..._...; . . . . . _ 1.3;. ... u. . - . . .. . . 39...... . ...>e.\ . , . , v . . 4 u'l —. . .c t. I . . 4 . u “tram” . e. a. 1m .(Wtza... . . r. .3». 17:92:03” .W . . ..a/.””1..em..w.”¥ . v . ; J r . . 3/4“». ., u g .3. no 4 ‘ f . . .1 v5 . with: .J” ”Manama? . . . q 13.91; ., w. . . 2...... 7.. . I a . 2 ram»... Bad. a.” (fin. an...» .. a . 2 2.3.04, 7... F .. luau... C. .. ., ....,.“..._.... ...~.....,...M........u.r. b. .45.... 6.3.2 v2... . .u..t It... 33.; .vr‘.._.g f. :23”. . ... z . . .Vriwnfi.’ G .. Er war? a “KIWHWJMW . .002 f at". .r . .1. 3 r .» .... . . we}. 1.3.2.... €34... 1.... ., . z... . t. v I ‘1 n. {9...}... . . Ham . ‘4 9:“. ( vignlt. . ’56,... A d1. 7.! .3“ Vinny fl... 4 .4 r1 :1. ‘ no.3“. I. .r .V... \‘ ... «pX.i <4, . 1" r4 Prawn” 1|. . :13 t. :34... .45 . yr O .. 91.2:sr it: .- 30:3. a... 4:... ,3. 1'. m ‘ ——_..,.r . at.“ L J Jr... 1.... . (13?! at... L. p Nitnnl .. :r “phutirdlbzw ......H\.,|‘ . 2 . ._ p... IL . . x . .. . .. 2n . .5 . . . . .. .3: . {v . ‘ . III—4“- ..|\.!\ .. ‘ . ‘ ,. . .....:2.»m.«mmq§«3£i~ i lllllllllllHHllllllllHllllllIlllllllllllllfllllllllllllllll 293 10381 5597 lijR-fiquy N MIChigan State mversity This is to certify that the thesis entitled INFORMATION CONTROL BEHAVIORS AND THE POLITICAL EFFECTIVENESS OF LOW INCOME URBAN BLACKS presented by JOHN ELLIOTT BOWES, II has been accepted towards fulfillment of the requirements for Ph . D. degree in Communication 5 fl 245/ 3101’ Affluent Date qygfg l/Z/lfl T 04639 ,_.,_.._....._ _..__ l [ .-__, .--._._—_-__—4-# ABSTRACT INFORMATION CONTROL BEHAVIORS AND THE POLITICAL EFFECTIVENESS OF LOW INCOME URBAN BLACKS BY John Elliott Bowes, II The primary research question of this study was to isolate factors which accounted for effective use of political information coming into urban black neighborhoods. It was reasoned that mass media, interpersonal contacts and organizations constituted the primary chan- nels through which such information from the general society entered the black community. Prior evidence indicated that while these channels are effective carriers of political and other specialized information for the middle class individual, they tend to be unprediCtable carriers for the inner city black despite their high level of use. Consequently, we developed independent variables characterizing basic modes of infor— mation gathering and use to see the extent to which they covaried with or were predictive of dependent political effectiveness measures. Variables characterizing information use were (a) the diversity or variety of inputs used, (b) selectivity or the extent to which one focuses on politically relevant information, and (c) mass media credibility. These three variables were hypothesized to relate posi- tively to dependent political effectiveness measures of political activ- ity, knowledge, and efficacy. A fourth independent variable, amount of input, was treated in interaction with the previous three. In these latter instances, it was reasoned that high amount alone would bear a null or negative relationship with dependent measures. However, under conditions of high selectivity, diversity or credibility, the amount of input would be predictive of higher political effectiveness. Hypotheses were tested by multiple regression with a probability sample of low income urban black residents, following data collection by personal interview in Cleveland, Ohio. Briefly, the following results were obtained. Diversity of mass media inputs and organiza— tional membership ties were significantly related in the predicted direction to at least one of the dependent political effectiveness measures. Selectivity demonstrated in news-public affairs use of the mass media and in interpersonal selectivity shown by a high propor- tion of political conversation contacts, was also related as hypothe— sized. Mass media credibility related positively and significantly to at least one of the dependent measures. Finally, the interaction of interpersonal and organizational selectivity with input amount was significantly associated with at least one of the dependent mea- sures. Overall, seven of the ten original hypotheses were partially or wholly confirmed in this investigation. Dependent measures of political knowledge and activity were best predicted by the independent information control variables. Political efficacy (the attitudinal ~component of political effectiveness) evidenced a relationship only with mass media credibility. Contrary to expectation, amount of infor- mation input, in several instances, showed strong positive associations with dependent measures, with the information control variables of selectivity, diversity and credibility held constant. ' INFORMATION CONTROL BEHAVIORS AND THE POLITICAL EFFECTIVENESS 0F LOW INCOME URBAN BLACKS By John Elliott Bowes, II A THESIS Submitted to Michigan State University in partial fulfillment of the requirement for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1971 c Copyright by JOHN ELLIOTT BOWES, II 1972 ACKNOWLEDGEMENTS I am greatly indebted to my academic advisor and dissertation director, Dr. Bradley S. Greenberg, who provided intellectual stimula- tion, continuing encouragement and friendly guidance throughout my doctoral studies. His advice and criticism was of invaluable help in the completion of this study. Also I am grateful to Drs. Hideya Kumata, Paul Hiniker and R. Vincent Farace, who served as members of my guidance committee and who offered valuable suggestions and comments in this undertaking. Special thanks are due to the other faculty members of the Department of Communication for their help and encouragement throughout my doctoral studies. I also wish to express my appreciation to Brenda Dervin, who was my colleague and co-worker in the implementation and initial data analysis of this study. Thanks are also due to the team of capable interviewers whose services we were fortunate to obtain in Cleveland, Ohio, and to a number of Michigan State University undergraduates who served as data coders. Special thanks are due the staff members of the University of North Dakota Computer Center and Academic Programming Office, particu- larly Jarvis Ehart, Alfred Lindem and John Williams, who offered invalu- able assistance in the data processing phase of this study. I also wish to thank my colleagues in the Department of Journalism at the University of North Dakota, especially Keith Stamm and Herbert Strentz, for their encouragement, help and moral support during the preparation ii of this thesis. Particular thanks are due Carolene Lerud for her able editing and typing skills in the preparation of this manuscript. I wish to express deep gratitude to my wife, Barbara, who helped in innumerable ways, ranging from computer programming to constructive nagging. Without her optimism, encouragement, and occasional push, this study would have been a far more difficult undertaking. Finally, thanks are due to my parents who have provided moral and financial support toward my education over the past years. John Elliott Bowes, II CHAPTER I . . . . TABLE OF CONTENTS Research Problem and Hypotheses . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . Theoretic Rationale . . . . . . . . . . . . . . . . . . . . Basic Research Concepts . . . . . . . . . . . . . . Interrelationship of Concepts . . . . . . . . Hypotheses and Rationale . . . . . . . . . . . . . . . . . CHAPTER II . . . Research Methodology Research Design . . . . . . . . . . . . . . . . . . Operationalization of Variables . . . Control Variables . . . . . . . . . . . . . . . . . . . . . Reliability of Measurement . . . . . . . . . . . . . . . . Statistical Analysis . . CHAPTER III . . . Findings and Interpretation . . . Diversity of Information Inputs and Political Effectiveness Selectivity of Information Inputs and Political Effectiveness Mass Media Credibility . . . . . . Information Input Amount and Political Effectiveness CHAPTER IV . . Summary and Conclusions . . Summary of Conceptual Framework . . . . . . . . . . . . . . Summary of the Empirical Study . . . . . . . . . . . . . . Limitations of the Study . Contributions of the Study Suggestions for Future Work . BIBLIOGRAPHY APPENDICES Appendix A: Appendix B: Appendix C: Appendix D: Letter of Introduction . . . . . . . . . . Sample Selection and Not—at-home Replacement Procedures . . . . . . . . . . . . Copy of Complete Questionnaire Troldahl' 8 Occupational Prestige Scale iv . 