) details the "conditions of self-knowl- edge" as follows: The process of knowing who one is is a pro- cess of definition . . . [and] the process of definition . . . consists in differentia- tion and association of the individual with other objects. An individual identifies him- self in terms of his conception of his relation- ship to the objects of his experience. The self-conception, then, is the sum of the indi- vidual's conceptions of his relationships to objects. Likely most important here is the belief that self-knowledge develops in a societal context. Individuals from a single society not only encounter similar objects to use in self- definition, other societal members assist the individuals in self—object definitions through communication (Cooley, 1902; Mead, 1934). Some uniformity in the process is, therefore, to be expected. In the case of maturing children one of the classes of objects that inevitably must be dealt with involves social roles. Of this category, occupational roles are especially notable. Empirical research has demonstrated that children three to six years of age already know that there are dif- ferent occupational roles and realize that both men and women fill them (Beuf, 1974). There is additionally, reason to believe that even quite young children begin to relate themselves to occupational roles (Moore, 1969). 12 How children see themselves in relation to different occupational roles varies. Where there is some conceivable uniformity, however, is in the attributes they use to dif— ferentiate the roles and themselves. There are three reasons for this expectation. First, there are attributes common to both self—concepts and occupational roles that are formed by societal norms and mores. To cite an example, there are socially defined aspects of the individual's sex that corre- spond with the "sex appropriateness" of different occupa- tions. Numerous theorists have, secondly, noted the impor- tance of the self-concept in vocational interests and choices (cf., Bordin, 1943; Super, 1953; Gonyea, 1961; Holland, 1963). Finally, empirical research into the principal categories of self-conception has yielded several "dimensions" that parallel attributes resulting from a distinct body of research on occupational role perceptions. A central issue, then concerns the specific aspects of the self-concept that become salient for children in their perceptions of occupational roles.3 A word of forewarning is called for here. The research on the self-concept that is considered in the succeeding pages is context-free. Mahoney (1973), among others, has shown that self-reported aspects of the self-concept vary considerably from condition to condition. As will be noted, several attributes thought to constitute self-identification, while not highly salient in general, do conceivably become more important in the context of l3 discerning occupational roles and the self. The following discussion will demonstrate the parallels between the attributes emerging from empirical research on self—conception and occupational role perceptions. This portion of the essay will be divided into three parts. Sep- arate sections will be devoted to personal, relational and material attributes that designate points of intersection be- tween the two bodies of literature. As these parts deal with points of intersection, there is limited direct research confirming the propositions to be offered. Pertinent evi- dence is drawn from diverse literatures to back the tenabil- ity of the arguments. Personal Characteristics of Self-Perception One of the first traits learned which remains inextric- ably tied to self-concept is the individual's "sex.” Children as young as seventeen months--and certainly by the age of three years—~know their sex (Rabban, 1950; Money and Ehrhardt, 1973). From these early ages children recognize that they are similar to some people and different from others with respect to this characteristic. Gordon (1968) furthermore found that sex was the third most frequently mentioned char- acteristic a sample of adolescents ascribed to themselves. Mahoney (1973) has indicated that sex is one of the attri- butes that individuals more often use in repeated self de- scriptions. 14 Olshan (1970) examined the structure of third, sixth and ninth graders' perceptions of ten social roles (e.g., mother, teacher, fireman). Sex emerged as a major dimension that sixth and ninth graders used to differentiate the role holders. Beuf (1974) determined that children are aware of the "sex appropriateness" of occupational roles portrayed on television. Miller and Reeves (1976) likewise concluded that sex was an attribute young viewers used to distinguish television occupational roles with the sex attribute proving to be more salient for males than females. These studies warrant the conclusion that sex is one aspect of the self-concept that children use to distinguish occupational roles. The sex attribute is more important in perceptions by males than females. H Children distinguish occupations on the basis of the sex of the role holders. H Male children distinguish occupations more a on the basis of sex than do female children. In terms of the number of people using the category and the consistency with which it is utilized over time, the individual‘s "sense of competence" is also one of the most pervasive aspects of the self-concept as measured by the "Twenty Statements Test" (TST) (Gordon, 1968; Mahoney, 1973). Competence, here, is a subjective evaluation of the general ability to cope effectively with varying situations. Self- descriptive terms that coincide with this category include "intelligent," "talented," "skillful" (Gordon, 1968:129). 15 Holland (1963) asked 638 high aptitude boys and girls to complete a set of sentences with a single word. The sentences took the form: "Physicists are ." and rotated the occupations engineer, physicist, teacher, account- ant, business executive and artist in the stem position. Of the responses obtained, there was a tendency to evaluate oc- cupations in terms of general abilities. In excess of thirty percent of all adjectives describing engineers and physicists, plus more than twenty percent of the evaluations of account— ants and artists, referred to their competence. Eleven per- cent of the adjectives describing business executives and five percent describing teachers dealt with the capabilities of the role holders. Olshan's (1970) study of children yielded similar results. Third, sixth and ninth graders alike differentiated social role holders in terms of per- ceived intellectual abilities. It is therefore offered that competence is a second personal attribute that children use to differentiate occupational role holders. H2 Children distinguish occupations on the basis of the perceived competence (intel- ligence) of the role holders. The "sense of moral worth" also appears as a major as- pect of the self-concept. This theoretic dimension, Gordon (1968:127) contends, parallels "Baldwin's formulation of the ethical socius" and "Mead's idea of self-respect." Sense of moral worth is variously conceptualized as the perception of "moral standing" and the "adherence to a valued code of moral 16 standards . . .“ with the category often represented in the TST by the self—descriptive adjectives "bad," "good," etc. (Gordon, 1968; Mahoney, 1973). Utilizing diverse samples, Gordon (1968) found that as few as twenty-two percent and as many as forty percent of the respondents expressed moral worth in their self—descriptions. Much along the same line, Olshan's (1970) examination of the dimensionality of children's perceptions revealed a "good-bad" attribute. Again, students in all three grades appeared to use this evaluative property in distinguishing the different social roles. A comparable finding is reported in Reeves' (1976) study of children's perceptions of fourteen television characters. A final personal characteristic that appears to be sal- ient for self-conception and occupational role perception is thus found in the discernment of moral worth. H3 Children distinguish occupations on the bas1s of the perce1ved moral worth (good- ness) of the role holders. Relational Characteristics of Self-Perception Gordon's (1968) and Mahoney's (1973) conceptualization of "self-determination" identifies an initial relational characteristic of the self-concept. Self-determination, here, refers to the relative ability of the individual to define goals, to behave and to otherwise freely act. The frequency with which this characteristic is self-reported varies con- siderably; Gordon (1968) indicates that responses fitting 17 this category result from between twenty—three and fifty— five percent of the subjects in different samples. This complex attribute is amplified by Weinstein's (1956) discussion of "authority." Weinstein reports that children are sensitive to the authority of occupations with regard to power over self and over the activities of others, too. Stone and Church (1973) propose that the desire for autonomy is in part learned from parents. This is supported by Weinstein's observations that as children grow older the meaning of authority changes to also imply "personal respon- sibility." He goes on to suggest that for upper "socioeco- nomic status" (SES) children, power over "one's own decision" is more important than being "boss." Weinstein's research shows that the importance of occupational authority in gen- eral is negatively related to social status. DeFleur (1964) similarly demonstrated the importance children assign to "power" in evaluating different occupations. No status dif- ferences were found by DeFleur. This study, however, con- ceptualized power solely in terms of power over other people. An initial relational characteristic common to self- identification and occupational role perceptions is thus found in the construct authority. Power over self and power over others represent constituent elements of this construct. It is tentatively offered that upper SES children value and therefore note autonomy more than lower SES children. 18 H4 Children distinguish occupations on the basis of the perceived authority of the role holders. H4 Upper socioeconomic status children dis- a tinguish occupations more on the basis of the perceived autonomy (who receives orders) of role holders than do lower socioeconomic status children. As mentioned earlier, societal members assist the indi- vidual in deciding who and what type of person he is. Other persons help the individual define relationships between self and physical as well as psychological objects. Both Gordon (1968) and Mahoney (1973) conceptualize a category of "judgments imputed pg others" (i.e., impressions and atti- tudes toward the individual as perceived by the individual). Attributes such as "p0pularity," "respect," and "love" are examples. Research indicates that, overall, this is one of the less consistent aspects of the self-concept. There are rea- sons, nonetheless, for proposing children's use of this at— tribute in role perceptions. First of all, while other people's judgments are important to most everyone at one time curanother,developmental psychologists agree that for chil- dren and adolescents peer evaluations are of great concern (cf., Stone and Church, 1973; Adams, 1973). Yarrow and Campbell's (1963) study reveals the significance that "social" and "interaction" categories play in young peOple's descrip- tions of others. Much along the same line, an evaluative trait of "few friends—many friends" is reported by Olshan 19 (1970). Again, this characteristic emerged as a significant aspect of children's discriminations of roles such as teach- er, fireman, etc. This leads to the proposition that chil- dren distinguish occupational roles with respect to judgments imputed to others. H Children distinguish occupations on the basis of the perceived popularity (degree liked) of the role holders. VEry closely related is a still more general category of the self-concept that is described as "interpersonal style." Gordon (1968:130) defines this as the "individual's typical nnéirlner of acting." Interpersonal style includes tendencies to be introverted as opposed to extroverted, and dispositions to be demanding versus supporting. Creason and Schilson (1970) interviewed 121 sixth-grade Youngsters concerning their vocational preferences. Respon- dents were asked to express their interests and the reasons behind their choices. The children's reasons for aspiring towé‘lrd the diverse occupations were grouped into seventeen categories. The reasons given most were nondescript state- ments (e.g., "I like it."). The second most frequently gj“"€311 answers indicated the children's interest in helping other people. This finding is in line with Rosenberg's ”‘9 5 '7) and Borow's (1973) observations that children show cleaéiir? interest in occupations that "DELE" other people. It i . 8 thus offered that a third relational attribute along Wthh ch' lldren distinguish occupational roles concerns the 20 occupations' apparent contributions to others. 6 Children distinguish occupations on the basis of perceptions of how much the role holders help other people. H ,Material Characteristics of Self-Perception There is consensus that James (1892) helped pioneer the "material resources" contribute to self-identifi- idea that Material referents can be holdings such as cars, cation. c:lothes and money or they can be abstract possessions such £153 "a secure future" (Gordon, 1968). In situations of gen- eazral self—description, references to "material possessions £111d resources" are infrequent, though they do occur. The significance of material referents is much greater j.11 :relation to occupational roles. DeFleur (1964) found that iFCDJ: children the material benefits accruing to occupational Mate- 17<>ZLe holders were second in importance only to power. 3E7j;a11 benefits, nonetheless, do not appear to be uniformly Weinstein (1956) examined the ef- important among children. feCts of the children's SES on perceptions of the importance As status increased the importance 0 f occupational income. These results parallel reports by of income decreased. Centers and Cantril (1946) and Bendix and Lipset (1953) VIkIjL<21n propose that persons lacking material possessions note tf1jL53‘ deficiency and express greater interest in having mate- 111a 1 objects. 21 A parallel set of possessions and.resOurces is found in the environmental settings associated with different occupa- tions. Work conditions constitute some of the more visible role attributes. Referring once again to Weinstein's (1956) survey, there is evidence that children's social status is loositively related to the value assigned to work conditions. '{Factors such as dress, cleanliness of work, control over ‘unorking time" were significantly more important for upper class respondents. The following conclusions are consequently tenable. bdarterial possessions and resources associated with occupa- 1::ixonal roles are salient in children's perceptions. The in— czcornes associated with different occupations constitute an Eitztzribute that is more critical for lower SES children. Higher socioeconomic status children differentiate occupa- tions in terms of work conditions. H Children distinguish occupations on the basis of the perceived income of role holders. H Lower socioeconomic status children distin- a guish occupations more on the basis of per- ceived income of role holders than do upper socioeconomic status children. H Children distinguish occupations on the basis of the perceived work conditions of role holders. H Upper socioeconomic status children dis- a tinguish occupations more on the basis of work conditions of role holders than do lower socioeconomic status children. ) t) l 22 Mahoney (1973) observes that "social status" can appear as an attribute of the individual's self-concept in the form of a categorical designation (e.g., "Middle Class") or by comparative references (e.g., "from a family of lower status than most college students"). Either way, the individual's social status still is a benefit resulting largely from an (occupational role. Children must deal with status both in .its derived and earned forms (Ausubel, 1954). A child's snocial status is initially a derived characteristic stemming frtmihis relationship with his parents. As the child grows colder, the requirement to establish anew one's social status becomes increasingly obvious. Various scholars have examined the issues of occupation- Eil prestige and work role aspirations (cf. Reiss, eg_al., 31961; Haller and Miller, 1963). Investigations by Galler (1951) and Steward (1959) have confirmed the expectation that (Even children are aware that different statuses and impor- ‘tance are ascribed occupations. Simmons (1962) found that children's occupational status rankings corresponded signif- icantly with adult rankings. For fourth-grade boys there ‘was a Spearman rho of +.87 and a rho equal +.54 for girls. There was a rS of +.94 between adult occupational status rankings and the rankings by eighth-grade boys and girls. Gunn (1964) adds support to Simmons' conclusions that chil- dren are sensitive to the status associated with different occupations and rank occupations in a manner very nearly Lu f‘ 23 identical to adults. H9 Children distinguish occupations on the basis of the perceived status (importance) of the role holders. To summarize briefly, the discussion has thus far centered first on television as a primary source which dis- seminates undifferentiated messages to children. Among the ‘undifferentiated messages are portrayals of the adult world, .including occupational roles. The discussion has further- Inore projected a paradigm of incidental learning. Of spe- <:ia1 concern are the images that children obtain about var- :ious vocations. An attempt has been made to identify major iattributes that children use to discriminate occupational :roles. The self-concept has been posited as an organizing Iprinciple that leads to the use of ten attributes in dis- ‘hinguishing work roles. Four of the attributes are var- :iously significant in perceptions by children of different sexes and socioeconomic backgrounds. Accepting tentatively the proposals of what elements constitute the structure of children's occupational percep- tions, a central issue of this essay can be pursued. The theoretic essay has detailed a model depicting how children conceivably gain impressions about occupational roles. At- tributes at the foundation of the occupational images have been proposed; of interest then is how information sources affect children's perceptions pf occupations along these 2E7 tributes. Specifically, in what ways--if any--does the 24 learning about occupational roles from television influence viewer images of the adult work world? Primary Sources and Cultural Homogenization It is important to recall that television is signifi— cant here, as it represents a class of information sources. {Television is an instance of a primary source which provides "standardized" messages and "distributes common information 'to many persons at once . . ." (Woelfel, 1976:69). Laswell (1948) contends that one of the principal func- tzions of communication systems is the "transmission of soc- :ial inheritance." This is a maintenance function designed t1) stablilize societies. Oral communication was initially :sufficient for this task, but this was before societies mush- Inoomed in size and complexity. Consequently, the mass media laave become basic to the transmission of a common social rueritage; some scholars refer to this as the emergence of 'Hnedia societies" (Peterson, Jenson and Rivers, 1965; Schramm, 1973). Referring specifically to the transmission of common occupational information by television, DeFleur and DeFleur (1967) have described this maintenance function as one of "homogenization." Persons that watch the same media program- ming are thought to gain fundamental information. In the case of children viewing various occupations on TV, it is decisively important that they are often denied direct ac- cess to work environments. What most children know about 25 some occupations, and indeed what some children know about most occupations, comes from their television experiences. The result is that television ip principle serves to equal- ize occupational knowledge, at least concerning the limited set of work roles covered in popular programming. The significance of this hypothesized effect is con- siderable. There have been numerous discussions of the dis— ;parity of occupational knowledge within social systems. Appreciable differences have been noted between male and :female children's occupational knowledge, with similar in- eaquities reported between children from advantaged and dis— aadvantaged social backgrounds (cf., Simmons, 1962; Amos and (Brambs, 1968). To the large extent that television is ac- <:essible to persons of varying sex and SES, it would be ex- ]oected to reduce these inequalities, especially among those asubgroup members who receive a proportionally larger part (of their occupational information from TV. Two surveys have sought to substantiate the existence of the hypothesized homogenization effect. DeFleur and DeFleur (1967) evaluated children's knowledge about occupa- tions the researchers classified as "personal contact occu- pations" (e.g., teacher, mailman), "television contact occupations" (e.g., lawyer, reporter), and "general cultural occupations" (e.g., bank president, electrical engineer). The adjectives "personal contact," "television contact," and "general cultural" referred to the expected sources of 26 of information. Children ranging in age from six to thirteen years old were interviewed. The results indicated pg sig— nificant differences between male and female respondents' knowledge about television contact occupations. Boys still knew more than girls about the general cultural occupations, however. Differences were also found in knowledge levels of (distinct social classes. Children at the bottom of the class structure knew less about the three sets of occupations than ciid upper SES children. Kawashima (1971) replicated the DeFleurs' study; this second survey was conducted in Japan. Kawashima reports sig- 11ificant differences between male and female children's knowl- enige about television contact occupations. Differences were aalso found in knowledge levels between upper and lower SES (children. Again, males and upper SES children were shown to lenow more about occupational roles. These studies would appear to cast doubt on the reality 0f the prOposed homogenization effect. Were it not for one factor this clearly would be the case. The studies share a common problem; neither investigation associated actual view- ing behavior with occupational knowledge. DeFleur and DeFleur, as well as Kawashima, reported bivariate analyses showing a positive relationship between occupational knowl- edge and general TV viewing by the total samples. In the analyses comparing males versus females, and upper versus lower classes, actual exposure to television programming 27 featuring the occupational roles was not controlled. The general proposition is that television is a source of societal information and persons watching the same pro- gramming gain much of the same information. This prOposi- tion gains greater plausibility with a slight modification. Persons from distinct subgroups that receive common infor- rnation from television should develop similar impressions as the shared source becomes their primary or single source of information.4 H There will be greater homogeneity in the a occupational images held by male and female children with shared single sources of in- formation than in the occupational images held by male and female children with mixed sources of information. 10 H10 There will be greater homogeneity in the b occupational images held by lower and upper SES children with shared single sources of information than in the occupational images held by lower and upper SES children with mixed sources of information. The immediately preceding discussion taps one of the 'potential aspects of homogenization. But possibly the more theoretically interesting effect does not center around the reduction of social differences, but rather on a more subtle result of learning from a common source. Primapy Sources and Stereotypes In Lippman's (1922) discussion of "public Opinion" he spoke of the "pseudo—environments" that are created by the mass media. Lippman felt that the "real environment [was] 28 altogether too big, too complex and too fleeting for direct acquaintance" and this left man in a susceptible condition (p. 16). To describe the result of this condition Lippman borrowed the phrase stereotyping. As originally used, "a stereotype was the plate made by taking a mold of a printing surface and casting type from it" (Peterson, Jenson and Rivers, 1965:23). Lippman offered that since the public was receiving much of its information about the world from the same media, the public's minds were similarly being molded to see the world in a certain way. This idea can be extended to other forms of mediated information where direct experi- ence is limited. With reference to occupational roles the effect is not only that aggregate perceptions of distinct groups are made similar by media exposure, but also that for group members receiving their information from the same source perceptual differences are minimized altogether. A voluminous literature now exists on the causes, the processes, and the results of stereotyping. Despite this, there is still disagreement as to what is a stereotype in a social—psychological sense (Brigham, 1971). Looking across the stereotyping research, several conceptual patterns be- come evident. From the outset, stereotypes are thought to be "general- izations" about objects falling into particular categories such as races, ethnic groups, occupations (Vinacke, 1949; Brigham, 1971; Dipboye and Anderson, 1961). Some researchers 29 have conceptualized stereotypes as "overgeneralizations" or "incorrect" (cf., Katz and Braley, 1935; Centers, 1951). Investigators have often addressed stereotyping in nominal categories reflecting extremes (cf., Katz and Braley, 1935; Walker, 1958). Finally, conceptualizations indicate that stereotypes require group agreement (LaViolette and Silvert, 1951; Fishman, 1956). These patterns in conceptual definitions of stereotypes deserve some comment. Specifically, the second and third characteristics evoke extremely complex issues. To pose an example, some might find it difficult to know when a general— ization is in fact an overgeneralization. Beyond this, the statement that a generalization is incorrect also requires some standard to guide evaluations. Vinacke (1949) and Fishman (1956) are among those researchers that have ex- pressed concern-for this requirement. Brigham (1971), more- over, has argued the rarity with which objective standards are available. Therefore, the inclusion of invalidity as a criterion in the conceptual definition of steredtypes seems unwarranted. In like manner, there is little good reason to address stereotypes as nominal phenomena. The categorical study of stereotyping has obvious limitations: (1) To have subjects evaluate groups in nominal categories tends to force general- izations (Brigham, 1971). (2) Such generalizations are often forced to appear at attribute extremes. (3) A categorical 30 approach precludes understanding the degree to which objects must be stereotyped as extremes for major consequences to result. Thus for the remainder of this discussion the term ster- eotypes will refer to generalizations for which there is group consensus. Degrees of stereotyping will describe the amount of variation in group consensus. Within this concep— tualization the issue is not whether television programming portrays occupations stereotypically (though there is evi- dence that this is in fact the case). The crucial factor is whether the reception of information from a shared single source leads to greater homogeneity of perceptions along selected attributes (i.e., reduced variation about the aver— 292)- A cultural subgroup (e.g., viewers of programs "X") could conceivably be made more stereotypic in their perceptions by witnessing a common TV model or set of TV models. If this subgroup's primary or single source of occupational informa- tion is a TV program or set of programs, it would be expected that their perceptions about the occupation would be similar. The assumption is that for audiences viewing the same range of models, their perceptions will cluster around the averages of attribute portrayals. An additional feature of this perspective on stereotyping suggests the feasibility of a primary source reducing stereo- types, too. Primary sources should also be able to increase 31 the variance in perceptions. Consider the situation where societal members have stereotypic views of a particular occupation.' The portrayal of that occupation on television in a counter-stereotypic fashion ought to shift the beliefs of viewers. Thus for the culture as a whole, there would be an increase in perceptual variation. Research by Atkin and Miller (1975), plus Miller and Reeves (1976), gives credence to this argument. Hll There will be greater stereotypic views of occupational roles by children with shared single sources of information than by children with mixed sources of infor- mation. Primary Sources and Status Conferral Switching to more individualized effects, a third poten- tial outcome revolves around the result of a primary source gaining social standing itself. Societal views of television are well summarized by Greenberg and Roloff (1974). This primary source has achieved credibility and prestige. These achievements are potentially passed on to the information and topics it portrays. Referring principally to the news media, Lazarsfeld and Merton (1948) argue that the media "confer status" on issues and groups merely by giving them recognition. This contention is bolstered by Hovland's (1954) description of the "halo" effect of the media, plus Klapper's (1960) discussion of the "prestige native to the media" and how persons appearing in the media gain public 32 "stature." Lemert (1966) proposes that Lazarsfeld and Merton's notions of status coincide with the concept of credibility as defined by Berlo, Lemert and Mertz (1966). Empirical evidence is provided supporting the position that media at— tention can enhance credibility. In a later article Lemert (1969) enlarges the status domain to include "prominence." Prominence was Operationalized in part by measuring sub- jects' perceptions of how "important" were persons appear- ing in the media. Again, empirical evidence confirms the hypothesis that media coverage increases audience perceptions of prominence. Returning to occupational roles, Moore (1969) notes tele- vision's glamorization of selected occupations. Programs featuring characters in work roles frequently show the major characters as doing exciting things. In addition, there is little question that media-emphasized occupations (e.g., doctor, lawyer, police officer, paramedic) are shown as helping other people and contributing to society. Such re- peated occupational portrayals are analogous to Lemert's conceptualization of prominence (i.e., "important"). Thus there is reason to propose that media emphasis of selected occupational roles enhances viewer perceptions of role prominence. A single study by Dominick (1974) has pursued the hy- pothesis that television emphasis of occupational roles re- sults in status conferral. Dominick surveyed a group of 33 elementary school children to establish their perceptions of the "prestige" associated with the roles of police officer and private detective. Prestige is not defined; there is brief mention that the respondents ranked ten occupations (only two of which are listed). The results of this survey do not confirm Dominick's argument that "exposure to crime and police shows" should lead to "greater prestige [being] attributed to law enforce- ment occupations" (p. 7). Dominick concluded that part of the results were due to the high prestige all youngsters attributed to police officers and private detectives. The median ranking for police officer was "2.07"; a median rank of "2.60" was given private detectives. The highest pres- tige rank possible was "1.0." Beyond this, Dominick reports that "the relative lack of variation may have reduced the correlation" between viewing behavior and perceived prestige (p. 9). The high prestige rankings reported by Dominick are puzzling. Investigators previously cited have demonstrated the overall accuracy of children's perceptions of occupation— al status. It would be helpful to know what were the other eight occupations. In sum, it is not clear just how adequate is this test of the status conferral prOposition. A clearer test is warranted. Such a test would require (1) a range of occupational roles, (2) measures of children's exposure to specific television programming and (3) a scale establish- ing the perceived importance of each media-emphasized 34 occupation. Recalling that status was one of the attributes hypothesized to contribute to children's occupational images, it is offered that: H12 The greater children's exposure to tele- a vision programming emphasizing an occupa- tional role, the greater the perceived relative importance of that occupation. Among children receiving proportionally b more occupational information from tele- vision, the greater the exposure to tele— vision programming emphasizing an occupa- tional role, the greater the perceived relative importance of that occupation. 12 Primary Sources and Perceived Role Distributions A second aspect of role prominence gains significance in the context of children's images pf the work world ag’a whole. The repeated emphasis of selected occupations by pri— mary sources makes them highly conspicuous. This becomes all the more certain when one considers the diversity and range of work roles as compared to television occupational por- trayals. To quote DeFleur: It is clear that it would be exceedingly dif- ficult for a young viewer to obtain much ac- curate information about the distribution of occupations by watching his television screen. We have noted the concentration of occupations related to the law, the preoccupation with upper status jobs, the infrequent representation of ordinary work roles and the imbalance of the sexes (1964:65) [emphasis added]. Other scholars have reported comparable inaccuracies in television job portrayals (cf., Katzman, 1972; Seggar and Wheeler, 1973; Downing, 1974; Tedesco, 1974; Miller and 35 Reeves, 1976). It is conceivable that the one area where occupational role distributions are most discrepant is in the number of men holding law enforcement, medical, or other pro- fessional positions (Katzman, 1972; Seggar and Wheeler, 1973; Tedesco, 1974). Similarly, women are infrequently shown holding occupations other than standard jobs such as nurse or secretary (Seggar and Wheeler, 1973). This writer performed a content analysis of occupational roles appearing on prime-time television. Analyses covered the programming offered by the three major networks during the Fall 1975 and Winter 1976 seasons. The TV viewing be- haviors of fourth, sixth and eighth graders in the Grand Ledge, Michigan area were also surveyed (Wakshlag and Korzenny, 1976). The prime—time shows viewed most by children biased the distribution of occupations held by men and by women. Specif- ically, more than forty percent of the employed men were shown in the roles of police officer, physician, private detective, fireman/paramedic and lawyer. Police officer was the most frequently portrayed male occupation, this role represents in excess of twenty percent of all occupations held by men. Doctors accounted for seven percent of the male roles, lawyers accounted for two percent. Finally, barely four percent of the employed men were pictured as part of the factory labor force. 36 In excess of fifty percent of the women with identifi- able vocations were depicted as housewives, nurses and teach- ers. Housewives constituted the largest single group, but sixty-nine percent of all females were still shown as working outside of the home. Nursing accounted for sixteen percent of the female occupations. In terms of "nonstandard" female work roles, four percent of the employed females were police officers, two percent were lawyers. One show portrayed a woman as a paramedic. As before, the major consequences of these portrayals stem from the condition where undifferentiated messages con- stitute a substantial portion of a receiver's information about a topic. Hl3 The greater children's exposure to re- a peatedly emphas1zed occupational roles on television, the greater the perceived number of persons thought to have the occupations. The greater children's exposure to oc- b cupational roles held by women on tele- vision, the greater the perceived number of women thought to have the occupations. 13 Among children receiving proportionally more occupational information from tele- vision, the greater the exposure to occu- pational roles held by males and females on television, the greater the perceived number of males and females thought to have the occupations. 13 c Summary This essay has generated a detailed paradigm. The following is argued: Many children are isolated from the 37 work world's realities. The vastness of the occupational arena initially precludes clear understanding of this area by children. Legal and normative sanctions combined with living patterns further restrict many children's opportuni- ties to directly witness different occupations. Addition- ally, parental and educational sources often provide rela- tively limited occupational information. Children and adolescents do have extensive contact with adult occupations as depicted on the primary source tele- vision. Moreover, through this entertainment-motivated con- 'tact, children conceivably gain assorted information. They Enotentially note, for instance, which jobs are held by the Clifferent sexes, which jobs carry authority, which jobs reap nuaterial benefits, etc. Not all children learn about occupa- txional roles in this manner. There are both environmental ade developmental factors that define the limits of incident- a 1 learning . The process of incidental learning is essentially one ofjperception and learning by contiguity. Viewers perceive true occupational roles shown on TV by noting similarities of true roles along underlying attributes. This association of tins occupations to each other, combined with perceptions of tlueir positions along the underlying attributes, yields a PjITture, an image of the work world. The attributes children use to differentiate the various IV OCcupations are numerous. There is good reason to believe, 38 nonetheless, that in part children use salient aspects of their self-concepts to discriminate occupational roles. Three categories of self—identification are of importance here. (1) Personal Characteristics (i.e., sex, sense of competence, sense of moral worth). (2) Relational Characteristics (i.e., self- determination, judgments imputed to others, interpersonal style). (3) Material Characteristics (i.e., material possessions and resources). Within these categories the attributes sex, self-determina- tion and material possessions are differentially important depending on the sex and socioeconomic standing of the per— ceiver. In terms of these underlying attributes, learning about occupational roles from a single shared source influences viewers' images in two ways. First, viewing should equalize occupational knowledge across the sexes and distinct social c1asses--at least concerning media-emphasized occupations. Learning from a common source, secondly, ought to minimize variations in perceptions along the salient attributes. This amounts to stereotyping of media-emphasized occupations. The communication source likewise influences receiver perceptions of the information. Where the shared primary source has gained an air of exclusiveness, topics dealt with Should gain prominence. With respect to television, viewers' 39 perceptions of the prominence of a set of occupations should be enhanced by learning about the roles from television. This should be especially true where information from secon- dary sources is relatively absent. Finally, it is offered that learning about the work world from television results in a biased perceptual image of the distribution of occupations. Content analyses reveal substantial inaccuracies in tele- vision's attempts to "mirror" the work environment. Again this hypothesized effect is contingent on availability of in— formation from other sources. 40 Notes Danes (1975) provides a succinct summary of the argu- ments surrounding theories of reinforced versus con— tiguous, conscious as opposed to unconscious, learning. Woelfel's (1974) theory of occupational choice contains a similar matrix. He demonstrates these principles using a "dummy coding" system. Woelfel's discussion further addresses the utility of multidimensional analyses in occupational research. Three recent articles are relevant to this question. Each offers a set of conceptual categories that attempt to "capture major dimensions" of self-conception. The category system by Stone, Dunphy and Ogilvie (1967) contains ninety-nine concept categories. Gordon (1968) identifies thirty categories; an equal number is found in Mahoney (1973). The Gordon and Mahoney categories are parsimonious and heuristic for the present topic. The term "shared source" will be used to designate the condition where respondents report near exclusive and relatively high exposure to occupations through a pri- mary information source. A "mixed sources" condition will denote contact with primary and secondary informa- tion sources, or solely secondary sources of occupational information. CHAPTER III METHODOLOGY This chapter details the research procedures used to test the hypotheses offered in the previous discussion. A beginning section describes the survey protocol and its de- velopment. Results of a series of pilot studies are re- ported; these studies guided the empirical operationaliza- tions of key variables. The survey sample and administration procedures are discussed next. Ethical considerations assoc- iated with these research procedures are reviewed. A final section is devoted to the discussion of statistical analyses. SurveyiQuestionnaire A survey questionnaire with six distinct parts was em- ployed for this study. The beginning of the questionnaire contained "paired-comparison" items designed to yield a representation of children's images of occupational roles. Fifteen occupational roles were compared. Based on content analyses of prime-time television in the 1975 Fall/Winter season, seven occupations were chosen due to their media em- phasis (i.e., police officer, doctor, nurse, private detective, paramedic, truck driver and lawyer). Eight additional occu- pations appear with varying frequency on television. The 41 42 work roles (i.e., housewife, teacher, banker, mail carrier, secretary, janitor, mechanic and factory worker) are not emphasized in TV programming, however. These eight roles were ultimately chosen on the bases of their use in previous research (cf., DeFleur and DeFleur, 1967; Kawashima, 1971) and to give a sex, a status as well as an occupational activ- ities (situs) balance to the questionnaire. Pilot studies with children ages nine to twelve provided guidelines for developing the paired—comparison items. Two issues were of major concern. Initially, in writing these questions, of importance was the choice of words that children understand as well as the choice of familiar adverb modifiers to express degrees of difference. Of parallel interest was the identification of a continuum along which children could meaningfully order objects in terms of perceived similarity. These preliminary inquiries revealed (1) that children fre- quently use the words "alike" and "different" in combination with modifiers "a little," "very much" and "completely" in describing objects' similarities and (2) that children can consistently discriminate objects