93 .100 \lr—Ir—I 13 15 27 27 27 37 54 55 57 59 59 61 68 75 77 89 89 92 98 .104 .110 .110 .111 .113 .130 Appendix Appendix Appendix Appendix Appendix Appendix Appendix Zero Order Correlations between Independent and Control Variables for Major Hypotheses Discussed in Chapter III . . . . . . . . . . . Zero Order Correlation Coefficients between Dependent Political Effectiveness Variables and Control Variables . . . . . . . . . . . . Intercorrelation Matrix of Major Independent and Dependent Variables . . . . . . . . . . . . . Summary Analysis of Variance Used in H8 through H10..................... Map of Sample Frames, Blocks Sampled, and Census Tracts Involved . . . . . . . . . . . . Sampling Record by Blocks . . . . . . . . Means, standard deviations, range, mode and unit of measurement for independent and dependent measures employed in this study. These data are also noted for component indices comprising interpersonal contact diversity . . .132 .133 .134 .135 .137 .138 .139 Table Table Table Table Table Table Table Table Table Table Table 10. 11. LIST OF TABLES Characteristics of the sample as finally constituted, and for an average sampled block . . . . . . . . . . . . . Characteristics of sampled households . . . . . . . . . . . Disposition of sampled households . . . . . . . . . . . . . Intercorrelations of three component scales comprising political effectiveness and the summated political effectiveness index . . . . . . . . . . . . . . . . . . . . Zero order product-moment correlations between interpersonal contact diversity and the component and summary measures of political effectiveness . . . . . . . Zero order and controlled product-moment correlation coefficients between diversity of mass media inputs and dependent measures of political effectiveness . . . . . . . Zero order and controlled product-moment correlation coefficients between diversity of organizational membership and dependent measures of political effectiveness . . . . . . . . . . . . . . . . . . . . . Zero order and controlled product-moment correlation coefficients between press and television news content viewing/reading selectivity and dependent measures of political effectiveness . . . . . . . . . . . . . . . . . Zero order and controlled product—moment correlation between selectivity of political organization membership and dependent measures of political effectiveness . . . Zero order and controlled product-moment correlation coefficients between selectivity of interpersonal contacts and dependent measures of political effectiveness . . . . . Zero order and controlled product-moment correlation coefficients between perceived credibility of the mass media (press and television) and dependent measures of political effectiveness . . . . . . . . . . . . vi 32 33 36 42 62 64 66 69 71 74 76 Table 12. Table Table Table 15. Table Appendix E: Appendix F: Appendix G: Appendix H: Appendix J: Appendix K: F ratio of diversity and input amount interaction for dependent political effectiveness measures under condition of (a) no controls and (b) covariable control F ratio of information input amount main effect, controlled for diversity of input, the interaction of diversity and amount, and other covariables F ratio of selectivity and input amount interaction for dependent measures of political effectiveness under conditions of (a) no controls and (b) covariable controls . . . . . . . F ratio of information input amount main effect controlled for selectivity of input, the interaction of selectivity and amount, and other covariables F ratio of mass media credibility and mass media amount for dependent measures of political effectiveness under conditions of (a) no controls and (b) covariable control . . . Zero order correlations between independent and control variables for major hypotheses discussed in Chapter III . . . . . . . Zero order correlation coefficients between dependent and political effectiveness variables and control variables . . Intercorrelation matrix of major independent and dependent variables Summary of analyses of variance used in H8 through H10 Sampling record by blocks Means, standard deviations, range, mode and unit of measurement for independent and dependent measures These data are also noted employed in this study. for component indicies comprising interpersonal contact diversity . . o o o o v o 79 81 82 85 86 .132 .134 .135 .138 .139 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure p... 0 ON \I o 00 LIST OF FIGURES Comparison of hypothetical media input, information control characteristics and efficacy across several research populations . . . . . . . . . . . . . . . . . . Illustrating information control variables with respect to information inputs and the criterion, political efficacy . . . . . . . . . . . . . . . . . . . . . . . . . Elaboration of hypotheses 8, 9 and 10 in terms of information inputs and measures which apply to each . . . Criteria used to define low income black sample frame and source data for selecting census tracts into sample . . . Items included as an index of political efficacy, their source, and the type of data yielded . . . . . . . . . Items included to form an index of political activity, their source and the type of data yielded . . . . . . Items included to form an index of political knowledge, their source and the type of data yielded . . . . . . . Indices used to compile an overall index of interpersonal contact diversity . . . . . . . . . . . . . . . . . . . Questions used as a basis for assessing diversity of mass media exposure . . . . . . . . . . . . . . . . . . . . . . Questions used as a basis for assessing amount of interpersonal contact . . . . . . . . . . . . . . Questions used as a basis for assessing amount of mass media exposure . . . . . . . . . . . . . . . . . . . . . . Questions used as a basis for assessing the relative credibility of television and the general circulation press . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of control variables used, their method of determination and the data yielded . . . . . . . . . . . . Main findings of study: Summary of hypotheses, method used to test each and outcome of the test . . . . . . . . viii 13 29 39 40 42 46 47 . 52 53 55 60 94 Appendix D: Troldahl's Occupational Prestige Scale . . . . . . . . .130 Appendix I: Map of Sample Frames, Blocks Samples, and Census Tracts Involved . . . . . . . . . . . . . . . . . . . . . . . .137 ix CHAPTER I RESEARCH PROBLEM AND HYPOTHESES INTRODUCTION This dissertation is composed of two basic parts: first, a conceptgal framework for analyzing certain patterns of communication behavior and their relationship to political effectiveness among low income minorities, and secondly, an empirical study of the relationship of several communication patterns to the political effectiveness of low income blacks living in Cleveland, Ohio.1 The importance of viable communication between the black community and the majority society and its government is almost self-evident. Dissatisfaction with the ability of government and the political system to respond effectively to the needs of low income minority neighborhoods has been a recurrent theme in the news of recent years. Similarly, this dissatisfaction has also received considerable, if not effective, attention from government and private agencies. The result has been a plethora of programs directed at the black community, often with disappointing results. The desired effects of these programs, of course, have ideally been betterment of living conditions based on increasing the 1In particular, our concern was for low income blacks living in mostly black neighborhoods among people of roughly the same income level and social status. In less precise terms we were interested in black ghetto residents. Cleveland suited this purpose well, because its sizeable black population is largely located in dense, crowded low income neighborhoods. ' 2 individual's ability to control conditions which affect him-—so-called self-sufficiency. Unfortunately, their results, particularly in terms of generating traditional political enthusiasm, have had mixed effects. Riesman & Glazer (1950) found, for example, that attempts to politicize Harlem residents repeatedly failed. Limited education and communication skills undercut self-confidence and social skills which seemed necessary for active political participation. Other investigators (notably Douvan & Walker, 1956; Easton & Dennis, 1965 and 1967; Greenstein, 1965; Siegel, 1965; and Jennings & Numi, 1968) offer similar conclu~ sions. These investigators basically agree that problems of political apathy are components of a self-perpetuating poverty cycle whereby the physical and emotional difficulties of ghetto life are reflected in political ineffectiveness and "passed on" to future generations. Meanwhile, civil rights groups, the poverty worker and, increas- ingly, government have labored to improve conditions. On paper at least these agencies offer a variety of services, e.g., employment counseling, health care. Private organizations also promote civic participation on