along a six—point continuum with a neutral phrase fixed as the seventh point. The paired- comparison response categories and questions used in this study appear in Appendix "A," pages 112-130.1 A second set of unidimensional scales was constructed to identify the underlying structure of children's occupational role perceptions. Ten personal, relational and material 43 attributes defined in the previous chapter were operational- ized using five-point response categories. Each of the oc- cupational roles was evaluated with respect to the attributes. Appendix "A," pages 131—146, contains the items. Five-point scales were used to measure children's fre- quency of interpersonal discussion about the topic occupations. Respondents were asked how often their parents and friends. talked about each of the work roles. A parallel set of ques- tions asked how often the respondents saw people with the different jobs. The interpersonal discussion as well as the direct contact items are indicators of exposure to "secondary" sources of information about the work roles. The specific questions appear in Appendix "A," pages 147—149. A fourth group of questions tapped respondents' percep- tions of the number of people actually holding work roles that are repeatedly emphasized in television programs. Police officers and doctors are the two male-dominated roles most heavily portrayed in current TV programming; nursing con- stitutes the major female work role-—second only to housewife. These questions took the form: "If you saw 25 people, how many would be police officers?" Subjects were allowed to write in answers ranging from 0 to 25. A subset of questions asked for perceptions of the num- ber of females in "non-standard" occupations. These items had a similar format to the previous three questions, though they focused on the work roles of police officer, paramedic and 44 lawyer. Page 150 of Appendix "A" contains these questions. Nielsen ratings for the Lansing region combined with three television-viewing studies conducted in Grand Ledge, Michigan, were used to isolate particular programs for TV exposure measures. Specific shows were chosen for inclusion in this questionnaire if they actively portrayed an occupa- tional role of interest and if the programs were viewed regu- larly by twenty percent or more of the area children. "Kate McShane" was an exception to this latter criterion; this short- run series, nonetheless, was included owing to its presenta- tion of a woman working as a lawyer. The final list of shows included eleven prime-time series and two locally syndicated programs (see Appendix "A," page 151). Children answering the questionnaire indicated approximately how many times each month they watch the individual prime-time shows and how of— ten each week they view the syndicated programs. Demographic information was obtained concerning the child's sex, age, grade in school and parents' occupations (Appendix "A," pages 112, 151). Questions were also asked con- cerning the number of color TV sets as well as the number of bedrooms in each respondent's home, how many automobiles the child's family had and finally whether the child owned a 10- speed bicycle (Appendix "A," page 151). Questions of this nature have been used in previous studies to measure family affluence when surveying young children.2 Before administering the total questionnaire, selected portions were pre—tested with thirty children in a combined 45 fifth- and sixth-grade classroom in Lansing. This pre-test supported the expectation that children would be able to deal with the survey items. With fgw_exceptions, the fifth and sixth graders could read the questions without assistance. Moreover, they answered the paired-comparison, the unidimen- sional attribute and the viewing behavior questions intellig- ibly. Sample To test the research hypotheses the questionnaire was administered to nonprobability, cluster samples of elementary and middle school students in Grand Ledge, Michigan. Grand Ledge is situated six miles west of Lansing. The 1970 Census indicates 6032 people living in this community; furthermore, a detailed demographic breakdown reveals that Grand Ledge in many ways mirrors the structure of a diversified midwestern community. Comparing Grand Ledge with Lansing, the economic and labor structures are very similar; there are appreciable differences in the average household sizes and the racial make- ups, however. The average household in Grand Ledge contains 3.21 persons as compared to 3.67 in Lansing; less than one ‘percent of the Grand Ledge population is a racial minority in contrast to 10.1% in Lansing. Four classes of fourth graders, three classes of sixth and three classes of eighth graders were surveyed. A total of 260 students participated. Of the total, thirty-one question- naires were incomplete--twenty-eight students were absent on one of the two days three students left naires blank. This and 70 eighth-grade 46 that the instrument was administered, substantial portions of their question- left a total of 88 fourth-, 71 sixth- survey protocols in completed form. To equalize grade sample sizes, eighteen of the fourth-grade questionnaires and one of the sixth-grade forms were randomly deleted. The resulting data set contained 210 cases repre- senting seventy respondents in each grade. A breakdown of the sample reveals these characteristics. There is an equal number of males and females; the average age is 145.3 months (12.11 years). Comparing grade by sex distributions the following results: Fourth Sixth Eighth Total grade grade grade Females 37 37 31 105 Males 33 33 39 105 Total 70 70 70 210 It is readily apparent that there are relatively minor dif- ferences between the gender breakdowns of the respondents in each grade level. There are more females in the lower and more males in the upper grades. Age in Months Mean Maximum Minimum Females 143.16 193 114 Males 147.48 188 114 47 Age in Months Mean Maximum Minimum Fourth grade 120.40 134 114 Sixth grade 145.00 158 138 Eighth grade 170.53 193 163 There are similarly small differences in the age distribu- tions for the two sexes and in the three grade levels. The average age of the male respondents is slightly higher, but there are no substantial aberrations to be contended with in the statistical analyses. Survey Procedures The total questionnaire was quite long. Its length, however, was reduced for individual subjects by using a ran- domization technique for the paired-comparison and unidimen- sional attribute items. .Each respondent was asked to make paired-comparison and attribute judgments for the seven media- emphasized roles and a subgroup of the remaining occupations. The media-emphasized occupations were evaluated by 210 respon- dents and the remaining eight work roles were evaluated by 105 respondents. Additionally, due to the length of the instrument the survey was divided into two sessions. Group administration required between 20 and 40 minutes on two consecutive days. The questions were read aloud to fourth graders; sixth- and 48 eighth-grade students worked by themselves. All survey ques- tions were answered independently by the children. In the unusual instance where subjects had difficulties, the prob- lems were handled on an individual basis by a survey adminis- trator or an assistant. Some mention is called for as to the ethical considera- tions associated with this study of human subjects. Three factors are of prime importance here: (1) respondent anonym— ity, (2) respondent consent and freedom from coercion, and (3) respondent debriefings. Concerning the first point, sub- jects were guaranteed that their answers would be treated confidentially. Identifying labels were removed from the questionnaires. All analyses have been conducted and re- ported with grouped data. Permission slips were sent home to the parents of the elementary school students. An example of the permission form can be found in Appendix "B." The Principal of the mid- dle school chose not to send home permission slips though the forms were offered by theresearch administrator. None of the parents at one of the elementary schools asked that their children not answer the questionnaire. A small number of children at the second elementary school did not participate as requested by their parents. At the beginning of each administration it was announced that the survey was not a test; the children were told that they did not have to answer the research instrument. Moreover, 49 the children were told that if there were any questions that they objected to, they were to leave those questions blank. After the survey was completed each of the middle school classes was debriefed. The Principals at both elementary schools decided against a debriefing session for the fourth graders. A final debriefing in the form of a report detail- ing the results of this study has been promised the adminis- trators at the three participating public schools. Appendix "C" contains a statement affirming this survey's compliance with all school policies regarding the use of school children as participants in research projects. Statistical Analyses Central to the first eight hypotheses are the principles and assumptions of "multidimensional scaling." Psychological distance is the fundamental concept, with distance correspond- ing to the degree of perceived similarity among objects (Helm, Messick and Tucker, 1959). Similar objects, in other words, are thought to be psychologically closer to one another than are dissimilar objects. Paired-comparison judgments for all possible pairs of objects yield a similarity matrix which can be arrayed in a dimensional space using any one of several multidimensional scaling programs. The computer program Galileo (version 3.9) is utilized in this study (Serota, 1974).3 The statistical analyses for this initial section are (Xxnducted at an aggregate level. Analyses are performed for 50 the total sample as well as for subgroups composed of males- females, upper socioeconomic status-lower socioeconomic status subjects.4 Perceived similarity values for the fifteen oc- cupations are averaged across subjects in the total sample and the four subgroup conditions. These mean similarity values are used to generate conceptual spaces representing the work roles along as many as N-l dimensions. The occupa- tional roles represent points in the spaces--these points are located by referencing the occupations' coordinates along the N-l dimensions. For the total sample the Galileo program generated a "normal solution" with thirteen dimensions in the "real" (Euclidian) space. All of the dimensions provide information about the interrelationships of the occupations, but there is typically a point of marginal return. A "scree test" is often used to indicate the maximally informative dimensions, but for this sample there was no point ("elbow") at which the eigen values declined sharply (Tatsuoka, 1971; Barnett, 1976). The total sample was therefore randomly divided in half and normal solutions were generated for each half. These dimen- sional structures were then rotated to a least-squares best- fit. Zero-order correlations were computed between the cor- responding dimensions thus producing a measure of reliability. The first six dimensions were reliable at levels greater than +.966. Correlations between the seventh dimensions were +.823, fior the eighth dimensions "r" equaled +.615. It was decided 51 that the remainder of the multidimensional analyses would be performed using the highly reliable, first six dimensions which accounted for eighty-four percent of the variance in the MDS space. The mean values of the occupations on the unidimensional attributes were also computed in the total sample, male-female and upper SES-lower SES conditions. In the five conditions the mean attribute values for the occupations were then sta- tistically regressed on the first six dimensions. The result- ing multiple correlations indicated the degree to which the attributes were present in the multidimensional spaces; the standardized beta weights revealed the attributes' positions with reference to the six dimensions. Hypotheses one through nine were tested in this fashion. Indices were then constructed for measures of exposure to television versus non-television sources of information about occupational roles. All indices were additive with the constituent variables standardized. Indices of general media exposure to the occupational roles, interpersonal discussions about occupations and direct contact with occupations were furthermore weighted using factor score coefficients.5 Similar procedures were followed in computing indices of GXposure to TV programs with Specific occupational content. Once again, for the index of shows representing programs focus- ing'on police officers, weights were assigned to the five Ennograms using factor scores. The remaining indices contained 52 no more than two items. No weights were assigned in these instances. To test Hypothesis 10, which concerned the homogeniza- tion of societal members, the interpersonal discussion and the direct contact indices were summed. This created a non- media sources index. A 2 x 2 matrix was then computed using the general media exposure index (media source) as the ordi- nate and the non—media source index as the abscissa. Media Source 0 + Non-Media Sources The quadrants were fixed by dividing each index at its median. Respondents fitting in quadrant four were noteworthy for two reasons. (1) They had indicated that they frequently viewed TV programs emphasizing the seven occupational roles. (2) Relatively speaking, these respondents' families and friends infrequently discussed the occupations, and the reSpondents infrequently saw persons with the seven work roles. This quadrant was labeled a single source condition; in relative terms these respondents shared a common "primary" source of information and had less Opportunity to gain information from "secondary" sources. Persons fitting in quadrants one, two 53 and three received a greater proportion of their occupation- al information from "secondary" sources. These three quad- rants were labeled a mixed sources condition. Within the single source vs. the mixed sources condi— tions the subjects were divided according to gender and af- fluence. Multidimensional spaces with the seven media- emphasized occupations were generated for males, females, lower and upper SES subjects. Hypothesis 10 was tested by correlating the spatial configurations and the mean distances between spatial concepts for demographic subgroups in the single source versus mixed sources conditions. A parallel procedure was followed to examine the hy- pothesis of a stereotyping effect (H ). Initially, for the 11 total seven media-emphasized occupations, comparisons were made between respondents in single versus mixed sources con— ditions. The average variances around the similarity esti- mates for the occupations in the conceptual spaces were compared across single and mixed sources conditions. Next, for each Specific occupation a 2 x 2 matrix was generated, thus dividing respondents into single as opposed to mixed sources conditions for individual work roles. Comparisons were made between the two source conditions by computing F- tests for homogeneity of variances around the mean attribute ratings for each occupation. Tests of the remaining two hypotheses required a shift from aggregated to non-aggregated data. Zero-order 54 correlations were computed between frequency of viewing par— ticular television programs emphasizing occupational roles and the perceived "importance" of the particular occupations (H12). Conditional analyses were conducted dividing the sample in half. The respondents were separated according to whether they were above or below the median in exposure to non-media sources about each specific occupation. This rep- resents a slight modification of the single-mixed sources paradigm, as there was but one median split for these anal- yses. The final hypothesis (H13) was tested by zero-order correlations between frequency of viewing specific TV pro- grams dealing with work roles and the perceived distributions of persons holding the media-emphasized occupations. Where statistically significant correlations were found, corrections were made for attenuation to give a more accurate estimate of the sizes of the relationships in the pOpulation.6 Here again, conditional analyses were performed dividing the sam— ple at the median score on the non-media sources indices for each occupation. An alpha level of p<.05 was established prior to anal- yses. Statistical values and significance levels of data counter to hypotheses are not reported. 55 Notes These questions are arranged by topic for purposes of presentation here. In the actual questionnaire, ques- tions from page 112 did come first, followed by the. measures on pages 113-130. Items tapping respondents' perceptions of the number of people actually holding different work roles (page 150) were then asked. Sub- jects finished the questionnaire for the first day by indicating their frequency of contact with secondary sources (pages 147—149) and their relative SES standing (page 151). The questionnaire on the second day began with the measures found on pages 131-146. Exposure levels to specific shows were asked next (page 151). The subjects then handed in their questionnaires and received a shorter booklet repeating previous questions as a check on reliability--particu1arly the reliability of ques- tions asked the previous day. The researcher is indebted to the directors and members of the Children and Social Television Learning (CASTLE: MSU) research team for these survey measures. Galileo (version 3.9) is an improved version of the computer program described in Serota's thesis. Largely through the efforts of Richard A. Holmes and Michael J. Bonkowski this program has been tailored to meet the needs of communication researchers. Documentation of this program is available through: Communication Research Services, Inc. P. O. Box 1024 East Lansing, MI 48823 Finally, the contributions of Michael Cody and Rick Holmes to this study need to be recognized. They were of great assistance in the analyses of the Galileo spaces. Socioeconomic status was measured by a weighted, summed index of the affluence measures described on page 153. Questions about the parents' occupations proved to be nearly useless, children do not appear to know what their parents do at work. Weights were assigned to the SES scales using factor score coefficients (see the following note). Lower—upper SES was decided by using a median split on the weighted index. 56 Rationales and techniques for deriving factor score co- efficients can be found in Nie 31 31. (1975), as well as Seibold (1975). The weighting coefficients for each variable correspond to the variable's loading on the underlying factor. The composite index results by multiplying the standardized variable by its respective factor score coefficient and then summing all of the variables which define the underlying factor. Corrections for attenuation are performed to obtain estimates of strength of relationships in the population. Measurement error reduces the size of a correlation. The formula provided by Nunnally (1967:218) corrects for the unreliability of measures. _ 12 12 / r11 r22 HI If both scales were perfectly reliable, E would be the expected correlation. The correlation can be made for one of two variables by placing that variable's reli- ability coefficient under the radical. CHAPTER IV RESULTS The empirical findings of this study are presented in a manner approximating the order that the research hypotheses were offered. Hypotheses are restated, descriptive and in- ferential statistical results are then reported as they support or fail to support the predictions. An overview of these results is presented in conclusion. Early in the first chapter it was proposed that children differentiate occupational roles along three personal attributes. H1 Children distinguish occupations on the basis of the sex of the role holders. H Children distinguish occupations on the basis of the perceived competence (intel- ligence) of the role holders. H3 Children distinguish occupations on the basis of the perceived moral worth (good- ness) of the role holders. Respondents evaluated the fifteen occupational roles along these three attributes using unidimensional measures. The six multidimensional coordinates for each occupation were then correlated with unidimensional means of each occupa- tion on these attributes. Standardized regression weights found in Table 1 indicate the degree of association between 57 58 mean attribute ratings and the six orthogonal dimensions. Multiple correlations appearing 13 33 extreme right-hand column indicate the degree 33 which the attributes are present 13 the multidimensional space.l An examination of Table 1 reveals that the occupations' personal attribute ratings are significantly related to the occupational loadings on at least two of the three initial dimensions. More importantly the multiple correlations are .82, .93 and .94 for the "sex," "intelligent" and "good" attributes, respectively. The latter two coefficients are statistically significant; the multiple "r" for the sex attribute approaches but fails to meet the alpha level for rejecting the null hypothesis with the six-dimensional solu- tion. All three multiple correlations are squared and cor- rected for shrinkage.2 In an ideal sampling situation these final coefficients are estimates of the amount of variance in the first six dimensions of the population space account- ed for by the personal attributes. In sum, intelligence and perceived goodness of the occu- pational roles appear to be major attributes in children's evaluations of occupations. There is limited evidence that sex is used as an attribute in differentiating work roles. This attribute is substantially less important, however. Hypotheses 4, 5 and 6 concerned relational attributes associated with work roles. It was predicted that children would discriminate between occupations on the bases of (4) 59 Ama u zv .ommxcflusm Mom cofluoonuoo umuwm coflumHmHHoo mamfluase cmumsom ecu no mess a mo. v a .1 oo. too. ms.