a neighborhood level or provide legal or health aid. Yet with frequent good reason, and with almost certain frustration of the officials involved, the services often are not effectively offered, are ill used, or are viewed with suspicion and rejected out- right (Clark, 1967; McIssac & Wilkerson, 1965). Prior poverty research offers limited help in explaining the ineffectiveness of political aid or information programs directed at the low income minority community for several reasons: First, and most basic, studies dealing with poverty and communication as covariables are few (see Greenberg & Dervin, 1970). In particular, 3 analysis of personal effectiveness in dealing with the environment together with a consideration of an individual's communication behaviors has received little attention in low income black neighborhoods. Nearly absent is information concerning interpersonal communication among low income minorities, mass media exposure, attitudes toward the media and media credibility. Nor is much known about how ghetto residents seek out advice on matters beyond their experience, either from "out- side" experts or from organizations within the minority community. Secondly, the bulk of poverty research expends considerable energy in describing the cyclical nature of poverty--essentially non- c0ping reactions to bad circumstances which simply perpetuate that state. The problem is that while the cycle itself is adequately demon- strated, relatively little attention is devoted to the variability of behavior within the poverty subgroup responsible for the cycle's maintenance. Obviously, ghetto residents are not equally victims of apathy and a sense of political ineffectiveness. Activism in the inner city has been increasingly visible. However, while the frequent demographic comparisons made between class levels have generally served to illustrate basic divisions in total range of society, they may have aided the stereotyping of each of those divisions. This may be especially true of the poor, since they are usually most remote from the scientific and reporting community. Finally, communication and various types of personal efficacy both as theoretical concepts and as operational variables have been often considered at a comparatively simple level of analysis. Principal reasons for this state, in part, hinge on the comparative difficulty of conducting field research in low income minority areas. Concepts 4 and operations designed for a general middle class sample often seem inappropriate to the ghetto circumstances. Wording and administration of items become more problematic; this often results in simplification to expedite field Operations or contentment with simple demographic correlates of political activity. The price, of course, is less precise theoretic and operational criteria for implementing study design. Perhaps at the root of this confusion, compounded by the research shortcomings just described, is a notion that consumption of information somehow is beneficial to the consumer. McCombs (1968), for example, implies that parallel increases in black media use and political activ- ism are related. Orum (1965) posits a similar relationship between black ogranizational participation (conducive to information gathering) and political activity. Yet in these instances there is little explana- tion as to how this information input promotes political efficacy. Beyond these specific cases, two broad areas of research have supported the information-~efficacy relationship. Demographic analysis of middle and upper class community influen- tials (Lane, 1959; Milbrath, 1965; Berelson, Lazarsfeld & McPhee, 1954) repeatedly show that politically active community influentials are heavy consumers of the mass media and have extensive interpersonal contacts, often with other influential persons. Complementing these data is evidence from studies of ”modernization" and diffusion of innovations largely conducted in less developed countries. These are concerned largely with basic agricultural practices and community participation. Almond & Verba (1963), for example, relate high mass media use to an elevated sense of local political effectiveness in five developing nations. Lerner (1963) hypothesizes that mass media 5 and modernization—-indicative of increasing political awareness-- jointly develop, each reciprocally advancing the other. Unfortunately, the analogy of increased communication and advanc— ing modernization may not be so easily transferred to the U.S. urban ghetto, sharing as it does some of the qualities of a less developed nation. Yet the little available evidence on this relationship gathered among the U.S. poor is contrary to the media—to—modernity hypothesis. Donohew & Singh (1968), for example, find that high television users were the least receptive to poverty program services and showed a greater sense of powerlessness. Case studies of ghetto communication behavior, especially among children, typically find that the high levels of communication have either a neutral or negative effect on one's ability to cope with conditions, political and otherwise (Liebow, 1967; Chilman, 1966; Minuchin, 1967). Possible reasons for this reversal are not beyond speculation. Middle and upper class influentials, the base of many demographic analyses, usually have an educational and socialization history far more conducive to political development than is available to the inner city resident. Innovative farmers or progressive peasants in less developed countries, in the aggregate, reflect the attainment of educa— tion and a sense of political effectiveness as they modernize. Both situations differ from the U.S. urban ghetto. The U.S. middle class influential and the ghetto resident share the same basic media system, but probably differ greatly in their ability to make effective use of it. -Also, the middle class subject is privy to well educated peers, often experts and specialists far removed from the daily interpersonal world of the ghetto black. The peasant in less developed lands, while 6 not possessing the educational and political sophistication of the U.S. middle class (or, perhaps, for that matter the ghetto resident), may be aided in his modernization efforts by rural extension workers or by specialized media instruction. Such help is frequently unavail- able or unacceptable to the urban black. Too, the base from which the peasant must evidence progress is usually far more rudimentary and simple when compared to the highly specialized, technological society faced by the U.S. urban black. The key to the discrepancy between the effects of information use in the ghetto and its use in the contrasting situations mentioned possibly lies in how individuals sort out and make sense of the informa- tion they are presented with--how they "control" or manage the informa- tion coming to them. The middle class person has his educational and social experience to help him in this process, while the rural peasant often has outside help or has only comparatively easy tasks to cope with. However, the urban black, equipped with poor educational and social skills, faces a complex, often hostile society. A sophisti- cated communication situation surrounds him, seemingly with little positive effect. Figure 1 summarizes this problem in relationship to the several research areas discussed above. Its purpose is to contrast differences made by information control characteristics upon consequent political and personal effectiveness. The balance of this chapter will consider in detail the communication environment of the low income black and the explanatory power of several information control characteristics associated with variations in political effec- tiveness. In particular, our attempt will be to isolate communication variables which improve explanation of ghetto political effectiveness over the present level of generalizations. Information Control Population Information Input Characteristics Efficacy Studies of High levels of media Present due to educa- High U.S. influen- and interpersonal tional and socializa- tials and information input tion history middle class Diffusion or High information Present due to greater High modernization compared to peers education in comparison studies on or baseline to peers, change agent innovative undeveloped society help, special media peasants programs U.S. low High media and Absence of those Low income interpersonal pertinent to middle blacks contact class or modernizing cultures, or unknown Figure 1. Comparison of hypothetical media input, information control characteristics and efficacy across several research popu- lations. THEORETIC RATIONALE The emphasis of this study rests on how individuals gather and interpret information and their