- HH.- mo. smm.i HH.- hoo.i mocmuhoaeH poo oo. «oo. oo.- mH.- sow. «om. mo. hao.- mcoauaocoo xuoz moflz oo. sow. NN.- tom.- om. o~.- «om. soo.- room Hmaumumz mo. too. oo.- m~.- so.. *No.- mm.u hmo.- mumnuo mash: mo. too. mm.- -.n «so. as. too.u som.- omens so. too. oo. oo.- ms. too. oo. tom. muoowo mm>amomm om. so. mm. HH.- om.- os.u oo.u iso.i humouo mm>flo HmcoflpmHmm mo. too. HH.- smm.- Ho. som.u soo.- soo.- oooo om. “mo. so. imm.- mo. o~.- Ho.u *No.- ucmoassmch mo. mo. mo.- HH.- oo.- som.u on. too. leans u hotshot xmm HMGOmHom o m o m N H bmm m o o a a a o .2 Hence u0m.mousnwuuu¢ anemone: cam HmcoHumHmm .Hchmumm so mmcfiumm cofiummsooo one: poacowm on com: mcoflmcmeflo accomonuuo xfim How mucmfioz coflmmoumom conwcumccmum .H magma 60 which work roles gave and received orders, (5) which role holders were generally liked by other people and (6) which occupations helped other people. Results found in Table 1 again indicate that the attributes are all significantly re- lated to one or more of the first three dimensions. The multiple correlation for the "give orders" attribute closely approaches, though fails to achieve, statistical significance. The remaining variables "receives orders," "liked" and "helps others" appear as attributes of considerable weight in chil- dren's evaluations. ‘ A final group of predicted attributes concern material and resource benefits associated with jobs. It was proposed that: H7 Children distinguish occupations on the basis of the perceived income of role holders. H Children distinguish occupations on the basis of the perceived work conditions of role holders. H Children distinguish occupations on the basis of the perceived status (importance) of role holders. Regression analyses, again in Table 1, provide evidence supporting these hypotheses. Multiple correlations for the "material" attributes are above .85; all are significant by the conservative statistical tests with eight degrees of freedom. In particular, work conditions are highly related to children's estimates of similarity in work roles. Wealth and job importance are similarly of value in accounting for the variance in children's occupational role perceptions. 61 A graphical representation of the attributes having statistically significant correlations with dimensions one and two in the Total N space can be found in Figure 1. Three attributes have been sketched into the two-dimen- sional plane; the designating symbols have been placed at the positive ends of the attributes. Locations of the occu- pations on a trait can be established by drawing a vertical axis from the attribute line to each work role in the space. The closer the role's location to the designating symbol, the greater the mean value of the occupation along that attribute. Figures 2 and 3 contain graphical representations of the attributes significantly related to the first and third plus the second and third dimensions. Once more, these plots correspond to the statistical configuration identified in Table l. The graphical configurations allow visual inspec- tion of occupation-trait spatial images held by children. Subordinate hypotheses were offered in relation to four of the attributes. Hypothesis la prOposed that males would distinguish occupations more on the basis of sex than females. Table 2 allows comparisons of the standardized regression weights for the two sexes. Clearly, there is no support for Hypothesis la. The multiple "r" is .82 for males and .81 for females in the six-dimensional solution. Further ex- ploratory examination of the multiple correlations for the F—"-— - .__...— -._-._--r- ._-‘- 62 J J0 A? (m c .13 .14 A3 Good A Liked 1 Police Officer 2 Doctor 3 Nurse 4 Private Detective 5 Paramedic 6 Truck Driver 7 Lawyer Figure l. 10 11 12 13 14 15 A Rich Housewife Teacher Banker :Mail Carrier Secretary Janitor Mechanic Factory Worker Graphical Representation of Three Attributes on Dimension One and Dimension Two of the Conceptual Space for the Total Sample. 63 A3/ A7//‘ A10/ Al Sex (female - male) A3 Good A5 Receives orders 1 Police Officer 2 Doctor 3 Nurse 4 Private Detective 5 Paramedic 6 Truck Driver 7 Lawyer Figure 2. 8 9 10 11 12 13 14 15 A7 Helps others A9 Nice work conditions A10 Job importance Housewife Teacher Banker Mail Carrier Secretary Janitor Mechanic Factory Worker Graphical Representation of Six Attributes on Dimension One and Dimension Three of the Conceptual Space for the Total Sample. 64 I33 543 ~*~“‘ 3*‘“‘”-‘”jfi“‘ D2 2 J4 5 A J ,5 A3// A3 Good 1 Police Officer 8 Housewife 2 Doctor 9 Teacher 3 Nurse 10 Banker 4 Private Detective 11 Mail Carrier 5 Paramedic 12 Secretary 6 Truck Driver 13 Janitor 7 Lawyer l4 Mechanic 15 Factory Worker Figure 3. Graphical Representation of One Attribute on Dimension Two and Dimension Three of the Conceptual Space for the Total Sample. 65 “ma u zv .mmmxcflucm Mow cofluoouuoo nopmm coflumamuuoo oameuase coumsom on» no mass 3 mo. v a k. 0v. Hm. mo.l vH.I mo. «ov.l OH.I shm.l mHmEmm oo. .oo. oH.- oo.u mo. «oo.i ma.- smm.- cams monophoass noo vm. smm. mo. mo.l «Hm. *HN. mo. «Na.l meEmh Hm. «hm. «HN.I hH.I «mm. «mm. Ho. tmm.l mam: mcofiuflcsoo xuo3 moflz mm. sum. 50. hm.l Hm. MH.I «Ne. «hm.l mHmEmh mm. «mm. tmv.l smv.l Hm. mm.l om. mm.l DAME SUHm mm. «mm. mH.I mH.I mo. mm.l Nm.l «mm.l mHmeh oo. too. oo. om.- oo.- .oo.. Hm.u «mm.u mam: mhmnuo moses mm. «mm. hH.I NN.I tam. mm. «ov.l «mv.l mHmEmm mm. «mm. tmv.l mH.I soc. vo.l «mm.l mN.I mam: Umxflq em. vm. mo. mo. NH. «me. 00. «we. mHmEmm oo. .oo. mo.- oH.- oH. tom. mo.- «so. moms msmoho mm>amomm mm. «mm. ma. OH.I mm.l NH.I HH.I «#5.! mHmEmh om. too. am. os.- om.i oH.- oo.- «oo.- was: humouo mw>ao mm. «mm. mo. VN.I HH. HN.I mm.l amb.l mHmeh Mb. «Hm. «mm.l «NM.I mo.l «Nv.l «om.l «mv.l OHMZ @000 we. «am. om. «mm.| NH. mH.I No. «we.1 mHmEmm on. two. oa.- .mm.: oo.i .mm.- oo.u two.. was: commassmucH we. , Hm. mo. mo.l no.l «wm.l mm. «Hm. mamfimh we. mm. mo.l ho.l mo.t «mm.l Hm. amv. GAME AmHmEom + mHmEv xmm m m v m m H QNM m D O D D Q Q .muommnsm masses .m> was: mcoa4 mousnwuuud Hmcoflmcmsficflca so mmcflumm coflummsooo new: beacons on com: meowmcosflo accomocuuo xflm wow mucmfloz cowmmoumom cmuflcumccmum .m wanes 66 nine attributes indicates no dissimilarity pattern between sexes. The attributes "receives orders" and "job importance" do seem to be of greater importance for males--R2 differences are substantial between the sexes along these two attributes. Aside from this, the breakdown by sex largely replicates the finding in the first table. Three subordinate hypotheses were prOposed concerning differences between subjects from varying SES backgrounds. H4 Upper socioeconomic status children distinguish a occupations more on the basis of the perceived autonomy (receives orders) of role holders than do lower socioeconomic status children. H Lower socioeconomic status children distinguish a occupations more on the basis of perceived in- come of role holders than do upper socioeconomic status children. H Upper socioeconomic status children distinguish a occupations more on the basis of work conditions of role holders than do lower socioeconomic status children. Comparisons found in Table 3 do not support the three hypotheses. Contrary to previous research, lower SES re- spondents (in this sample at least) distinguish occupations more on the basis of receiving orders than do upper SES re- spondents. The wealth associated with the fifteen work roles is more important in the occupational perceptions of more affluent respondents--the "rich" attribute has an R of .91 for upper SES children as opposed to an R of .83 for lower SES children. 67 Ams n CV .mmcxcsusm How coHuomuHoo Houmw scoumaosuoo mamouase cmsmsvm can no mosh o mo. v a k oo. so. oo.: oo.- os. os.i ss.u oo.: swam: No .oo No I ms 1 so . om u os n oo u Hoses mucosHOQEs men so. soo. oo.: oo.- so. oo. oo. oo.- seams mo soo oo . so u so os oo oo . noses ocosusocoo xuoz mosz so. «so. oo.- oo.- oo. oo.- oo. os.u swam: so so om I om . so os u om . so a Hoses nose oo. so. oo. oo.- so.i oo.- so.u oo.- seam: oo too mo u os . no u so u oo u oo u oozes humane oases oo. soo. oo.- oo.- «oo. os. «oo.- soo.n seams e e 0' 0' e 0|. 0|- e' O os so so so «so no so om has s omxss oo. so. so. oo.- oo. smo. so.: «so. seams so «so oo oo . om «mo mo soo smzos osmoso om>smomo os. oo. oo. oo.- oo.- os.u oo.: «so.u seams mo so ss so u om . ms . os u .oo . noses osmoso om>so oo. «oo. so.- «oo.- oo. oo.- .ms.- «oo.: seams so. soo. so.. so.. oo. «oo.- «os.u «oo.- oozes oooo oo. smo. os. «so.u oo. s~.- so. «oo.- some: so «so sow . sos . mo sow . oo u sso . sm3os scmosssmscs os. oo. oo. ss.- so.- «oo.- oo. «os. swam: os. so. oo.- oo.- oo.- «so.u os. hos. hoses . AmHmE + msmEmmv xo m m on on so on on so on .muoonnsm mmm momma .m> Hozos macEd mmusnsuuu< smconcoEocoss co mossumm coHummsooo com: weapons on com: mGOHmcoEHQ secomosuuo xom How mucmowz COommwummm couwcumcccum .m msnme 68 Finally, there are no differences between subjects' perceptions of occupations along the work conditions attri- bute when SES comparisons are made. These findings are in— congruent with studies reviewed earlier. Breakdowns by SES, much like the sex variable, mainly demonstrate the similar- ities between subgroups and replicate Total N results. Having established the general commonality of attributes contributing to children's perceptions of work roles, the analyses shifted next to the core issues of this study. Specifically, interest centered on the effect of children's receiving their occupational information from a "primary" source (television) on their overall images of the work world. This was examined by contrasting the occupational images held by respondents receiving information principally from a single institutional source (television) with the images held by re- spondents reporting 31333 sources of information. It was argued earlier that there ought to be greater homogeneity in the occupational images held by children with shared sources of information; television was proposed as a common source available to both males and females, upper and lower SES children. Thus it was hypothesized that: Hlo There will be greater homogeneity in the occu- a pational images held by male and female chil— dren with shared single sources of information than in the occupational images held by male and female children with mixed sources of information. 69 H10 There will be greater homogeneity in the b occupat1onal images held by lower and upper SES children with shared single sources of information than in the occupational images held by lower and upper SES children with mixed sources of information. In testing the hypotheses, attention centered on the seven occupations earlier identified as media-emphasized work roles. Respondents were grouped in the single vs. mixed sources conditions by using median splits on informa- tion source indices. These indices represented frequency of viewing work roles on television and exposure to non-media sources of information about the same seven occupations. Sex and SES splits were also made; care was taken to maintain very nearly equal cell sizes within source conditions. Empirical analyses took two forms. To begin with, dif- ferences in occupational images were compared for the two sexes in single and mixed sources conditions. Each occupa- tion's spatial positioning was compared. Counter to expec- tations there are greater differences in male and female spaces in the single source condition. The average mean difference is .707 in the single source as compared to .671 in the mixed sources condition (Table 4). Visual representa- tions of the differences are found in Figure 4 and 5. For males and females with mixed sources there is unex- pected congruence in their images of police officers, pri- vate detectives, truck drivers and lawyers. Note the varying degrees of agreement between the source conditions, 70 moses soo. muss: ooo. mocmsmmsso new: ossco oso. muss: oms.s houses muses sos. muses ooo.s sm>suo sense muses ooo. muses ooo. usomEesma muons new. moon: awe. o>suoouoa oum>oum muses sso.s muses smm. mossz muss: ooo. muses sos. sosooo muses som. muses ooo.s smossoo mosses soHusccou GOouHccoo sconH>osoev coHummnooo mmousom cmxoz so COsuomom condom camcom so COAuHmom smoumom so moocouommoo smoummm so moocmnommoo .mCOHuoccou moousom cows: .m> mousom osmssm so mpomnosm osmEmm cam osmz mo cams mommEH smooHummsooo somsumm moocmummmoo .s wanes 71 D - D Plane J 7 Dz'D3 Plane 1 Police Officer 5 Paramedic 2 Doctor 6 Truck Driver 3 Nurse 7‘ Lawyer 4 Private Detective Figure 4. Graphical Representation of the Mean Differences Between Male and Female Occupational Perceptions in the Single Source Condition. 72 D - D Plane 1 2 RIF/IN >l< ‘3 ,2 Plane .1” / A, D2"33 Plane 1 Police Officer 5 Paramedic 2 Doctor 6 Truck Driver 3 Nurse 7 Lawyer 4 Private Detective Figure 5. Graphical Representation of the Mean Differences Between Male and Female Occupational Perceptions in the Mixed Sources Condition. 73 especially for the occupation of truck driver. Generally speaking, the dissimilarities of perceptions between males and females are not great in either condition. But there is still largely consistent and apparently greater agreement in the mixed sources condition. Correlations were next computed between the male-female spaces in the two conditions. 'Table 5. Zero-Order Correlations of the First Four Corre- sponding Orthogonal Dimensions of the Occupational Spaces Held by Male and Female Subjects in Single Source vs. Mixed Soures Conditions.a Dimensions in Male Single Source Mixed and Female Spaces (Television) Sources First .9953* .9987* Second .9175* .9983* Third .9499* .9976* Fourth .9089* .8984* :1_ p <.05 The first four dimensions were chosen as there are four dimensions in the real space of males in the single source condition. The remaining spaces each have five real space dimensions. Zero-order relationships between the corresponding orthogo- nal.dimensions demonstrate the tendency for the two sexes to 'be in greater agreement in the mixed sources condition. It is important to note, nonetheless, that the correlations are high in both conditions and minimally discrepant. 74 The mean differences and the correlational findings do not buttress Hypothesis H In fact, the results are 10a counter to the prediction. The greater homogeneity between the male and female images in the mixed sources condition definitely warrants a re-evaluation of the underlying theo- retic position. Parallel analyses were conducted comparing lower and upper SES respondents' occupational images in the two source conditions. Once more there is greater disagreement in the occupational images for respondents reportedly receiving a larger prOportion of their information from television (Table 6). Police officers, nurses, private detectives, truck drivers and lawyers are perceived more similarly by lower and upper SES respondents with mixed sources. Only the means for the work roles doctor and paramedic appear in the hypothesized manner when SES and source distinctions are made. Differences between mean perceptions overall are smaller between SES subgroups within both conditions. There is a mean difference of .572 units between spaces for SES groups in the single source condition; a mean difference of .377 units separates spaces for SES groups with mixed sources. Pearson r statistics (Table 7) for the corresponding di- mensions are again quite high for all respondents. Four dimensions in the lower vs. upper SES spaces are correlated +.93 in both source conditions. The fifth spatial dimensions are very dissimilar for respondents with a common single 75 muss: mom. muss: mom. mucoummmoo coo: muse: soo. muses moo. sesams muses mos. muses oos. . sm>sso xosse muses ooo.s muses soo. usemsmsma muses mam. muons 5mm. m>ouomuoo mum>oum muses ooo. muse: ooo. mossz muses mss.s muse: mos. sosooo muses ooo. muses som. smasomo mosses c0suoccoo coouoccou scoomw>osoev cOsummsooo moousom cmxoz so soouomom mouoom oamcom so coauomom swoummm so mmocmuommoo smoummm so moocoummmsn .mcoHuHccoo moonsom cmxaz .m> condom msmcom as muoonnsm mmm some: can Hosos so csom mommEH secooummsooo soo3uom neoconmmmso .m wanes 76 D1 - D2 Plane >+< {N4 Plane \, 2 3 Plane 1 Police Officer 5 Paramedic 2 Doctor 6 Truck Driver 3 Nurse 7 Lawyer 4 Private Detective Figure 6. Graphical Representation of the Mean Differences Between Lower and Upper SES Subjects' Occupation- al Perceptions in the Single Source Condition. 77 D - D Plane >*< {A ane \d \ / / 7 \ Dz ’ D3 Plane 1 Police Officer 5 Paramedic 2 Doctor 6 Truck Driver 3 Nurse 7 Lawyer 4 Private Detective Figure 7. Graphical Representation of the Mean Differences Between Lower and Upper SES Subjects' Occupation- al Perceptions in the Mixed Sources Condition. 78 source of information-—"r" equals +.1l. Among the respon- dents with differing SES backgrounds and relatively mixed sources of information the correlation for the fifth dimen- sion remains above +.98. Table 7. Zero—Order Correlations of the First Five Corre- sponding Orthogonal Dimensions of the Occupational Spaces Held by Lower and Upper SES Subjects in Single Source vs. Mixed Sources Conditions.a Dimensions in Lower Single Source Mixed and Upper SES Spaces (Television) Sources First .9978* .9990* Second .9908* .9940* Third .9658* .9982* Fourth .9355* .9967* Fifth .1140 .9807* at p <.05 a e c c I I The first f1ve d1mens1ons were chosen as there are f1ve dimensions in the real spaces in all conditions. 10 is not accepted. The comparatively b greater agreement of work role images among subjects with Hypothesis H mixed information sources was unforeseen. Television has not homogenized occupational perceptions in the manner ex- pected. In the single source condition, greater differences appear in the mean spatial images reported by lower and upper SES children; zero-order correlations comparing di- mensional arrays of occupations are consistently lower. 79 Not only is the hypothesis not supported, findings are also in the opposite direction. Chapter I also argued for a homogenization of images in the form of a stereotyping effect. While earlier hypoth- eses predicted increased agreement in mean perceptions of occupational roles, Hypothesis H11 proposed a reduction in variances around mean occupational perceptions for persons sharing a common information source. H11 There will be greater stereotypic views of occupat1onal roles by ch11dren w1th shared single sources of information than by chil- dren with mixed sources of information. In the data collection children were allowed to indicate that they were not sure of the answers to the paired-com- parison and unidimensional attribute questions. For the following analyses, responses indicating such uncertainty were treated as missing data.3 Respondents were categorized into single and mixed sources conditions. Average variances for the paired-com- parison measures of the seven television occupations were computed. An average variance for every occupation was thus based on approximately 200 observations repeated across six different work roles (i.e., 1200 + measures). The average variances around the means in the paired comparison items are all in the hypothesized direction, save for the occupation private detective. Children with single sources of information are in greater agreement as to the 80 Table 8. Average Variances for the Paired-Comparison Measures of Seven Television Emphasized Occupa- tions Among Subjects with Single vs. Mixed Sources of Information. Average Variance Average Variance Among Subjects Among Subjects With Single With Mixed Occupation Sources Sources Police Officer 1.266 1.479 Doctor .763 .849 Nurse .685 .749 Private Detective 1.062 1.020 Paramedic .988 1.118 Truck Driver .689 .843 Lawyer 1.107 1.138 overall similarities among police officers, doctors, murses, paramedics, truck drivers and lawyers. Statistical tests are deferred here, as there is uncertainty as to the number of degrees of freedom associated with each average variance. Attribute measures for all seven occupations were ex- amined. Though the "sex" and "gives orders" attributes' multiple "r's" fell short of statistical significance in the six-dimensional solution, there was evidence of the attri- butes' existence in children's evaluations. All ten attri- butes were therefore examined across the seven occupations for the hypothesized stereotyping effect (Tables 9-15). It should be mentioned first that there is mixed evi- dence concerning the reduction in variances around attribute 81 Table 9. F-tests Comparing Variances Around the Mean Attri- bute Ratings of Police Officers by Subjects with Single vs. Mixed Sources of Information. Attribute Condition N Variance F Value Sex Single 53 .30 Mixed 157 .31 1.03 Intelligent Single 17 1.19 Mixed 58 1.57 1.32 Good Single 27 .58 Mixed 74 1.21 2.08* Gives orders Single 22 1.19 Mixed 60 1.16 Receives orders Single 42 50 Mixed 124 50 Liked Single 17 1.19 Mixed 58 1.34 1.13 Helps others Single 33 .12 Mixed 90 .68 5.62* Rich Single 16 .46 Mixed 48 .66 1.42 Nice work conditions Single 11 1.86 Mixed 41 1.25 Job importance Single 40 .10 Mixed 123 .27 2.66* * p < .05 (one-tailed test) Table 10. F-tests Comparing Variances Around the Mean Attribute Ratings of Doctors by Subjects with Single vs. Mixed Sources of Information. Attribute Condition N Variance F Value Sex Single 52 .21 Mixed 158 .32 1.53* Intelligent Single 28 .00 Mixed 110 .29 0.00* Good Single 39 .10 Mixed 122 .07 Gives orders Single 32 .63 Mixed 97 .56 Receives orders Single 40 .20 Mixed 123 .30 1.45 Liked Single 29 .49 Mixed 86 .45 Helps others Single 41 .19 Mixed 139 .06 Rich Single 18 .85 Mixed 43 1.15 1.35 Nice work conditions Single 26 .66 Mixed 63 .68 1.02 Job importance Single 47 .09 Mixed 137 .07 * p < .05 (one-tailed test) Table 11. F-tests Comparing Variances Around the Mean Attribute Ratings of Nurses by Subjects with Single vs. Mixed Sources of Information. Condition Attribute N Variance F Value Sex Single 41 .63 MiXed 167 .43 Intelligent Single 14 1.14 Mixed 59 .78 Good Single 18 1.06 Mixed 92 .44 Gives orders Single 10 2.32 Mixed 55 1.58 Receives orders Single 26 80 Mixed 104 40 Liked Single 14 2.11 Mixed 76 .70 Helps others Single 20 1.52 Mixed 90 .36 Rich Single 12 1.48 Mixed 49 .73 Nice work conditions Single 14 2.06 Mixed 65 1.02 Job importance Single 14 1.14 Mixed 77 .61 * p < .05 (one-tailed test) Table 12. F-tests Comparing Variances Around the Mean Attribute Ratings of Private Detectives by Subjects with Single vs. Mixed Sources of Information. Attribute Condition N Variance F Value Sex Single 50 .44 Mixed 158 .36 Intelligent Single 16 1.60 Mixed 63 .65 Good Single 14 1.14 Mixed 52 6O Gives orders Single 17 1.62 Mixed 66 1.19 Receives orders Single 33 .96 Mixed 87 .51 Liked Single 11 1.82 Mixed 30 1.68 Helps others Single 20 1.09 Mixed 55 .42 Rich Single 13 1.69 Mixed 40 1.28 Nice work conditions Single 14 1.87 Mixed 37 1.54 Job importance Single 21 1.05 Mixed 66 .57 * p < .05 (one-tailed test) 85 Table 13, F-tests Comparing Variances Around the Mean Attribute Ratings of Paramedics by Subjects with Single vs. Mixed Sources of Information. Attribute Condition N Variance F Value Sex Single 42 .37 Mixed 165 .37 Intelligent Single 21 .19 Mixed 92 .51 2.69* Good Single 27 .00 Mixed 108 .54 0.00* Gives orders Single 22 .35 Mixed 84 .85 2.46* Receives orders Single 25 .50 Mixed 87 .41 Liked Single 24 .17 Mixed 97 .53 3.21* Helps others Single 30 .00 Mixed 116 .21 0.00* Rich Single 14 1.08 Mixed 38 1.33 1.24 Nice work conditions Single 13 .92 Mixed 50 1.47 1.59 Job importance Single 30 .26 Mixed 121 .30 1.15 * p < .05 (one-tailed test) 86 Table 14. F-tests Comparing Variances Around the Mean Attribute Ratings of Truck Drivers by Subjects with Single vs. Mixed Sources of Information. Attribute Condition N Variance P Value Sex Single 46 .35 Mixed 159 .31 Intelligent Single 16 1.07 Mixed 45 1.31 1.23 Good Single 17 .72 Mixed 47 .69 Gives orders Single 27 .50 Mixed 80 .46 Receives orders Single 24 .41 Mixed 83 .94 2.32* Liked Single 10 1.95 Mixed 48 1.16 Helps others Single 13 .77 Mixed 52 .84 1.09 Rich Single 16 .76 Mixed 67 .84 . 1.10 Nice work conditions Single 19 .81 Mixed 80 .83 1.03 Job importance Single 16 1.53 Mixed 52 .62 * p < .05 (one-tailed test) 87 Table 15. F-tests Comparing Variances Around the Mean Attribute Ratings of Lawyers by Subjects with Single vs. Mixed Sources of Information. Attribute Condition N Variance F Value Sex Single 42 .48 Mixed 167 .36 Intelligent Single 29 .14 Mixed 80 .48 3.51* Good Single 25 .16 Mixed 87 .69 4.32* Gives orders Single 27 .63 Mixed 68 .76 1.21 Receives orders Single 30 .53 Mixed 88 .64 1.21 Liked Single 14 .00 Mixed 53 1.00 0.00* Helps others Single 25 .00 Mixed 83 .28 0.00* Rich Single 14 .29 Mixed 75 .95 3.33* Nice work conditions Single 15 1.55 Mixed 61 1.05 Job importance Single 34 .00 Mixed 84 .28 0.00* * p < .05 (one-tailed test) 88 means. The variances associated with mean attribute ratings for two occupations offer no support. Perceptions'of nurses (Table 11) and private detectives (Table 12) are not more similar for respondents in the single source condition. For the other five occupations there is evidence of the hypothe- sized stereotyping effect. Tests for homogeneity of variances are employed at this point.4 F-values for the variances around the perceived "intelligence" and "rich" attributes are in the direction predicted for all five remaining occupations. Three of the "intelligence" and one of the "rich" variance ratios are statistically significant. Four of the five F-values for the "helps others" attribute were in the predicted direction; three are significant. Among the "good," "receives orders," "liked," "work conditions" and "job importance" attributes three of the five ratios are in the expected direction. Of these, statistical significance is achieved by all of the F-values for the "good" attribute, by two of the "liked" and two of the "job importance" F-values. Variances around the ratings along the final two attributes, "sex" and "gives orders," were as predicted in two cases each--one F-ratio in each instance is significant. There is evidence of a stereotyping effect especially for the occupations paramedic and lawyer. Of the ten vari- ance ratios associated with the work roles, eight are in the hypothesized direction. Five of the paramedic attribute 89 variances and six of the lawyer attribute variances are sta- tistically significant. The remaining three occupations all have between five and seven variance ratios in the predicted direction; slightly less than half of these are significant. That so many of the variance ratios are in the predicted direction is encouraging. This is especially the case here. Under most circumstances, one would expect the variances for the large sample to be smaller.5 Extreme or deviant values have less effect on the standard error as sample size in- creases. In this analysis, it was predicted that variances would be smaller in the condition with decidedly fewer cases-- a prediction clearly counter to what would normally be expected. The findings reported here are encouraging, but not con- vincing. Hypothesis Hll is not uniformly supported. Chil- dren in the single source condition do tend to demonstrate greater homogeneity in terms of reduced variance about the mean, especially with the paired-comparison items. Caution is warranted, moreover, in suggesting that many of the uni- dimensional attribute findings are even tentatively suppor— tive of the hypothesis. An unexpected "edge-effect" appears in the scaling of the unidimensional measures. In the single source condition several of the attributes have no variance (e.g., Table 15), the subjects clustered at the highest values allowed. 90 Tests of the final two hypotheses required a shift in units of analysis. Bivariate and conditional bivariate analyses were performed for the 200 plus cases. The first research question at this level addressed the status conferral notion proposed by Lazarsfeld and Merton and later more narrowly defined by Lemert. Witnessing and learning about jobs through the "primary" source television was thought to carry with it an enhancement of the perceived importance of the occupational roles. H12 The greater children's exposure to tele- vision programming emphasizing an occupa— tional role, the greater the perceived relative importance of that occupation. H Among children receiving proportionally 12 . . . b more occupational information from tele- vision, the greater the exposure to tele— vision programming emphasizing an occupa- tional role, the greater still the perceived relative importance of that occupation. Measures of children's viewing of Specific shows empha- sizing occupational roles were correlated with their per- ceptions of how important were the media-emphasized roles. Measures of viewing behaviors were constructed concerning specific occupations as follows: Police officer--weighted summed index of the programs "Police Woman," "Hawaii 5-0," "Kojak," "Adam 12" and "Ironsides." Doctors--summed index of the programs "Emergency" and "Medical Center." Nurses--summed index of the programs "Emergency" and "Medical Center." 91 Private detective--summed index of the programs "Rockford Files" and "Cannon." Paramedics--measure of the program "Emergency." Truck drivers--measure of the program "Movin' On." Lawyers——measure of the program "Petrocelli." Table 16. Zero-order Correlations Between Frequency of Viewing Television Programs Emphasizing Partic- ular Occupational Roles and the Perceived Importance of the Occupations. TV Programs Emphasizing: N Zero-Order 'r' Police Officers 193 .06 Doctors 203 .01 Nurses 200 .09 Private Detectives 203 .09 Paramedics 204 .11 Truck Drivers 206 .06 Lawyers 202 ' .31* *p < .05 Zero-order correlations (Table 16) between frequency of viewing television programs emphasizing particular occupa- tional roles and the perceived importance of the occupations revealed no substantive findings. A single correlation was significant at the p < .05 level; the relationship between 92 viewing "Petrocelli" and the perceived importance of lawyers was +.3l. Conditional, correlational analyses were further performed. Table 17 indicates that the correlations relating to the work role lawyer behaved as expected. Table 17. Conditional Zero-Order Correlations Between Frequency of Viewing Television Programs Emphasizing Particular Occupational Roles and the Perceived Importance of the Occupation by Subjects in Single vs. Mixed Sources Conditions. TV Proqrams Single Mixed Emphasizing: Source Sources Police Officers -.01 .14 (n=102) (n=9l) Doctors .06 -.07 (n=115) (n=88) Nurses .05 .13 (n=100) (n=100) Private Detectives .10 .08 (n= 98) (n=105) Paramedics .04 .19* (n=107 (n= 97) Truck Drivers .07 .04 (n=102) (n=104) Lawyers .38* .25* (n=101) (n=101) * p < .05 A test for differences between independent correlations, nonetheless, was not significant (z=l.102). Correlations for the single source and mixed sources conditions were 93 otherwise irregular. The correlation between frequency of viewing "Emergency" and perceived importance of paramedics in the mixed sources condition was +.l9 and significant. These all but solitary findings were singularly unimpressive. Mean "job importance" values for the seven occupations were examined for the total sample. "Importance" values, in principle, could range from one through five. The seven media-emphasized occupations were evaluated in the succeeding manner: Occupation Mean Importance Rating Doctors 4.838 Police Officers 4.638 Paramedics 4.612 Lawyers 4.502 Nurses 4.353 Private Detectives 4.167 Truck Drivers 4.010 Such consistently high evaluations of media-emphasized roles are noteworthy. Returning to Figure 2, it can be seen how six of the media-emphasized roles cluster as compared to the remaining occupations. These consistently high evaluations of the seven work roles were most likely instrumental, in part, in producing the small, nonsignificant relationships found in Tables 16 and 17. Notwithstanding, this provides no ElSEE evidence for the hypothesized status conferral effect. A final set of hypotheses similarly concerned children's perceptions of the work world at the individual level. Argu- ments were made to demonstrate the relative isolation of many children from work environments. It was additionally 94 proposed that due to this isolation it was difficult for children to gain accurate information about the distribution of work roles. Children's images of what proportion of the population had a particular job were thought to be distorted by viewing repeatedly portrayed work roles in television. 813 The greater children's exposure to repeatedly a emphasized occupational roles on television, the greater the perceived number of persons thought to have the occupations. The same line of reasoning was carried a step further to deal with the perceived occupations held by women. H13 The greater children's exposure to occupa- b tional roles held by women on television, the greater the perceived number of women thought to have the occupation. The three most heavily emphasized occupational roles were focused on initially. Three nonstandard work roles for women were then examined. In the total sample the average perceived number of work role holders broke down in this manner: Out of every 25 people an average of 5.428 were thought to be police officers, 4.598 were thought to be doctors, and 4.703 were thought to be nurses. Out of every 25 police officers 5.952 were thought to be women. Out of every 25 paramedics 4.454 were thought to be women. Out of every 25 lawyers 5.556 were thought to be women. Given the complexity of asking children's estimates of how many peOple have particular jobs, a reliability check was included in the data collection. The measures of 95 internal consistency ranged from "r" equals +.35 to "r" equals +.60. The average correlation was +.47. Resulting significant correlations were corrected for attenuation to gain better estimates of the relationship sizes in the population. Table 18. Zero-order Correlations Between Frequency of View- ing Television Programs Showing Repeatedly Empha- sized Occupational Roles and the Perceived Number of Persons Having the Occupations. TV Programs Emphasizing: N Zero-Order 'r' Police Officers 191 .17* a [.22] Doctors 202 .26* [.38] Nurses 202 .23* [.33] * O p < 05 aThe values in brackets are the zero-order correlations corrected for attenuation. Zero-order correlational tests (Table 18) with job specific viewing indices provided significant relationships between watching relevant TV shows and perceptions of how many people have the portrayed occupations. After correcting for unreliability, the correlations were +.22 for the occu- pation police officer, +.38 for doctor and +.33 for nurse. Attention turned next to conditional analyses contrasting children with primarily television versus television plus 96 Table 19. Conditional Zero-order Correlations Between Frequency of Viewing Television Programs Showing Repeatedly Emphasized Occupational Roles and the Perceived Number of Persons Having the Occupations by Subjects in Single vs. Mixed Sources Conditions. TV Programs Single Mixed Emphasizing: Source Sources Police Officers .32* a -.03 [.41] (n=101) (n= 90) Doctors .19* .34* [.27] [.49] (n=114) (n= 88) Nurses .22* .25* [.31] [.36] (n=101) (n=101) * p < .05 aThe values in brackets are the zero-order correlations corrected for attenuation. outside sources of information. Among children receiving their information from single vs. mixed sources (Table 19), the difference between the uncorrected independent correla- tions concerning police officers is statistically signifi- cant (z=2.458). There is reason to believe, in other words, that with respect to the occupation police officer the rela- tionship between viewing behaviors and perceptions of numbers of role holders is greater if the children have limited outside information about the role. Differences between the conditional correlations relating to doctors and is nurses are not significant. Thus Hypothesis Hl3 a 97 endorsed by these empirical results. The prediction that children's occupational distribution images will especially be influenced in single source conditions (Hl3 ) does not receive uniform support. C Table 20. Zero-order Correlations Between Frequency of Viewing Television Programs Showing Females in Non-Standard Occupational Roles and the Perceived Number of Females Having the Occupations. TV Programs Show— ing Females as: N Zero-Order 'r' Police Officer 202 .20* a [.29] Paramedic 203 .17* [.29] Lawyer 198 .20* [.31] * p < .05 aThe values in brackets are the zero-order correlations corrected for attentuation. Table 20 contains the zero-order correlations between the frequencies of viewing television programs portraying females in nonstandard work roles (i.e., "Police Woman," "Emergency" and "Kate McShane") and the perceived numbers of females having the occupations. The relationships are small, still they are consistent and significant. Corrected correlations rurver around +.30 for viewing measures and perceptions of how many women have these jobs. 98 Table 21. Conditional Zero-order Correlations Between Frequency of Viewing Television Programs Showing Females in Non-Standard Occupational Roles and the Perceived Number of Females Having the Occu— pations by Subjects in Single vs. Mixed Sources Conditions. TV Programs Show- Single Mixed ing Females as: Source Sources Police Officer .33* .05 [.48] (n=105) (n= 97) Paramedic .19* .17* [.32] [.29] (n=106) (n= 97) Lawyer .35* .01 [.54] (n= 99) (n= 99) * p < .05 aThe values in brackets are the zero-order correlations corrected for attenuation. Conditional analyses found in Table 21 show substantial, significant differences between children in the contrasting source conditions. Children's perceptions of the number of women police officers and the number of women lawyers are dependent upon the availability of secondary sources of information. Differences between the conditional Pearson r's are significant for children's perceptions of how many police officers (z=2.049) and lawyers (z=2.472) are females. Rela- tively speaking then, there is no relationship between TV viewing and perceptions of how many women have the two work roles if children have sources of information other than 99 television. These data sustain Hypothesis H Children's per— l3b ceptions of the number of women in nonstandard roles are related to their television viewing behvaviors. Examining respondents in single source and mixed sources conditions (Tables 19 and 21), there are statistically significant dif- ferences between the correlations related to two occupations —-police officer and lawyer. Hypothesis H receives appre- 13 c ciable though not total support from these findings. Summary This discussion has detailed the empirical findings relevant to hypotheses derived in Chapter I. The first nine hypotheses were tested using multiple regression analyses. Seven of the nine predictions were confirmed. Multiple cor- relations for two attributes, "sex" and "gives orders," fail to reach .05 levels of significance using the conservative estimate of eight degrees of freedom. Children appear to use the attributes relating to intelligence, goodness, who re- ceives orders, popularity, helpfulness, wealth, work con- ditions and job importance in their perceptions of occupational roles. No systematic differences were found in the percep- tions by subgroups. Tests for homogeneity of occupational images for media- emphasized occupations were conducted. Mean perceptions by males-females, lower—upper SES children were compared. Con— trary to expectation, there was greater similarity in the 100 images held by children with more diverse as opposed to less diverse information sources. The differences between condi- tions were not great; the directions of the findings were, however, counter to the hypotheses. The homogeneity of variances around mean perceptions was examined next. Comparisons of the average variances for the paired—comparison measures, with one exception, appear in the manner predicted. By these measures, persons receiving a proportionally large part of their information from a primary information source tend to have more stereotypic views than persons with secondary or a mix of primary and secondary information sources. F-tests of variances around means of attributes associated with occupational roles pro- vide less consistent results. Children in the single as opposed to the mixed sources condition do not uniformly have more stereotypic views. The hypothesis of a status conferral effect traceable to the primary source television was not sustained. Hypotheses of television influencing perceptions of occupation distributions were supported. Conditional analy- ses further demonstrated partial support for the predictions that the effect would be greatest where secondary information is relatively absent. 101 Notes As noted by Rosenberg and Sedlak (1972:245), "the orthogonal axes in a multidimensional solution do not necessarily correspond to meaningful psychological dimensions; the location of these orthogonal axes is typically arbitrary and serves simply as a convenient way to describe the output configuration." These axes, however, can be used to establish the degree that a trait or attribute is related to the configuration. The square-root of the summed, squared zero-order cor- relations between the individual occupation's attribute ratings and the corresponding dimension loadings is the multiple correlation. This multiple R can be used as an indicator of the attribute's presence in the space. This correction is to counter the upward bias of the multiple R due to, among other things, the ratio of the number of predictor variables to the sample size. Thiel (1971:178) provides the following formula which was used to correct the multiple correlation appearing in the SPSS output. R2-(k-l/N—k) (1-R2) Adjusted R2 N number of cases (15) k = number of predictor variables (6) By removing the uncertain cases for each measure, Spurious findings of a clustering effect were avoided. This move, unfortunately, also reduced cell sizes considerably. The F-test for homogeneity of variances is normally a two-tailed test, where an F ratio is determined by placing the larger variance estimate over the smaller. In this study, a prediction was made calling for the variance estimate of the larger sample (mixed condition) to be smaller. Thus a one—tailed test will be used in interpreting those ratios appearing in the predicted direction. See Blalock (1972) for greater discussion about sampling distributions. CHAPTER V DISCUSSION This final chapter has two parts. A beginning portion is devoted to an examination of the major empirical findings. Following this, the focus shifts to a re-examination of the strengths and weaknesses of the underlying theoretic position. It is here that suggestions for future research are tendered. Perceived Occupational Attributes An integral step to the study of occupational percep- tions is found in the identification of attributes consitut- ing work role images. Ten attributes were hypothesized as contributing to occupational images held by children and ad- olescents. Eight of the attributes, indicated by significant multiple correlations, were as predicted in six-dimension space analyses. All ten attributes would be confirmed by a five—dimension space analysis-—the recovery of one degree of freedom is the critical factor. Examination of the spatial arrays of male, female, lower and upper socioeconomic status children reveals no consistent patterns. Though previous research suggests that male chil- dren more often note the sex appropriateness of roles, there was no apparent difference in this study. One interpretation 102 103 of this is that the males and females in this sample are equally sensitized to sex-roles. Could it be that the recent "liberation" movements are having an effect on chil- dren's occupational views? Are females at an early age now becoming more aware of sex roles, or is it that males are now assigning less importance to sex-role differences? Either explanation has merit. No differences were found with respect to the importance of the authority and material benefits attributes for lower versus upper SES children. This, too, is counter to expec— tation. A possible source of this aberration rests in the sample. Possibly major social and economic differences are minimally present. Extreme opulence and destitution were not apparent in this community. Furthermore, it is hard to gain detailed information about most subjects' SES standing. With children a researcher has the added complexity of re- spondents actually not knowing much about their parents' jobs and incomes. The scales used to measure affluence were highly reliable (r's > +.90). Possibly lower—upper quartile analyses of the scales would yield different results. Such comparisons are not made in this study, as lower-upper quartiles when further divided into information source con- ditions would reduce cell sizes below desirable levels. Cultural Homogenization A central concern of this study rests on the effects of primary sources on aggregate perceptions of social roles. 