success in the environment as a result. A person's ability to obtain needed services from government is deeply involved in this process. The individual who has little information on what he can do to correct abuses or to seek help will not benefit much from the protection and services government offers. Politically effective attitudes and behaviors, to the contrary, depend on some skill in utilizing available information inputs. Political effectiveness and its relation to communication typi- cally is viewed as part of a larger theoretical background concerned with general differences in human effectiveness. A recurrent theme in literature on this subject is that information acquisition and its proper use are basic to effective human functioning. Berlyne 8 (1966), for example, theorizes that information seeking is an instinc- tual response of all successful organisms to their environment. Others (Ascroft, 1969; Cofer & Appley, 1964; Dervin, 1970; Rotter, 1966) discuss information seeking and evluation as fundamental to human success. Information provides the base for effectiveness; it alerts the individual to dangers which must be avoided and to benefits which will improve the quality of life. However, several theorists (Ackoff, 1958; Dervin, 1970) stipulate the need to control or manage information input. Essentially, an individual must be able to discriminate among and evaluate each of the various information inputs he perceives. Awareness of alternative actions available, their outcomes, how the outcome can be reached, and which action is best in which circumstances are all necessary kinds of information which allow the individual to achieve effective control in the environment. The inability to process information and its consequences for political effectiveness is clear in several studies (Douvan & Walker, 1956; Mussen & Wyszynski, 1952; Rotter, 1966). Here it is apparent that the politically ineffective is also one unable to use information effectively and consequently sees the world as oppressive and uncon— trolled. Feelings of political ineffectiveness are part of this larger sense of chaos. Withdrawal, fatalism and inability to associate behav- ior and reward are products of this condition. Deficiencies in education and environment which face the inner city resident offer some explanation for this state. Descriptions of the arid educational climate of the inner city are plentiful. Reisman & Glazier (1950) noted, for example, that low energy levels, 9 limited and ineffective formal education in this setting, result in political apathy. Easton & Dennis (1965) conclude that the low income child has a retarded sense of political effectiveness which becomes worse with age. Minuchin gt a; (1967) found that ghetto children suffer from memory problems, poor time orientation, and limited percep- tual skills--all conditions which increase the difficulty of recogniz- ing and dealing with problems. Douvan & Walker (1956) noted that feelings of incompetence experienced by ghetto residents in one area or activity tended to generalize and become a segment of this incompe— tence these individuals generally felt in most areas of life. The ability to function effectively based on proper information use is not, however, strictly a function of the individual. Simply put, the individual is dependent upon information available from the social system. Others, be they television announcers or neighbors, are components in an information delivery network operating for each individual. Because this system is responsible for presenting behav- ioral alternatives and information concerning their outcomes, its openness to new alternatives is a measure of its success. Closed social systems which restrict the flow of information in a manner analogous to their biological counterparts are unable to adapt to changes or improve existing conditions and eventually fail (Ascroft, 1969; Miller, 1965; Roling, 1970; Watzlawick_e£_al, 1967). The urban black ghetto is by no means a closed information system. However, it is one where only a limited range of inputs are present, and thus shares some of the qualities of the closed system. Basic reasons are not hard to find. Communication in the inner city seems to center upon two 10 disjointed, non-complementary systems. The mass media system serving as the primary conveyor of "outside" information into the ghetto, generally is oriented toward a white middle class consumer society, not the needs and culture of the inner city. The interpersonal network of the low income black, on the other hand, typically, is restricted to a tight circle of family and friends who disregard or disdain infor- mation that is unfamiliar or comes from white dominated sources. This disjointed pattern stifles the effective flow of information between a technical and affluent general society and a ghetto community which lags in this development. Deprived of vital information from society and government, political effectiveness has little basis for development. BASIC RESEARCH CONCEPTS Four basic information control characteristics are hypothesized to be essential to how well the ghetto black overcomes this disjointed situation and its negative consequences for political effectiveness. Essentially, it is reasoned that they intervene in the presentation of information from the mass media, interpersonal, and organizational sources, and the utilization of this information to political advantage. These characteristics comprise the independent variables of this study: I. Diversity: A variety of information inputs is necessary to gather sufficient information to improve control over rewards. In a sense specific to politics, a variety of infor- mation inputs maximizes coverage of political events and minimizes potential bias due to the input of one specific viewpoint. II. III. 11 Selectivity: Even in the relatively isolated climate of the urban ghetto, potential information input can be massive. Television is ubiquitous and supplies an outpouring of infor- mation having little specific political bearing. Interpersonal contacts, also can supply massive input, and if traditional characterizations of ghetto communication are accepted, this communication tends as well to be homogeneous and apolitical. In this situation to effectively utilize the political content available, the individual must sort, accepting some inputs and rejecting the rest. This sorting procedure places strong responsibility on the individual's ability to handle informa- tion. Television and typical interpersonal communication behavior in the ghetto provide few or no cues as to which inputs are important and which are not. Organizations ideally assist the individual to obtain political information, under- stand it, and provide social support for its utilization. Moreover, they can perform this function in a manner consonant with the culture and interests of the inner city black man. However, some kinds of organizations may merely serve a kill- time function, forestalling political action. Credibility: Credibility of information sources is probably of greater concern in the ghetto than in the balance of the population. "Outside” media and interpersonal information sources face suspicion of bias and self-interest from the black community.. However, ”credibility" may imply uncritical acceptance of content leading to fantasy and escape, rather than critical analysis of content and enhanced political 12 effectiveness as a result. IV. Amount: Amount is simply an indication of the quantity of information one has contact with within a period of time. It connotes no effort to be selective or to maximize some sort of goal in attending this information. In the ghetto, "amount" tends to be considerable, equalling and in some cases surpassing the communication activity of middle class individuals. Television consumption, for example, is consid- erable, gregariousness frequently is high within a circle of friends and family, there is evidence of growing organiza- tional membership and press consumption is good. Whether "amount” contributes or detracts from political effectiveness may in good part depend on the presence of the other control characteristics. The dependent variable of this study, political effectiveness, has been operationalized in numerous ways, of which three seem most common. At an attitudinal level, political efficacy concerns the individual's subjective feeling that his actions are effective in making the government consider his needs. High "efficacy” notes that the individual feels himself to be politically proficient, even if in reality he is not (Campbell gt_al, 1954; Robinson, 1968). A second approach gauges the degree of political involvement evidenced by one's overt behaviors. In this sense, the person who, for example, runs for office or gives money to a party, shows greater effectiveness than the individual who merely votes or does nothing (Campbell et a1, 1954; Matthews & Prothro, 1966; Robinson, 1968). A third variant 13 concentrates on one's knowledge 3f the political system. Here individ— uals with high levels of factual recall of such things as names of governors or current events are considered more politically effective (Matthews & Prothro, 1966; Robinson, 1967). Realizing the diverse human attributes subsumed under "effective- ' several researchers have combined these elements into comprehen— ness,l sive indices (for example, Matthews & Prothro, 1966; Litt, 1963). Political effectiveness as conceptualized in this study will include these three components, attitude, behavior and knowledge, which are discussed in greater detail in the next chapter. INTERRELATIONSHIP 0F CONCEPTS Because of their mediating position between sources of informa— tion, and consequent effectiveness, information control characteristics are involved in a number of relationships. These are illustrated in the paradigm below.1 Information Inputs Information Control Criterion A. Interpersonal 1. Diversity Political effectiveness contacts Attitudes (efficacy) Political knowledge B. Mass Media 4. Amount 2. Selectivity Political behaviors C. Organizations 3. Source credibility Figure 2. Illustrating information control variables with respect to information inputs and the criterion, political efficacy. For example, information presented by the mass media, to have any effect on efficacy must of course be attended to. Since this is true 1The paradigm presented here has causal implications. For example, selective media use appears to lead to heightened political effective— ness. However, the "one time” nature of this study and its reliance on correlational technique for hypothesis testing voided causal testing on an empirical level. 14 of the other information inputs, "amount" serves to qualify or mediate the effect of the other three principal control variables; hence, its position in the paradigm.1 Given that there is some non-zero amount, diversity, selectivity and credibility then affect the degree of political effectiveness. Though not directly considered in this study, the paradigm also implies that information control variables, besides "amount," may interact with each other to collectively affect political efficacy. For example, a diverse array of interpersonal contacts might only be effective if they are chosen in a selective manner. Otherwise considerable energy may be wasted in contacting a great number of incompetent or uninformed sources. The same factors are likely to affect organizational membership since the ability of the organizations one attends to affect political activity and attitudes hinge partly on how well chosen they are and how they complement one another. Similarly, the credibility of a given information source may be based on careful judgment or on an uncritical belief in what is largely fantasy content. Thus, one of the major points of the paradigm is to show that political effectiveness is affected by a complex of infor- mation control variables.2 1To an extent, too, amount is related to diversity and selectivity in that if there is zero amount, then obviously there is no information input upon which information control characteristics such as diversity can be demonstrated. This phenomenon is compensated for in analysis. 2Not all control characteristics need apply to all inputs, how— ever. In particular, given certain operational constraints in the study design, source credibility was impractical to assess for organiza- tional and interpersonal inputs. In the case of organizations, credi— bility had poor operational independence from "amount.” It would be unlikely that one would join organizations whose informational services were not to some extent believed. Thus, the totality of groups belonged to likely represents those that are credible information 15 The present study, though, limits its inquiry to the primary relationship of the three conceptually simple information control variables and their interactions with a fourth, amount, in terms of their consequences for political effectiveness. Doubtless, other interactions such as those suggested above as well as more complex predictive combinations of variables (such as the combined effect of diversity, selectivity and amount upon political effectiveness) could be examined. The difficulty, however, lies in the lack of a strong theoretical base on which to make more complex hypotheses and the consequent risk of capitalizing on chance data fluctuations in verifying them. Such extensions may be a logical continuation of the present study and are discussed in Chapter IV. HYPOTHESES AND RATIONALE The initial set of hypotheses together with their justifications deal with the relationship of information control variables to political effectiveness. A second, later set concerns the relationship to effec- tiveness of three control variables, diversity, selectivity and credi- bility, in interaction with a fourth, amount. Our purpose in this second instance is to test the traditional notion that more information inputs are conducive to greater levels of political effectiveness against other information control variables which may offer more expla- nation. sources for the individual as well. Much the same argument applies to interpersonal contacts where credibility has poor operational inde- pendence from selectivity. It would be unlikely, for example, for one to seek "expert" advice from a non—credible source. Further, time considerations brought on by the volume of contacts specified by each respondent ruled out collection of interpersonal contact credibility data from a procedural standpoint. Credibility is, however, evaluated in terms of the mass media since these sources seem to be the ghetto's primary means of contact with the white majority society. 16 Diversity and Political Effectiveness Diversity refers to the variety of different interpersonal and organizational information sources one uses in daily life. The emphasis here is not on amount or number of information sources, but rather their ability to reflect varied political opinion and advice. Most inner city evidence shows that diversity is limited. A number of studies characterize the ghetto resident as one having homogeneous, unvaried interpersonal contacts limited to family and close-by friends. While gregariousness is high, conversations tend to be limited in scope and avoid matters of achievement or advance- ment. The literature also suggests that contacts tend to be more casual and spontaneous, less tied to goals and personal gain than in middle class society. Essentially, information stressing change and self-accomplishment has little chance of passage through this tight, homogeneous network resulting in apathy and low achievement (Liebow, 1967; Epstein, 1961; Chilman, 1965). Though none of these studies reflect specifically on the consequences their findings have for political effectiveness, the association they make of homogeneous, non-achievement oriented contacts to an atmosphere of apathy and with— drawal shows relevance for political effectiveness. Little information is available concerning variability in the different media used by ghetto residents. Since television ownership and consumption is nearly universal, it is doubtful that use of this medium, excluding the question of how much it was used, would evidence much variability. However, the findings of several studies suggesting that the broadcast media imparts little useful information to the inner city resident, indicate that use of the press may be more relevant 17 to political effectiveness (Wade & Schramm, 1969; Block, 1970). Press use, too, is not a universal phenomenon and may well evidence variabil- ity across ghetto households. Roughly half of the inner city families in one study read the general circulation press in a regular basis, but thorough reading (entire paper) occurred only in 17 percent of the cases reported (Greenberg & Dervin, 1970). Black press readership, while respectably high shows variation in the several studies measuring this phenomenon. Lyle (1967) found, for example, that over 70 percent of Los Angeles blacks regularly read the major Negro weekly. On the other hand, a sample of Detroit rioters, showed only a 16 percent readership of the major black weekly (Singer, 1967). The black press particularly seems capable of providing political content attuned to the sympathies of the ghetto. In the recent past, comment emphasiz- ing civil rights has increased while mention of whites correspondingly has gone down (Rosen, 1964). Also, perceived