104 Numerous scholars have indicated that one function of mass communication involves the transmission of a common social heritage. Drawing upon this notion, hypotheses were tested which compared the homogeneity of mean occupational images between subgroups in single and mixed information sources conditions. Contrary to hypotheses, subgroups receiving a proportionally larger amount of information from a shared, primary source were legs homogeneous than subgroups report- ing primary and secondary information sources. These findings are at once puzzling and damaging. Why should there be greater differences in the aggregate occu- pational images of people who indicate sharing a common in- formation source? The effectiveness of this information source is called into question. Post hoc explanations of these findings can be found in several arenas. Two such interpretations are noted at this time. Shibutani (1965) has discussed the processes of be- lief formation. Among other things, Shibutani proposes that the clear definition of a social situation comes from reality testing. Uncertainty typically exists when a comparative function is not allowed, regardless of the original informa— tion source. Children in the single source condition indi- cated appreciably less discussion about and direct contact with the media-emphasized roles, hence little opportunity for testing their ideas. 105 Closely paralleling Shibutani's idea is a discussion by Woelfel (1976). Woelfel calls attention to the "error" that emerges in "low feedback systems" such as the mass media. Respondents with mixed sources of information re— ported media viewing to some extent, but they also indicated contact with and interpersonal communication about the work roles (a high feedback situation). It is possible that the respondents had sufficiently integrated interpersonal commun— ication networks in the mixed condition to allow corrections in perceptual errors. Neither of these post hoc hypotheses can be tested with this data set; both lines of research potentially have explanatory value, however. Stereotypes The hypothesis of a stereotyping effect receives extreme— ly mixed and inconclusive support. Variances around paired- comparison estimates appear largely as expected, though the differences are not great. Tests for homogeneity of occupa- tional ratings on attribute scales are inconclusive. Vari- ances around mean perceptions of nurses are not as predicted with the unidimensional attribute scales. Moreover, standard errors associated with the role private detective are not in the expected direction using either empirical measure. Hypotheses tested here do not deal directly with the stereotyping of particular occupations or stereotyping with regard to a particular attribute. The theoretic position advanced in Chapter I is cast at a higher level of 106 abstraction in hOpes of allowing generalizations. Notwith- standing this, the completely unexpected results related to nurses and private detectives deserve comment. An underlying assumption of this study has been that there are basic differences and inequities in the secondary information held by respondents. If nothing else, individuals were thought to have different role models to observe. In the case of nurses this may not be a tenable assumption. The students in the Grand Ledge sample do share a common school nurse. It would be valuable to know the extent to which sub— jects reporting contact with nurses and discussions about nurses were in fact indicating contact with and communication about the school nurse. There is no way of determining this here, but if it is the case, the small variances in attribute perceptions in the mixed condition are explicable. Variances around the ratings of the private detective role are another matter. There is no obvious explanation except possibly that no real differences between single and mixed conditions exist in terms of secondary information. The role is unique from all the others, as more than sixty percent of respondents claim to "never" talk about private detectives, and more than sixty-six percent report "never" seeing a private detective. Five-point scales were used to measure levels of exposure to non-media information. Eighty- seven percent of all respondents score "2" or less on the direct contact scale; ninety—three percent score two or less 107 on the interpersonal discussion scale. These subjects, therefore, could not be greatly differentiated in terms of their secondary information about private detectives. Aside from these two matters, any future research of this hypothesized effect ought to be conducted with more precise, and ideally unbounded, scales. The "edge effects" encountered in the attribute scales for several occupations make data interpretations difficult and necessarily inde- terminate. Status Conferral These results are in such agreement that further dis- cussion is very nearly superfluous. There is virtually no evidence to sustain this hypothesis. Figure 2 demonstrates the unusual clustering of the media-emphasized occupations along the "job importance" attribute. The seven occupations all rate above four on a five-point scale. No support, how— ever, is found for the prediction that frequency of TV viewing correlates with perceived importance of media occupa- tions. Results show the seven media-emphasized occupations to be generally perceived as important irrespective of tele- vision viewing behavior. Perceived Role Distributions Consistent, though small, positive correlations exist between watching an occupation on television and perceptions of how many people have the particular job. Having access 108 to secondary information does not appreciably affect, in any consistent manner, these perceptions of repeatedly emphasized roles. Differences between conditional zero-order correla- tions relating to portrayals of women in nonstandard occupa- tions are somewhat of an exception. Secondary information about police officers and lawyers seems to discount any effect of seeing women in the roles on television. Some might argue that this latter finding again demon- strates the effects of reality testing. That the effect does not also occur for female paramedics suggests additional com- plexity that is not understood. No other unusual, systematic differences appear with regard to primary and secondary infor- mation about paramedics. Conclusions and Suggestions The implications of these findings are considerable, especially for theories and research related to social learn— ing from primary information sources. Governmental and edu- cation institutions, plus commercial information and some entertainment enterprises, attempt to guide, teach and pro- voke awareness in mass audiences. With societies growing ever larger there is good reason to believe that such efforts will continue. Communication researchers will no doubt be called on more and more to assist in these activities. Popular notions about the ability of undifferentiated messages to inform and coalesce mass societal units are un— substantiated by the empirical results of this study. 109 Primary information sources, by themselves, appear to be inadequate for increasing cultural homogenization. Males- females and lower-upper SES subjects were more dissimilar in their role perceptions where a primary source was pre- dominant. But why are there systematic differences between these subgroups when television is the primary source?1 Is there in fact systematic "slippage" in the transmission of culture within subgroups owing to their use of primary infor- mation sources? (Nuxmztwo questions are ripe for future inquiry. Stereotyping as a function of primary vs. a mix of pri— mary and secondary information receives enough support here to justify additional research with more precise scaling and hopefully more equal cell sizes. In particular, the measurement of standard error around concepts in multidimen- sional spaces remains a challenge. This study averaged the variances of all paired-comparison items for each concept. This is too crude. The variance for eeeh concept should ideally be ascertained along eeeh dimension in the space. The solving of this methodological problem is needed. The systematic relationships found between exposure to specific media messages and perceptions of the distribution of occupational roles are intriguing. Investigators of this aspect of social role imagery would do well to seek addi- tional antecedent variables that contribute to the dependent phenomenon. The variance accounted for is still quite small. 110 Equally important, an additional step needs to be taken to link the dependent variable to subjects' expectations for themselves and their friends. This is especially true for the portrayals of females and males in nonstandard roles. As mentioned in Chapter I, several researchers have begun to address this topic. A next move is thus found in carefully relating this research to the literature on "role definers" and "role models."2 111 Notes Woelfel (1976:69) proposes that "these errors are not systematic . . . but are more or less random." Appearing under the rubric of "reference groups," and "significant others" there is a large body of literature discussing how individuals learn "subject-object" rela- tionships with reference to themselves. To this can be added the research by Bandura and others on "observa- tional learning." APPENDIX A Survey Questions T.) u. (fl, “unlmmm 112 What is your name? What is your sex? (Check one) Girl Boy How old are you? What.month is your birthday in? (Check one) ___danuary ___duly ___February ___August ___March ___Septemxar ___April ___October ___May ___November June ___December What grade are you in? Today we want to know'what you think about several jobs and your ideas about the people that have these jobs. This is not a test, there are no wrong answers. we want to know what you think. Please work quickly and try to answer all the questions. If there is any question that you object to, leave it blank. The first group of questions ask you to think about two jobs and then decide if the two jobs are "alike" or "different." Read the following questions and put an §_on the line above what yeu think. Here is an example: What do you think about Paramedics and Ambulance Drivers? Are they: completely very alike I'm not different very completely alike much sure different different alike 113 What do you think about Police Officers and Private Detectives? Are they: completely very alike I'm not. different very completely alike .much sure different different alike What do you think about Lawyers and Paramedics? Are they: crnpdetely very alike I'm not. different 'very completely alike nuxxi sure different different alike What do you think about Doctors and Private Detectives? Are they: completely very alike I'm not different very ccmpletely alike much sure different different alike What do you think about yourself and Police Officers? Are you: completely very alike I'ntnot different very completely alike much sure different different alike What do you think about Nurses and Paramedics? Are they: completely very alike I'm not different ‘very completely alike nuch sure different different alike What do you think about yourself and Lawyers? Are you: completely very alike I 'm not different very completely alike much sure different, different alike What do you think about Doctors and Truck Drivers? Are they: completely very alike I'nlnot different very completely alike much sure different different alike 8. 9. 10. 12. {1'3 . ‘ 14. 114 What do you think about Police Officers and Lawyers? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about Nurses and Private Detectives? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about yourself and Paramedics? Are you completely very alike I'ntnot. different very completely alike much sure different. different alike What do you think about Police Officers and Nurses? Are they: cozrpletely very alike I'm not different very completely alike much sure different different alike What do you think about Private Detectives and Paramedics? Are they: completely very alike I'm not. different. very completely alike Hamil sure different different alike What do you think about yourself and Truck Drivers? Are you: completely very alike I'm not. different very completely alike much sure different different alike What do you think about Lawyers and Truck Drivers? .Are they: completely very alike I'm not different 'very completely alike much sure different. different alike . is: . 1'6 . l7. 18. 19. 20. 21. 115 What do you think about Doctors and Nurses? Are they: completely very alike I'm.not. different ‘very completely alike much sure different different alike What do you think about Nurses and Lawyers? Are they: completely very alike I'm.not different very completely alike much sure different different alike What do you think about Private Detectives and Truck Drivers? Are they: completely very alike I'm.not different very' completely alike ‘much sure different different alike What do you think about Doctors and Paramedics? Are they: completely very alike I'nlnot different very completely alike Hmch sure different different alike What do you think about yourself and Private Detectives? Are you: completely very alike I'm.not different very completely alike :much sure different different alike What do you think about Police Officers and Doctors? Are they: completely very alike I'm not different very completely alike much sure different. different alike What.do you think about Paramedics and Truck Drivers? Are they: completely very alike I'm not different very completely alike much sure different. different alike 22. 23. 25. 26. 27. {/7'\\ 282/ 116 What do you think about Doctors and Lawyers? Are they: completely very alike I'Hant different very completely alike much sure different different alike What.do you think about Police Officers and.Truck Drivers? Are they: completely very alike I'm.not different very completely alike much sure different different alike What do you think about yourself and Nurses? Are you: completely very alike I'm.not. different very completely alike much sure different different alike What.do you think about Nurses and Truck Drivers? Are they: completely very alike I ' m not different very completely alike much sure different different alike What do you think about Private Detectives and Lawyers? .Are they: completely very alike I'm not. different very completely alike mmph sure different. different alike What do you think about Police Officers and Paramedics? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about yourself and Doctors? Are you: completely GEE§" alike I'm.not different very completely alike mmch sure different different alike 29. 30'. 31. 32.; 33. ..;_32i;> 35. 117 What.do you think about Mail Carriers and Bankers? Are they: completely very alike I'm.not different very completely alike much sure different different alike What.do you think about Housewives and Teachers? .Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about Doctors and Mail Carriers? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about Housewives and Secretaries? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about Paramedics and Bankers? Are they: completely very alike I'm.not different very completely alike much sure different different alike What.do you think about Teachers and Janitors? Are they: completely very alike I'm.not. different very completely alike much sure different different alike What do you think about Private Detectives and Bankers? Are they: completely very alike I'm not. different very completely alike much sure different different alike 36. oz; 38,. 39 .7‘ 40. .41“; 42. 118 What do you think about Paramedics and Mail Carriers? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about Teachers and Doctors? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about yourself and Teachers? Are you: completely very alike I 'm not different very completely alike much sure different different alike What do you think about Housewives and Truck Drivers?‘ Are they: completely very alike I 'm not different very completely alike much sure different different alike What do you think about Lawyers and Mail Carriers? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about Housewives and Mechanics? Are they: completely very alike I 'm not different very completely alike much sure different different alike What do you think about Teachers and Factory Workers? Are they: completely very alike I 'm not different very completely alike much sure different different alike 43. 44. 45. 46. 47. 49. 119 What do you think about Nurses and Bankers? Are they: completely very alike I 'm not different very completely alike much sure different different alike What do you think about Housewives and Private Detectives? Are they: completely very alike I'm not different very completely alike mmich sure different different alike What do you think about Truck Drivers and Mail Carriers? Are they: corpletely very alike I'm not different very corpletely alike much sure different different alike What do you think about Teachers and Private Detectives? Are they: completely very alike I'm not different very corpletely alike much sure different different alike What do you think about Teachers and Mail Carriers? Are they: completely very alike I 'm not different very completely alike mmch sure different different alike What do you think about yourself and Housewives? Are you: completely very alike I'm not different very corpletely alike much sure different different alike What do you think about Teachers and Bankers? Are they: completely very alike I'm not different very completely alike much sure different different alike 50. 51. 52?. 54. 55. 56 I.) Whatcbyou 120 completely alike What do you completely alike What do you completely alike What do you completely alike What do you completely alike What do you completely alike What do you corpletely alike think about Housewives and Janitors? Are they: very alike I 'm not different very completely much sure different different alike think about Police Officers and Mail Carriers? Are they: very alike I'm not different very completely much sure different different alike think about Housewives and Doctors? Are they: very alike I 'm not different very completely much ' sure different different alike think about Teachers and Nurses? Are they: very alike I 'm not different very cotpletely much sure different different alike think about Teachers and Truck Drivers? Are they: very alike I'm not different very completely much sure different different alike think about Police Officers and Teachers? Are they: very alike I 'm not different very completely much sure different different alike think about Housewives and Bankers? Are they: very alike I'm not different very coipletely much sure different different alike 57. 58. 59. 60. 617- 62. 63. 121 What do you think about Teachers and Paramedics? Are they: coipletely very alike I 'm not different very completely alike much sure different different alike What do you think about Housewives and Factory Workers? Are they: completely very alike I ' m not different very completely alike much sure different different alike What do you think about Teachers and lawyers? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about Housewives and Paramedics? Are they: completely very alike I 'm not different very corpletely alike much sure different different alike What do you think about Police Officers and Housewives? Are they: completely very alike I 'm not different very completely alike much sure different different alike What do you think about Lawyers and Bankers? Are they: corpletely very alike I 'm not different very coIpletely alike much sure different different alike What do you think about Teachers and Mechanics? Are they: completely very alike I 'm not different very completely alike much sure different different alike .64; 65. 66. 67. 68. 69. 70. What do you 122 completely alike 'What do you completely alike ‘What do you completely alike What do you completely alike What do you completely alike What do you completely alike What do you completely alike think about Housewives and Nurses? Are they: very alike I'm.not. different 'very completely much sure different different alike think about yourself and Bankers? Are you: very alike I'm not. different very completely much sure different different alike think about Private Detectives and Mail Carriers? Are they: very alike I'm.not. different ‘very completely much sure different different alike think about Doctors and Bankers? Are they: very alike IWm1not different very completely much sure different different alike think about Truck Drivers and Bankers? Are they: very alike I'mlnot different very completely much sure different. different alike think about Housewives and Mail Carriers? Are they: very alike Itm.not different very completely much sure different. different alike think about Teachers and Secretaries? Are they: very alike I'm not. different very completely .much sure different different alike 71. 73. 74. 75. 76. 