need for such publications is reported highest in the inner city (Frazier, 1965; Lyle, 1967). Thus the individual who uses the full range of media inputs, particu— larly other than television, may likely evidence a greater degree of political efficacy as a result. There is little direct evidence, too, that inner city residents belonging to a more diverse array of organizations are essentially more politically effective. However, according to Lane (1959) and Berelson E£.El (1954), group membership generally supports three politi- cal activities. First, in joining, the individual gives support inten- tionally or otherwise to the group's political activities. Secondly, the group provides a channel of communication from the individual to certain elites often used to pass on political messages and attitudes 18 to government authorities. Finally, the organization provides a plat- form and social support for the individuals own political opinions. The intuition from this would be that varied organizational membership would offer more opportunity for political expression and transmission of political feelings to governmental authorities; hence, greater . political effectiveness. Support for this intuition is indirect and tentative. For exam- ple, several studies (Rogers, 1966; Menzel & Katz, 1955) and others find those with extensive cosmopolite, specialized contacts, including organizations, tend to be more innovative in their profession. Parker and others(1968) find a positive relationship between the diversity of information scientists are exposed to and their level of research output. Jain (1970) found weak indication of such a relationship among agricultural extension professionals. Though these studies do not directly confirm the importance of organizational diversity, they seem sufficient justification for testing the proposition, given the faults of generalizing findings from scientists and professionals to the political effectiveness of the ghetto resident. These findings also, of course, lend general support to the viability of interpersonal contact and mass media diversity as information control variables. Essentially, it seems likely that the ghetto resident with a more diverse array of interpersonal and organizational information sources would be better equipped to make useful, coping political decisions. These hypotheses were tested: H1: Interpersonal contact diversity of low income urban blacks is~positively related to their level of political effective- ness. Diversity of mass media used by low income urban blacks is_ppsitively related to their level of political N 19 effectiveness. H : The organizational membership diversity of low income urban blacks is positively related to their level of political effectiveness. Selectivity and Political Effectiveness Selectivity refers to one's ability to choose information from that available which maximizes his political effectiveness. Since considerable information of a diverting, entertaining sort competes for the individual's time and attention, his ability to minimize time spent on these inputs and concentrate on more politically useful ones is a basic criterion of selectivity. In the ghetto, this ability is hard to come by. Dervin (1970) concludes in a review of poverty research that: "...the high use of television (with its homogeneous, non-means oriented content) and the high use of in-ghetto interpersonal ties (also homogeneous and non-achievement oriented) leave the American urban poor in an information void." Several studies (Wade & Schramm, 1969; Block, 1970) suggest that the principal media of the inner city--radio and television-- are not those which clearly impart useful information. Essentially, then, politically useful content in the mass media may well be ”lost" in a distracting outpouring of entertainment content. Under these circumstances, not only is the politically useful content (news, public affairs, editorial) not effectively used, but the time spent displaces other activities which might be of more direct political benefit. This problem might not be as salient were it not for the consider— able time spent by the ghetto resident with television in particular. Ghetto TV consumption is massive, moreso in fact than for any other sector of the population (Greenberg & Dervin, 1969; U.S. Government, 1968). Though there is some indication that discretion is exercised 20 in what is viewed, more often program selection appears haphazard and unplanned (Allen, 1968; Greenberg & Dominick, 1968). In the case of radio and the press there is little information in terms of what content is preferred and selected out. Radio may largely serve as background music. Allen (1968), for example, found "anything" cited as the most preferred radio programming among black respondents. Some preference was found among adult females and children for popular or gospel music and among men for sports. Favored press content seems limited to scanning of the headlines and want-ads. Only 17 percent of low income black respondents in one study indicated that they read the entire paper (Greenberg & Dervin, 1970). This evidence, in view of typically homogeneous, entertainment based radio programming and the presence of and preference for considerable non— editorial content of the press, possibly indicates limited acquisition of political information from these media. The political consequences of selectivity exercised in joining organizations has a more developed body of theory. Based on this, the position taken in this study is that the ghetto resident who belongs to civic action or other need-oriented organizations will evidence higher levels of political efficacy than those who do not. However, membership in other types of organizations directed toward religious or entertainment activity would at best offer little support to politi- cal activity or at worst displace political activity and reduce conse— quent effectiveness. A number of studies offer support. Traditional views of ghetto organizational life classed residents either as ”isolated," feeling no need in their apathy to join groups, or "compensating" by busily 21 forming groups, imitating established groups in the general society but having little political power (Orum, 1965; Axelrod, 1956; Komarovsky, 1946; Frazier, 1965). Mbre recent studies (Orum, 1965; Marvick, 1965) show that blacks are increasingly joining political organizations to serve political ends. Low income blacks in one study showed a rate of 12 percent among a control group of low income whites, an advantage that the author contends came about in the past 20 years (Orum, 1965). Selectivity considered in light of interpersonal contacts centers usually on one's propensity for seeking out expert advice and help. Particularly in a ghetto setting, problems are severe and complex, often needing above average expertise to solve. The network of homog- eneous, low need achievement contacts that exists for the typical inner city resident has little to offer in this regard. Further, a ghetto norm of reciprocal exploitation spurred by financial insecurity makes a friend's advice often suspect. Advice from outside agencies and experts, particularly from the white community, are viewed with suspicion or rejected outright. Many inner city residents believe, with some justification, that welfare workers and allied advice givers merely want their clients to accept middle class values or be content with the status quo. Thus with an underlying feeling of interpersonal mistrust and a disdain for "outside" advice givers, expert advice, no matter how well meaning, is usually shunned in preference to the less expert counsel of friends (Clark, 1967; Liebow, 1967; Henry, 1965). It would be erroneous to characterize selective information seeking by all low income blacks in light of this basically pessimistic 22 assessment. In the main, it represents a modal, stereotypical view. The contradictions present in the discussion, particularly those dealing with organizations, are perhaps indications that not all inner city residents are cut off from beneficial, expert information or organiza— tional channels through which they can voice their political opinions. Moreover, accounts of ghetto political activism in the popular press and increasing self-help efforts indicate a growth of political exper— tise. The following hypotheses are designed to relate this expertise or effectiveness to selective information control behaviors: H4: Selectivipy by low income urban blacks in the use of politi— cally relevant mass media content is positively related to their level of political effectiveness. H5: Selectivity evidenced py low income blacks in their member- ship in politically active opganizations is ppsitively related to the individual's level of political effectiveness. H6: Selectivity evidenced byplow income urban blacks in the use of expert advice or contact with political information sources is positively related to their level of political effectiveness. Credibility of Mass Media Inputs and Political Effectiveness For information to be politically useful, it must have the confi- dence of the consumer. In the black community especially, suspicion of racial bias extending from distortion of news events to omission of black personalities and events in the news seems to dilute the value of media information. A number of studies (Lyle, 1967; Ingram, 1969; CBS, 1968; Geiber, 1960) point out the suspicion blacks hold for the mass media, particularly the general circulation press. Several accounts of black feeling toward the media (Lyle, 1967; Frazier, 1965) also report that the black operated press, while not suspected of racial bias, appears to distort and sensationalize news about the black community. Perhaps as a consequence of these negative feelings, 23 one study (Greenberg & Dervin, 1970) reports that low income blacks far more than whites prefer "other people" rather than the mass media as local news sources. While there is no direct evidence on the effects of low media credibility upon political effectiveness, disbelief of press content cuts off an important source of political information. Therefore we hypothesize: H7: Perceived media credibility is positively related to the level of political effectiveness for urban low income blacks. Information Input Amount and Political Effectiveness Information input amount refers to the quantity of information an individual acquires through various interpersonal, organizational and media channels. It does not take account of whether this informa- tion comes from diversified sources, or whether the sources are care— fully selected or believed. Input amount merely indicates the total information inputs the individual is exposed to in a period of time. Earlier we made extensive reference to the quantity of past research which related information input amounts to various criteria of effectiveness, political and otherwise. The prime complaint ex- pressed was that these analyses left little grounds to explain why this relationship has not been adequately sustained among U.S. low income samples. The hypotheses stated above tested the relevance of several information control variables for political effectiveness. Essentially, our purpose now is to gauge their importance for the amount of information input present in terms of consequent political effectiveness. We are proposing interaction hypotheses which imply that a large amount of information input can only be effective politi- cally in the presence of operative information control characteristics. Therefore: 24 H8: For low income urban blacks, information input amount inter- acts with input diversity to explain significantly more variation in political effectiveness than both conditions alone. H9: For low income urban blacks, information input amount inter— acts with input selectivipy to explain significantly more variation in_political effectiveness than both conditions alone. H10: For low income urban blacks, media information ipput amount interacts with media credibility to explain significantly more variation in political effectiveness than both condi- tions alone. H8 and H9 posit that the relationships hold for media, organizational and interpersonal inputs. They are illustrated in Figure 3 in terms of the several information inputs which apply to each. For example, interpersonal ”amount” will be gauged in terms of ”gregariousness" or number of contacts made during a sample period of time. The two information control characteristics which apply to this situation, diversity and selectivity, will then be tested for their ability to explain, together with amount, variation in political effectiveness. In conclusion, it should be pointed out that a number of the independent variables employed in this study, both conceptually and operationally, bear resemblance to those of other analyses. For exam— ple, diversity as conceptualized in this study partly approximates "cosmopoliteness" measures of Rogers (1966) and others. Measures of organizational membership approach those commonly employed as a predictor of upper and middle class political ”influentials” (Lane, 1959). Selectivity of news consumption, too, has been employed in studies of this genre, particularly in terms of the relative use of one mass medium compared to another (Allen, 1968). Finally, credibil- ity has been conceptualized as a mediator of information or communicator effectiveness in a variety of contexts (Lane, 1969). INPUT 1. family-peer, interpersonal network 2. organizational membership 3. mass media (press, 25 MEASURE OF AMOUNT gregariousness or number of contacts made in a period number of organizations belonged to time spent viewing or reading broadcast) Figure 3. INFORMATION CONTROL CHARACTERISTIC (H8) (H9) (Hg) (H9) <38) (H9) (H10) diversity of contacts or variation in terms of race, occupation, etc. selectivity or use of expert, specialized information sources diversity or number of categorically different organizations belonged to selectivity or membership in political organizations diversity or total kinds of media regularly used (TV, radio, general circulation press, ethnic press) selectivipy or proportion of news—editorial content consumed credibility or perceived freedom of medium from racial bias Elaboration of hypotheses 8, 9 and 10 in terms of information inputs and measures which apply to each. 26 However, the present adaptation of these concepts to inner city communication problems has changed them substantially. Diversity, for example, does not only concern distance of contacts (as "cosmOpo- liteness" in diffusion/modernization studies), but also measures a respondent's ability to surmount the racial and social isolation of the inner city. Organizational membership, while seemingly a clear- cut contributor to political effectiveness in middle class society, has traditionally a much more ambiguous position in the black community. Clearly the question whether organizations generally or particular ones contribute to or negate black political effectiveness is far from answered in past studies. Selectivity of mass media news consump- tion concerns here more the relative importance of selective information use ip_a particular medium (such as proportion of TV viewing devoted to news) rather than the more traditional notion of relative importance to political effectiveness of a particular medium's use (such as high television consumption contrasted to high magazine use). Credibility, finally, has also been changed in that it is framed specifically in terms of the perceived racial bias of a source rather than the tradi- tional view of its relative trustworthiness or expertise. Chapter II will describe in detail these operationalizations and the measurement model used to empirically test the above hypotheses. CHAPTER II RESEARCH METHODOLOGY The present study was conducted as part of a larger field study which examined the communication behaviors of low income blacks, partic- ularly in relationship to their consumer behaviors, political effective— ness and general ability to cope with their surroundings. This chapter describes the research design used, the nature of the research setting, sampling procedures, the Operationalization of the main variables, data collection methods, reliability and validity of measurement, and the statistical analysis used in the study. The Research Desigp The basic design employed was a field survey conducted in densely populated urban low income black neighborhoods. The goal for using this approach was to chart the nature and effectiveness of ongoing communication patterns in their natural setting with minimal interven— tion by the study's instruments or research personnel. This design, too, was a ”one time only” method, meaning that we were not able to establish cause and effect types of relationships between communication patterns and consequent political effectiveness. Thus, our efforts were limited to describing the nature of relationships which might exist in a causal sense. Field designs also suffer from an inability to control extraneous variables. In the present study, we attempted control in two ways: (1) we attempted to select our samples so as to hold constant possible 27 28 confounding variables (e.g., race, locale), and (2) some variables were controlled statistically by partialling out their contribution to variance demonstrated by the hypothesized relationships (e.g., age, income, education, sex). The Research Setting The site for the research