77. What do you 123 completely alike What do you cotplete ly alike What do you completely alike What do you cotpletely alike What do you completely alike What do you corpletely alike What do you cmpletely alike think about Police Officers and Bankers? Are they: very alike I'm not different very completely much sure different different alike think about Housewives and lawyers? Are they: very alike I'm not different very completely much sure different different alike think about Nurses and Mail Carriers? Are they: very alike I'm not different very filetely much sure different different alike think about yourself and Mail Carriers? Are you: very alike I'm not different very cotpletely much sure different different alike think about Police Officers and Private Detectives? Are they: very alike I 'm not different very corpletely much sure different different alike think about lawyers and Paramedics? Are they: very alike I'm not different very completely much sure different different alike think about Doctors and Private Detectives? Are they: very alike I 'm not different very completely much sure different different alike 29. 30. 31. 32. 33. 34. 35. 124 What do you think about Janitors and Bankers? Are they: corpletely very alike I'm not different very completely alike rmmji sure different different alike What do you think about Private Detectives and Mechancis? Are they: completely very alike I'm not. different very completely alike much sure different different alike What do you think about Lawyers and Factory WOrkers? Are they: completely very alike I'm not different very completely alike much sure different different alike What do you think about Secretaries and Bankers? Are they: corpletely very alike I 'm not different very corpletely alike much sure different different alike What.do you think about Doctors and Janitors? Are they: completely very alike I'm.not different very completely alike ‘much sure different. different alike What do you think about Lawyers and Secretaries? Are they: completely very alike I'm.not different very completely alike much sure different different alike What.do you think about Nurses and Factory workers? Are they: coipletely very alike I'm not different very cotpletely alike mimii sure different different alike 3'6 . 37. 38. 39. 40. 41. 42. 125 What do you think about Police Officers and Secretaries? Are they: completely very alike I'm not different very corpletely alike much sure different different alike What do you think about Paramedics and Janitors? Are they: corpletely very alike I'm not different very completely alike much sure different different alike What do you think about Lawyers and Mechanics? Are they: corpletely very alike I'm not different very completely alike much sure different different alike What do you think about yourself and Secretaries? Are you: coIpletely very alike I'm not different very cotpletely alike much sure different different alike What do you think about Mechanics and Bankers? Are they: corpletely very alike I 'm not different very completely alike much sure different different alike What do you think about Truck Drivers and Secretaries? Are they: corpletely very alike I 'm not different very cotpletely alike much sure different different alike What do you think about Secretaries and Mail Carriers? Are they: corpletely very alike I 'm not different very completely alike much sure different different alike 43. 44. 45. 46. 47. 48. 49. What do you 126 completely alike What do you corpletely alike What do you crmpdetely alike What do you completely alike What do you completely alike What do you completely alike What do you completely alike think about Truck Drivers and Factory Workers? Are they: very alike I'mlnot different very completely much sure different different alike think about Police Officers and Mechanics? Are they: very alike I'mlnot different very completely much sure different. different alike think about yourself and Janitors? Are you: very alike I'm.not. different ‘very completely much sure different different alike think about Janitors and Mail Carriers? Are they: very alike I'm not. different very completely much sure different different alike think about Doctors and Factory Wbrkers? Are they: very alike I'm not different ‘very completely much sure different different alike think about Paramedics and Secretaries? Are they: very alike I'm.not different very completely much sure different. different alike think about Mechancis and Mail Carriers? Are they: very alike I'm not different very completely much sure different different alike 50. 51. 52. 53. 54. 55. 56. 127 What do you think about Secretaries and Janitors? Are they: completely very alike I'm.not different very completely alike much sure different different alike What do you think about Mail Carriers and Factory WCrkers? Are they: completely very alike I'm.not. different very completely alike much sure different different alike What do you think about yourself and Mechanics? Are you: completely very alike I'm not. different “very completely alike much sure different different alike What do you think about Secretaries and Mechanics? Are they: completely very alike I'm.not. different. very completely alike much sure different different alike What do you think about Truck Drivers and Janitors? Are they: completely very alike I'm.not. different very completely alike much sure different different alike What do you think about Nurses and Janitors? Are they: completely very alike I'm.not. different ‘very completely alike much. sure different different alike What do you think about Ruck Drivers and Mechanics? Are they: completely very alike I'm not different very completely alike much sure different different alike 57. 58. 59. 60. 61. 62. 63. 128 What do you think about Doctors and Secretaries? Are the: cotpletely very alike I 'm not different very completely alike much sure different different alike What do you think about yourself and Factory Workers? Are you: corpletely very alike I 'm not different very completely alike much sure different different alike What do you think about Nurses and Secretaries? Are they: coipletely very alike I'm not different very corpletely alike much sure different different alike What do you think about Janitors and Mechanics? Are they: corpletely very alike I'm not different very completely alike much sure different different alike What do you think about Bankers and Factory Workers? Are they: corpletely very alike I'm not different very completely alike much sure different different alike What do you think about Doctors and Mechancis? Are they: completely very alike I 'm not different very completely alike much sure different different alike What do you think about Private Detectives and Janitors? Are they: completely very alike I 'm not different very completely alike much sure different different alike 64. 65. 66. 67. 68. 69. 70. 129 What do you think about Paramedics and Factory Workers? Are they: corpletely very alike I 'm not different very completely alike much sure different different alike What do you think about Private Detectives and Secretaries? Are they: corpletely very alike I 'm not different very corpletely alike much sure different different alike What do you think about Secretaries and Factory Workers? Are they: corpletely very alike I'm not different very cotpletely alike much sure different different alike What do you think about Paramedics and Mechanics? Are they: corpletely very alike I 'm not different very completely alike much sure different different alike What do you think about Police Officers and Factory Workers? Are they: coipletely very alike I 'm not different very corpletely alike much sure different different alike What do you think about Nurses and Mechanics? Are they: co'rpletely very alike I'm not different very corpletely alike much sure different different alike What do you think about Private Detectives and Factory Workers? Are they: corpletely very alike I'm not different very coIpletely alike much sure different different alike 71. 72. 73. 74. 75. 76. 77. What do you 130 completely alike What do you complete 1y alike What do you corpletely alike What do you completely alike What do you cotpletely alike What do you corpletely alike What do you completely alike think about Mechanics and Factory Workers? Are they: very alike I'm not different very completely much sure different different alike think about lawyers and Janitors? Are they: very alike I 'm not different very completely much sure different different alike think about Police Officers and Janitors? Are they: very alike I'm not different Very corpletely much sure different different alike think about Janitors and Factory Workers? Are they: very alike I'm not different very completely much sure different different alike think about Police Officers and Private Detectives? Are they: very alike I'm not different very completely much sure different different alike think about lawyers and Paramedics? Are they: very alike I 'm not different very completely much sure different different alike think about Doctors and Private Detectives? Are they: very alike I 'm not different very corpletely much sure different different a1 ike Circle your answer 131 1. 10. How often do police officers tell other people what to do? very often I'm not not very not often often sure often at all How nice are the places where police officers work? very nice I'm not not very not nice nice sure nice at all How important are police officers? very important I 'm not not very not important important sure important at all Are police officers always usually men and usually always men men women women women How much do other people like police officers? a lot a little I'm not not very not at sure much all How often do other peOple tell police officers what to do? very often I 'm not not very not often often sure often at all How helpful are police officers to other people? very helpful I'm not not very not helpful helpful sure helpful at all How intelligent are police officers? very intelligent I'm not not very not intelligent intelligent sure intelligent at all How good are police officers? very good I'm not not very not good good sure good at all How rich are police officers? very rich I'm not not very not rich rich sure rich at all Circle your answer 132 1. 10. How often do doctors tell other people what to do? IFm not sure often not very often very often How nice are the places where doctors work? very nice I'mxnot not very nice sure nice How important are doctors? very important I'm not not.very important sure important Are doctors always usually men and usually men men women women How much do other people like doctors? a lot a little I'm not not very sure much How often do other peOple tell doctors what to do? I'm.not sure often not very often very often How helpful are doctors to other people? very helpful I'm not not very helpful sure helpful How intelligent are doctors? very intelligent I'm not not very intelligent sure intelligent How good are doctors? very good I'm not not very good sure good How rich are doctors? very rich I'm not not very rich sure rich not often at all not nice at all not important at all always not at all not often at all not helpful at all not intelligent at all not good at all not rich at all Circle your answer 1 33 l. "(.10. How often do nurses tell other peOple what to do? very often I'm.not often sure not very often How nice are the places where nurses work? very nice I'm not nice sure HOw important are nurses? very important I'm not important sure Are nurses always usually men and men men women How much do other people like nurses? a lot a little I'm not sure not very nice not very important usually not very often How often do other peOple tell nurses what to do? very often I'mlnot often sure not very often How helpful are nurses to other people? very helpful I'm not helpful sure How intelligent are nurses? very intelligent I'm.not intelligent sure How good are nurses? very good I'm.not good sure How rich are nurses? very rich I'm not rich sure not very helpful not very intelligent not very not very rich not often at all not nice at all not important at all always not at all not often at all not helpful at all not intelligent at all not good at all not rich at all Circle your answer 134 1. 10. How often do private detectives tell other people what to do? very often I'mlnot not very not often often sure often at all How nice are the places where private detectives work? very nice I'mlnot not very not nice nice sure nice at all How important.are private detectives? very important I'mlnot not very not important important sure important at all Are private detectives always usually men and usually always :men men women mrmrmi vmrman HOW'mUCh do other peOple like private detectives? a lot a little I'm not not very not at sure much all How often do other people tell private detectives what to do? very often. I'm not not.very not.often often sure often at all How helpful are private detectives to other peOple? very helpful I'mtnot not very not helpful helpful sure helpful at all How intelligent are private detectives? very intelligent I'm.not not very not intelligent intelligent sure intelligent at all HOW’gOOd are private detectives? very good I'm not not very not good good sure good at all How rich are private detectives? very rich I'm.not not very not rich rich sure rich at all Circle your answer 1 35 1. How often do paramedics tell other people what to do? very often I'm not not very not often often sure often at all 2. How nice are the places where paramedics work? very nice I'm not not very not nice nice sure nice at all 3. How important are paramedics? very important I'm not not very not important important sure important at all 4 . Are paramedics always usually men and usually always men men women women women 5. How much do other peOple like paramedics? a lot a little I'm not not very not at sure much all 6. How often do other people tell paramedics what to do? very often I'm not not very not often often sure often at all 7. How helpful are paramedics to other peOple? very helpful I'm not not very not helpful helpful sure helpful at all 8. How intelligent are paramedics? very intelligent I'm not not very not intelligent intelligent sure intelligent at all 9. How good are paramedics? very good I'm not not very not good good sure good at all 10 . How rich are paramedics? very rich I'm not not very not rich rich sure rich at all Circle your answer 136 1. 10- lbw often do truck drivers tell other pe0ple what to do? very often I'm not not very not often often sure often at all How nice are the places where truck drivers work? very nice I'm not not very not nice nice sure nice at all How important are truck drivers? very important I'm not not very not important important sure important at all Are truck drivers always usually men and usually always men men women women women How mmich do other people like truck drivers? a lot a little I'm not not very not at sure much all lbw often do other people tell truck drivers what to do? very often I'm not not very not often often sure often at all How helpful are truck drivers to other peOple? very helpful I'm not not very not helpful helpful sure helpful at all How intelligent are truck drivers? very intelligent I 'm not not very not intelligent intelligent not sure intelligent at all How good are truck drivers? very good I'm not not very not good good sure good at all lbw rich are truck drivers? very rich I 'm not not very not rich rich sure rich at all Circle your answer 137 1. 10. How often do lawyers tell other people what to do? I'mtnot sure very often often not very often How nice are the places where lawyers work? very nice I'm not nice sure How important are lawyers? very important I'm.not important sure Are lawyers always usually men and men men women How much do other peOple like lawyers? I'm.not sure a lot a little not very nice not very important usually not very much How often do other people tell lawyers what to do? I'm.not sure very often often not very often How helpful are lawyers to other people? helpful I'm.not sure very helpful How intelligent are lawyers? very intelligent I'm.not intelligent sure How good are lawyers? very good I'm not good sure How rich are lawyers? very rich I'mtnot rich sure not‘very helpful not very intelligent not very not very rich not often at all not nice at all not.important at all always not at all not often at all not helpful at all not intelligent at all not good at all not rich at all Circle your answer 138 l. 10. How often do housewives tell other people what to do? very often I'm not not very not often often sure often at all lbw nice are the places where housewives work? very nice I'm not not very not nice nice sure nice at all lbw important are housewives? very important I'm not not very not important important sure important at all Are housewives always usually men and usually always men men women women women How much do other people like housewives? a lot a little I'm not not very not at sure much all How often do other people tell housewives what to do? very often I'm not not very not often often sure often at all How helpful are housewives to other people? very helpful I'm not not very not helpful helpful sure helpful at all lbw intelligent are housewives? very intelligent I'm not not very not intelligent intelligent sure intelligent at all How good are housewives? very good I'm not not very not good good sure good at all How rich are housewives? very rich I'm not not very not rich rich sure rich at all Circle your answer 139 1. How often do teachers tell other people what to do? very often I'm not not very not often often sure often at all 2. lbw nice are the places where teachers work? very nice I'm not not very not nice nice sure nice at all 3. lbw important are teachers? very important I 'm not not very not important important sure important at all 4 . Are teachers always usually men and usually always men men women women women 5. lbw much do other peOple like teachers? a lot a little I'm not not very not at sure much all 6. How often do other peOple tell teachers what to do? very often I'm not not very not often often sure often at all 7. How helpful are teachers to other people? very helpful I'm not not very not helpful helpful sure helpful at all 8. lbw intelligent are teachers? very intelligent I'm not not very not intelligent intelligent sure intelligent at all 9. How good are teachers? very good I'm not not very not good good sure good at all 10. How rich are teachers? very rich I'm not not very not rich rich sure rich at all Circle your answer 140 1. 10. How often do bankers tell other people what to do? I'm.not sure very often often not very often How nice are the places where bankers work? very nice I'm.not nice sure How important are bankers? very important I'm not important sure .Are bankers always usually men and men men women How'much do other peOple like bankers? a lot a little IFm not sure How often do other peOple tell bankers very often I'm not often sure not very nice not very important usually not‘very much what to do? not very often How helpful are bankers to other people? I'm.not sure very' helpful helpful How intelligent are bankers? very intelligent I'm.not intelligent sure How good are bankers? very good I'm.not good sure How rich are bankers? very rich I'm not rich sure not very helpful not very intelligent not‘very not very rich not often at all not nice at all not important at all always women not at all not often at all not helpful at all not intelligent at all not good at all not rich at all Circle your answer 141 l. 8. 9. 10. How often do mail carriers tell other peOple what to do? I'm.not sure very often often not very often How nice are the places where mail carriers work? I'HlDOt sure very nice nice How important are mail carriers? very important I'm.not important sure Are mail carriers always usually men and men men women not very nice not very important usually women How much do other people like mail carriers? I'm.not sure a lot a little not very much not often at all not nice at all not important at all always women not at all How often do other people tell mail carriers what to do? very often I'm.not often sure How helpful are mail carriers to other very helpful I'm not helpful sure How intelligent are mail carriers? intelligent I'm not sure very intelligent How good are mail carriers? good I'm not sure very good How rich are mail carriers? I'm not sure very rich rich not very often people? not very helpful not very intelligent not very not very rich not often at all not helpful at all not intelligent at all not good at all not rich at all Circle your answer 142 1. How often do secretaries tell other people what to do? very often I'Hant not very not often often sure often at all 2. How nice are the places where secretaries work? very nice I'mtnot not very not nice nice sure nice at all 3. How important are secretaries? very important I'm.not not very not.important important sure important at all 4. Are secretaries always usually men and usually always men :men women women women 5. How'much.do other people like secretaries? a lot a little I'Hant not very not at sure much all 6. How often do other people tell secretaries what to do? very often I'm not not very not often often sure often at all 7. How helpful are secretaries to other people? very helpful I'm not not very not helpful helpful sure helpful at all 8. How intelligent are secretaries? very intelligent I'm.not not very not intelligent intelligent sure intelligent at all 53. How 900d are secretaries? very good I'm not not very not good good sure good at all 11). How rich are secretaries? very rich I'm.not not very not rich rich sure rich at all Circle your answer 143 l. 1(). How often do janitors tell other peOple what to do? very often I'm not not very often sure often How nice are the places where janitors work? very nice I'm.not not very nice sure nice How important are janitors? very important I'm not not very important sure important Are janitors always ususally men and usually men memx vrmmmi women How'mmch do other people like janitors? a lot a little I'm.not not very sure much How often do other people tell janitors what to do? very often I'm not not very often sure often How helpful are janitors to other people? very helpful I'm.not not very helpful sure helpful How intelligent are janitors? very intelligent I'm not not very intelligent sure intelligent How good are janitors? very good I'm not not very good sure good HOw rich are janitors? very rich I'm not not very rich sure rich not often at all not nice at all not hmportant at all always not at all not often at all not helpful at all not intelligent at all not good at all not rich at all Circle jour answer 144 1. 10. lbw often do mechanics tell other people what to do? very often I'm not not very not often often sure often at all lbw nice are the places where mechanics work? very nice I'm not not very not nice nice sure nice at all How important are mechanics? very important I'm not not very not important important sure important at all Are mechanics always usually men and usually always men men women women women lbw much do other peOple like mechanics? a lot a little I'm not not very not at sure often at all lbw often do other peOple tell mechanics what to do? very often I'm not not very not often often sure often at all How helpful are mechanics to other people? very helpful I'm not not very not helpful helpful sure helpful at all lbw intelligent are mechancis? very intelligent I 'm not not very not intelligent intelligent sure intelligent at all lbw good are mechanics? very good I 'm not not very not good good sure good at all lbw rich are mechanics? very rich I'm not not very not rich rich sure rich at all Circle_your answer 145 1. 2. 10. How often do factory workers tell other people what to do? very often I'm not often sure How nice are the places where factory workers work? very nice IFm not nice sure How important are factory workers? very important. I'm not important sure Are factory workers always usually men and men men women not very often not very nice not very important usually women How'much do other people like factory workers? a lot a little I'm not sure How often do other people tell factory workers what very often I'mlnot often sure How helpful are factory workers to other people? very helpful I 'm not helpful sure How intelligent are factory workers? very intelligent I'manot intelligent not sure How good are factory workers? very good I'mlnot good sure How rich are factory workers? very rich I'mlnot rich sure not‘very much not very often not very helpful not very intelligent not very not very rich not often at all not nice at all not.important at all always not at all to do? not often at all not helpful at all not intelligent at all not good at all not rich at all Circle your answer 146 l. 2. How often do you tell other peOple what to do? very often I'm.not not very often sure often not often at all How nice are the places where you would like to work? very nice I'm not not.very nice sure nice HOW'important are you? very important I'm not not very important sure important How much do other peOple like you? a lot a little I'm.not not.very sure much How often do other people tell you‘what to do? very often I'm not not very often sure often How helpful are you to other people? very helpful I'm.not not very helpful sure helpful How intelligent are you? very intelligent I'm.not not very intelligent sure intelligent How good are you? very good I'Hant not very good sure good How rich are you? very rich I'm.not not very rich sure rich not nice at all not.important at all not at all not often at all not helpful at all not intelligent at all not good at all not rich at all Circle your answer 1. How often do your parents and friends talk about police officers? How often do your parents and friends talk about doctors? How often do your parents and friends talk about nurses? How often do your parents and friends talk about private detectives? ....... How often do your parents and friends talk about paramedics? ....... How often do your parents and friends talk about truck drivers? .......... How often do your parents and friends talk about lawyers? 147 ............... very often very often very often very often very often very often very often often often often often often often often sometimes sometimes sometimes sometimes sometimes sometimes sometimes not very often not very often not very often not very often not very often not very often not very often never never never never never never never Circle your answer 1. How often do you see a police officer? ....... How often do you see a doctor? How often do you see a nurse? How often do you see a private detective? .... How often do you see a paramedic? ............ How often do you see a truck driver? How often do you see a lawyer? ............... very often very often very often very often very often very often very often 148 often often often often often often often sometimes sometimes sometimes sometimes sometimes sometimes srmemimes not very often not very often not.very often not very often not very often not very often not‘very often never never never never never never never 149 Write jour answer on the line beside each question 1. If you saw 25 people, how many would be Police Officers? 2. If you saw 25 people, how many would be Doctors? 3. If you saw 25 peOple, how many would be Nurses? 4. If you saw 25 Police Officers, how many would be women? 5. If you saw 25 Paramedics, how many would be women? 6. If you saw 25 lawyers, how many would be women? 150 Here are sore shows on TV. These programs are on television four (4) times each month. lbw many times do you usually watch these shows each month? Circle your answer four three two one I don' t watch times times times time this show 1. Medical Center ........... 4 3 2 1 0 2. M.A.S.H .................. 4 3 2 l 0 3. Police Woman ......... . . . . 4 3 2 1 O 4. Cannon ................... 4 3 2 l 0 5. Hawaii 5-0 ............... 4 3 2 1 0 6. Rockford Files ........... 4 3 2 l 0 7. Ehergency. . ......... . . . . . 4 3 2 l 0 8. NbVin On ......... . ....... 4 3 2 l 0 9. Kojak .................... 4 3 2 l 0 Here are two shows that were on TV last winter. About how many times each month did you watch these programs? 10. Kate NbShane ............. 4 3 2 l 0 ll. Petrocelli .............. 4 3 2 l 0 Here are two shows that are on everyday. How many times do you watch them each week? 12. Adam 12 .............. 5 4 3 2 l 0 l3. Ironsides ............ 5 4 3 2 l 0 151 What kind of job does your father have? What exactly does he do at work? What kind of job does your mother have? What exactly does she do at work? How many working color TV sets do you have in your How many cars and trucks does your family have? lbw many bedrooms are in your house? Do you have your own 10 speed bicycle? house? Yes APPENDIX B Permission Form Example 152 APPENDIX B Permission Form Example May 19, 1976 Dear Parent: On May 25th and 26th the fourth grade students at Neff Elementary School will be interviewed. The purpose of these interviews is to find out what children think about different jobs and the people that have these jobs. This project is examining how television influences young people's ideas about various occupations. The interviews will be conducted by representatives from Michigan State University. This project is sponsored by the U.S. Department of Health, Education and Welfare as part of a series of studies on children's development and learning. We hope that your child can take part in this study. The children's answers will be treated confidentially. If you do 22E want your child to participate, please complete the form below and return it to the school. Thank you. Eugene Golanda Principal Neff Elementary School Grand Ledge, Michigan I request that my child not participate in the occupations project at school. (Child's Name) (Parent's Signature) Return this form to school with your child by May 24th. APPENDIX C Statement of School Policy Compliance 153 Gary R. Heald has complied fully with all school policies regarding the use of school children as participants in research projects. .2 'I A 1.44/11 0.... r. #5 ~42. Richard Jo s Principa1(, Beagle Middle School Grand Ledge, Michigan ~ wag/4%] /~/f% Eugene Golanda Principal Neff Elementary School Grand Ledge, Michigan ' . $45- 7é Ric ard L. Johnso Principal Greenwood Elementary School Grand Ledge, Michigan BIBLIOGRAPHY BIBLIOGRAPHY Adams, J. Understanding Adolescence. Boston: Allyn and Bacon, Inc., 1973. Amos, W. and Grambs, J. Counseling the Disadvanteged Youth. Englewood Cliffs, New Jersey: Prentice-Hall, 1968. Atkin, C. and Miller, M. Experimental Effects of Television Advertising on Children. Paper presented to the Inter- national Communication Association, Chicago, Illinois, 1975. Ausuabel, D. Theory and Problems of Adolescent Development. New York: Grune and Stratton, 1954. Barker, R. Ecological Psychology. Stanford: Stanford University Press, 1968. Barnett, G. Bilingual Information Processing: The Effects of Semantic Structure. Unpublished Ph.D. Dissertation, Michigan State University, 1976. Blalock, H. Social Statistics. New York: McGraw-Hill, 1972. Bendix, R. and Lipset, S. Class, Status and Power; A Reader in Social Stratification. Glencoe, Illinois: Free Press, 1953. Berlo, D., Lemert, J. and Mertz, R. Dimensions for Evalu- ating the Acceptability of Message Sources. Unpublished manuscript, Department of Communication, Michigan State University, 1966. Beuf, A. Doctor, Lawyer, Household Drudge. Journal of Communication, 1974, 24, 142-145. Bordin, E. A Theory of Vocational Interest as Dynamic Phenomena. Educational and Psychological Measurement, 1943, 3, 49-65. Borow, H. Career DevelOpment in Adolescence. In J. Adams (Ed.) Understandigg Adolescence. Boston: Allyn and Bacon, 1973. 154 155 Brigham, J. Ethnic Stereotypes. Psychological Bulletin, 1971, 76, 15-38. Brodbeck, A. The Mass Media as a Socializing Agency. Paper presented to the American Psychological Assoc- iation, San Francisco, California, 1955. Bruner, J. On Perceptual Readiness. In R. Harper, C. Anderson, C. Christensen and S. Hunka (Eds.) gee Cognitive Processes. Englewood Cliffs, New Jersey: Prentice-Hall, 1974. Bruning, J. and Kintz, B. Computational Handbook of Sta- tistics. Atlanta, Georgia: Scott, Foresman and Co., 1968. Centers, R. An Effective Classroom Demonstration of Stereo- types. Journal of Social Psychology, 1951, 34, 41-46. Centers, R. and Cantril, H. Income Satisfaction and Income Aspiration. Journal of Abnormal and Social Psychology, 1946, 41, 64-69. Collins , W. Learning of Media Content: A Developmental Study. Child Development, 1970, 41, 1133-1142. Cooley, C. Human Nature and the Social Order. New York: Charles Scribner and Sons, 1902. Creason, F. and Schilson, D. Occupational Concerns of Sixth- grade Children. Vocational Guidance Quarterly, 1970, 18, 219-224. Danes, J. Mass Communications‘lmpact Upon the Image: A New View of the Image and a Model for Change. Unpub- lished manuscript, Department of Communication, Michigan State University, 1975. DeFleur, M. Occupational Roles as Portrayed on Television. Public Opinion Quarterly, 1964, 28, 57-74. DeFleur, M. and DeFleur, L. The Relative Contribution of Television as a Learning Source for Children's Occupa- tional Knowledge. American Sociological Review, 1967, 32, 777-789. Dipboye, W. and Anderson, W. Occupational Stereotypes and Manifest Needs of High School Students. Journal of Counseling Psychology, 1961, 8, 296-304. 156 Dominick, J. Children's Viewing of Crime Shows and Atti— tudes on Law Enforcement. Journalism Quarterly, 1974, 51, 5-12. Dornbush, 8., Hastorf, A. Richardson, 8., Muzzy, R. and Vreeland, R. The Perceiver and the Perceived: Their Relative Influence on the Categories of Interpersonal Perception. Journal of Personality and Social Psychology, 1965, 1, 434-440. Downing, M. Heroine of the Daytime Serial. Journal of Communication, 1974, 24, 130-137. Fishman, J. An Examination of the Process and Function of Social Stereotyping. Journal of Social Psychology, 1956, 43, 27-64. Galler, E. Influence of Social Class on Children's Choices of Occupations. Elementary School Journal, 1951, 51, 439-445. Gonyea, G. Dimensions of Job Perceptions. Journal of Counseling Psychology, 1961, 8, 305-313. Goodson, 8. Occupational Information Materials in Selected Elementary and Middle Schools. Vocational Guidance Quarterly, 1968, 17, 128-131. Gordon, D. Self-Conceptions: Configurations of Content. In C. Gordon and K. Gergen (Eds.) The Self in Social Interaction. John Wiley and Sons, 1968. Gunn, B. Children's Conceptions of Occupational Prestige. Personnel and Guidance Journal, 1964, 42, 559-563. Hale, G., Miller, L. and Stevenson, H. Incidental Learning of Film Content: A Developmental Study. Child Development, 1968, 39, 69-77. Haller, A. and Miller, I. The Occupational Aspiration Scale, Theory, Structure and Correlates. East Lansing: Michigan Agricultural Experiment Station, Technical Bulletin 142, 1963. Hastorf, A., Richardson, S. and Dornbush, S. The Problem of Relevance in the Study of Person Perception. In R. Tagiuri and L. Petrullo (Eds.) Person Perception and Interpersonal Behavior. Stanford: Stanford University Press, 1958. Helm, C., Messick, S. and Tucker, L. Peychological Law and Scaling Methods, Princeton, New Jersey: Educational Testing Service, 1959. 157 Himmelweit, H., Oppenheim, A. and Vince, P. Television and the Child. London: Oxford University Press, 1958. Holland, J. Explorations of a Theory of Vocational Choice, Part I: Vocational Images and Choice. Vocational Guidance Quarterly, 1963, 11, 232-239. Hollander, M. and Parker, H. Occupational Stereotypes and Needs: Their Relationship to Vocational Choice. Vocational Guidance Quarterly, 1969, 18, 91-98. Hovland, C. Effects of the Mass Media of Communication. In G. Lindzey (Ed.) Handbook of Social Psychology. Cambridge, Massachusetts: Addison-Wesley, 1954. James, W. Psychology: The Briefer Course. New York: Henry Holt, 1892. Kaniuga, N., Scott, T. and Gade, E. Working Women Portrayed in Evening Programs. Vocational Guidance Quarterly, 1974, 23, 134-137. Katz, D. and Braly, K. Racial Prejudices and Racial Stereo- types. Journal of Abnormal and Social Psychology, 1935, 30, 175-193. Katzman, N. Television Soap Operas: What's Been Going On Anyway. Public Opinion Quarterly, 1972, 35, 200-212. Kawashima, J. Television Viewing and Occupational Knowledge Among Japanese Children. Unpublished Ph.D. Dissertation, Michigan State University, 1971. Klapper, J. The Effects of Mass Communication. New York: Free Press, 1960. Klineberg, O. The Scientific Study of National Stereotypes. International Social Science Bulletin, 1951, 3, 505-515. Krugman, H. and Hartley, E. Passive Learning from Tele- vision. Public Opinion Quarterly, 1970, 34, 184-190. Kuhn, M. Major Trends in Symbolic Interaction Theory in the Past 26 Years. SociolOgical Quarterly, 1964, 5, 61-79. Laswell, H. The Structure and Function of Communication in Society. In L. Bryson (Ed.) The Communciation of Ideas. New York: Harper, 1948. Lathrop, R. The American Educational Structure. In E. Herr (Ed.) Vocational Guidance and Human Development. Boston: Houghton Mifflin, 1974. 158 LaViolette, F. and Silvert, K. A Theory of Stereotypes. Social Forces, 1951, 29, 257-262. Lazarsfeld, P. and Merton, R. Mass Communication, Popular Taste and Organized Social Action. In L. Bryson (Ed.) The Communication of Ideas. New York: Harper and Brothers, 1948. Lemert, J. Two Studies of Status Conferral. Journalism Quarterly, 1966, 43, 25-33, 94. Lemert, J. Status Conferral and Topic Scope. Journal of Communication, 1969, 19, 4-13. Liebert, R., Neal, J. and Davidson, E. The Early Window: Effects of Television on Children and Youth. New York: Pergamon Press, 1973. Lippman, W. Public Opinion. New York: Harcourt-Brace, 1922. Maccoby, E. Why Do Children Watch Television? Public Qpinion Quarterly, 1954, 18, 239-244. Maccoby, E. The Effects of Television on Children. In W. Schramm (Ed.) The Science of Human Communication. New York: Basic Books, 1963. Mahoney, E. The Processual Characteristics of Self—Conception. The Sociological Quarterly, 1973, 14, 517-533. Mead, G. Mind, Self and Society. Chicago: University of Chicago Press, 1934. Miller, M. Television and Sex-Typing in Children: A Review of Theory and Research. Unpublished manuscript, Depart- ment of Communication, Michigan State University, 1976. Miller, M. and Reeves, B. Dramatic TV Content and Children's Sex-Role Stereotypes. Journal of Broadcasting, 1976, 20, 35-49. Money, J. and Ehrhardt, A. Man and Woman, Bey and Girl. Baltimore, Maryland: Johns Hopkins University Press, 1973. Moore, W. Occupational Socialization. In D. Goslin (Ed.) Handbook of Socialization Theory and Research. Chicago: Rand McNally, 1969. Mowrer, 0. Learning Theory and the Symbolic Processes. New York: John Wiley and Sons, 1960. 159 Nie, N., Hull, C., Jenkins, J., Steinbrenner, K., and Bent, D. Statistical Package for the Social Sciences. New York: McGraw-Hill, 1975. Nunnally, J. Psychometric Theory. New York: McGraw Hill, 1967. Olshan, K. The Multidimensional Structure of Person Per- ception in Children. Unpublished Ph.D. Dissertation, Rutgers University, 1971. Peterson, T., Jenson, J. and Rivers, W. The Mass Media and Modern Society. New York: Holt, Rinehart and Winston, 1965. Piaget, J. The Language and the Thought of the Child. New York: Meridian Books, 1955. Rabban, M. Sex Role Identification in Young Children. Genetic Psychology Monographs, 1950, 42, 81—158. Reeves, B. The Dimensional Structure of Children's Percep- tions of Television Characters. Unpublished Ph.D. Dissertation, Michigan State University, 1976. Reiss, A., Duncan, D., Hatt, P. and North, C. Occupetions and Social Status. New York: Free Press, 1961. Roberts, D. The Nature of Communication Effects. In W. Schramm and D. Roberts (Eds.) The Process and Effects of Mass Communication. Urbana: University of Illinois Press, 1971. Roberts, D. and Schramm, W. Children's Learning from the Mass Media. In L. Deighton (Ed.) The EncyclOpedia of Education. New York: Macmillan, 1971. Roe, A., and Siegelman, M. The Origin of Interests. Washington: American Personnel and Guidance Associa- tion, 1964. Rosenberg, M. Occupations and Values. Glencoe, Illinois: Free Press, 1957. Rosenberg, S. and Sedlak, A. Structural Representations of Implicit Personality Theory. In L. Berkowitz (Ed.) Advances in Experimental Social Psychology. New York: Academic, 1972. Schilder, P. Mind: Perception and Thought in Their Construc- tive Aspects. New York: Columbia University Press, 1942. 160 Schramm, W. Men, Messages and Media. New York: Harper and Row, 1973. Schramm, W., Lyle, J. and Parker, E. Television in the Lives of Our Children. Stanford: Stanford University Press, 1961. Seggar, J. and Wheeler, P. World of Work on TV: Ethnic and Sex Representations in TV Drama. Journal of Broad- casting, 1973, 17, 201-214. Seibold, D. A Complex Model of Multidimensional Attitude and Overt Behavior Relationships: The Mediating Effects of Certainty and Locus of Control. Unpublished Ph.D. Dissertation, Michigan State University, 1975. Serota, K. Metric Multidimensional Scaling and Communication: Theory and Implementation. Unpublished M.A. Thesis, Michigan State University, 1974. Sheffield, F. Theoretical Considerations in the Learning of Complex Sequential Tasks from Demonstration and Practice. In A. Lumsdaine (Ed.) Student Response in Programmed Instruction. Washington, D.C.: National Academy of Sciences National Research Council, 1961, 943, 13-32. Shibutani, T. Improvised News. Indianapolis, Indiana: Bobbs-Merrill, 1966. Simmons, D. Children's Rankings of Occupational Prestige. Personnel and Guidance Journal, 1962, 41, 332-336. Smith, D. The Vocational Aspect of Guidance in the Elementary School. Unpublished manuscript, Harrisburg: Pennsylvania Department of Education, 1968. Stewart, L. Relationship of Socioeconomic Status to Children's Occupational Attitudes and Interests. Journal of Genetic Psychology, 1959, 59, 111-136. Stone, L. and Church, J. Childhood and Adolescence. New York: Random House, 1973. Stone, P., Dunphy, D. and Ogilvie, D. The General Inquirer: A Computer Approach to Content Analysis. Cambridge, Massachusetts: M.I.T. Press, 1967. Super, D. A Theory of Vocational Development. The American Psychologist, 1953, 3, 185-190. 161 Super, D., Starishevsky, R., Marlin, N. and Jordoan, J. Career Development: Self-Concept Theory. Princeton, New Jersey: College Entrance Examination Board, 1963. Tatsuoka, M. Multivariate Analysis Techniques for Educa- tion and Psychological Research. New York: Wiley, 1975. Tedesco, N. Patterns in Prime Time. Journal of Communica- tion, 1974, 24, 119-124. Thiel, H. Principles of Econometrics. New York: Wiley, 1971. Thompson, J. Career Development in the Elementary School: Rationale and Implications for Elementary School Counselors. The School Counselor, 1969, 16, 208-209. Torgerson, W. Theory and Methods of Scaling. New York: John Wiley and Sons, 1958. Van Hoose, W. and Leonard, G. Vocational Guidance in the Elementary School. Guidance Journal, 1966, 5, 61-64. Vinacke, W. Stereotyping Among National-Racial Groups in Hawaii: A Study of Ethnocentrism. Journal of Social Psychology, 1949, 30, 265-291. Vygotsky, L. Thought and Language. Cambridge, Massachusetts: M.I.T. Press, 1962. Wakshlag, J. and Korzenny, F. Untitled Children's Viewing Survey (CASTLE project), Department of Communication, Michigan State University, 1976. Walker, K. A Study of Occupational Stereotypes. Journal of Applied Psychology, 1958, 42, 122-124. Weinstein, E. Weights Assigned by Children to Criteria of Prestige. Sociometry, 1956, 19, 127-132. Woelfel, J. "Symbolic Interaction." Unpublished manuscript, Department of Sociology, University of Wisconsin, 1967. Woelfel, J. Significant Others and Their Relationships to Students in a High School Population. Rural Sociology, 1972, 37, 86-97. Woelfel, J. A Theory of Occupational Choice. Unpublished manuscript, Department of Communication, Michigan State University, 1974. 162 Woelfel, J. Communication Across Age Levels. In H. Oyer and E. Oyer (Eds.), Aging and Communication. Baltimore, Maryland: University Park Press, 1976. Yarrow, M. and Campbell, J. Person Perception in Children. Merrill—Palmer Quarterly, 1963, 9, 57—72. "ITilililllillliii“