study, Cleveland, Ohio, was chosen for several reasons. First, it constituted a sizeable urban area of approximately 800,000 residents, of whom 25 percent were black. Census information showed that the black population was relatively confined to all—black neighborhoods, not evenly distributed throughout the city. Secondly, Cleveland at the time of the field research (8/69) was one of the first major cities to have a black mayor, a factor we considered indicative of significant variation in black political activity. For these reasons, this city seemed favorable to the research goals of this study. The sampling frame was located entirely within the Cleveland standard metropolitan statistical area (SMSA). Several sources of information then were used to define areas of predominantly low income blacks including census tract data compiled by the Welfare Federation of Cleveland and block data from the 1960 U.S. Census of housing. The basic criteria used are listed in Figure 4. The goal in sample frame selection was to locate census tracts which showed the poorest social environment and the highest density of black persons. With only a few exceptions, eligible areas had to have a non—white household density of 75 percent or better. Additional screening information on eligible tracts was obtained from Mack Clemmons & Associates, a market research organization familiar 29 CRITERION DATA SOURCE a. aid to dependent children, rate per 1000 families WFC b. general relief cases, rate per 1000 families WFC c. illigitimate births per 100 live births WFC d. incidence of male juvenile delinquency, rate per 1000 males, age 12-17 years old WFC e. proportion of deteriorated and dilapidated housing Census f. proportion of homes having incomplete or inoperative plumbing Census g. proportion of non—white occupancy Census h. proportion of crowded homes (more than 1.01 person/room) Census i. assessed valuation of owner occupied dwellings Census j. fewest average number of rooms per dwelling Census Figure 4. Criteria used to define low income black sample frame and source of data for selecting census tracts into sample. (WFC = Welfare Federation of Cleveland, Census = U.S. Govern- ment, 1960 census of housing) with Cleveland's black neighborhoods. Immediately prior to field operations, visual checks of the sampled area eliminated some vacant or burned out blocks, commercial or industrial areas, or white, middle income areas not detectable in census data. Sampling An area cluster sampling design was used to locate respondents within the sampling frame. Basically, this method consists of random selection of blocks or primary sampling units (PSU's) within a frame and then systematic selection of households and respondents within households once the blocks have been determined. The point of the procedure is to select enough PSU's to secure adequate representation of the area within the frame, yet not so many that costs rise due to interviewer time spent traveling to dispersed blocks. Initially a total of 40 blocks were selected randomly. From each block, five ”clusters” of three households each were selected systematically, meaning that every nth household from a random 30 starting position constituted the first household in a cluster.1 In terms of cost of interviewing and representativeness of the sample, the compromise between number of blocks and households sampled on each was considered an optimum balance for this survey situation. In considering the size of the sample, an ”ideal" compromise of cost of interviewing versus precision was estimated to be a sample of 350 completed interviews or a precision of i 5 percent in 95 samples out of 100, given an equal probable estimated outcome. Some 600 house- holds were originally selected to provide in addition to the 350 homes to be interviewed, 250 households to serve as replacements. However, an extremely high number of not—at-home respondents, vacant dwellings and presence of some ineligible upper income black respondents soon exhausted the original and replacement addresses of the first sample. This situation made necessary a second sample to provide additional low income black replacements. The supplementary sample was selected from a relatively confined area lying completely within the geographical bounds of the original sample frame. Procedures for selection of blocks, households, and respondents within both frames were identical. In the second frame, a total of 150 households or 10 blocks were listed to provide replacements (the precise disposition of households is shown in Table 3). Thus, across the two samples, a total of 750 households were selected by random methods.2 A map 1n was equal to the number of household units on the block minus 15, divided by 5. See Appendix B for details. 2But disproportionately. The disproportionality was due to the second sample frame, meaning that blocks (and ultimately respon- dents) from some census tracts had better than equal chance of inclu- sion. The bias introduced by this procedure tended to emphasize the very low income blacks in the frame. Consequently, the respondents were probably more homogeneously low income than would have been the case with equal probability or simple random sampling. 31 of the sample frames, blocks selected and the census tracts involved are included in the appendix (see Appendix I). Comparison of basic housing characteristics for the entire Cleveland metropolitan area to the sampled blocks is given in Table 1. Here the relatively deteriorated housing conditions, low assessed valuation, higher incidence of renter occupied housing and density of black residents in the areas sampled is clearly seen. Once an interviewer had a list of 15 households systematically sampled from each selected block, she contacted them in their order of listing. If no one answered at a listed household, the interviewer selected a replacement from the "extra" replacements incorporated into each block, the five extra blocks incorporated into the first sample, or the second supplementary sample.1 If contact was made, non-blacks, minors, and those with severe language problems were screened out. Of the eligible respondents remaining, interviewers were instructed to try to secure interviews with males, if possible. This procedure was used to compensate for the relative abundance of cooperative, at—home females. As a consequence, a comparatively bal— anced (60.7% female, 39.3% male) sex split was maintained. Respondent Characteristics Descriptive statistics of respondent characteristics show (see Table 2) a fairly typical portrait of the inner city resident. High 1These replacement procedures were employed in this order with the interviewer first attempting to secure replacement households from the block originally selected. Call-backs to not—at—home or uncooperative addresses were attempted only when a definite appointment could be made with the householder for a more convenient time. Concern for interviewer safety ruled out evening interviews, a time when call- backs usually are most successful. All interviewing was limited to the normal working day of 9 a.m. to 5 p.m. Table 1 . 32 Characteristics of the sample as finally constituted, the City of Cleveland, and for an average sampled block.** Total Population Total Housing Units Total Deteriorating & Dilapidated Housing Units Total Renter Occupied Housing Total Black Households Total Housing Units with 1.01 persons or more per room (crowding) Value-- Owner Occupied Housing Final Sampled Cleveland Blocks* 790,000 (approx.) 20,810 282,914 (100%) 6,105 (100%) 269,891 (occupied) 50,436 (17.8%) 2,079 (34.1%) 148,668 (55.1%) 4,282 (70.1%) 67,464 (25.0%) 5,102 (84.4%) 27,686 (10.3%) 1,317 (21.6%) $14,300 *includes only those blocks where interviews were **data from the 1960 U.S. Census of Housing. Average per Sample Block 594.57 174.43 59.4 122.34 145.77 37.62 $10,014.00 (70% of city average) completed. Table 2. D \l . 00 *base: Variable Sex Age Family income Sources of income in household** . One parent households Respondent's formal education Adults per household Children per household (18 years old and under) n = 366 33 Characteristics of sampled households. Description of Variables* Males Females 21-30 years 31—40 41-50 51 or more $25 or less/week 26-50 51-100 101-200 201 or more Wages and Salaries, Welfare, Unempl. Comp. Social Security Other Have only one parent (mother or father) range mean standard deviation mode range mean standard deviation mode range mean standard deviation mode **households may have indicated more than one source. 39.3% 60.7 30.3% 26.5 23.5 19.7 5.9% 12.7 32.6 33.3 15.5 75.5% 34.4 11.7 19.2 40.3% 1-16 years 10.02 2.51 12 1-5 adults 2.01 0.84 2 0-11 children 2.18 2.33 O 34 incidence of one-parent families, a relatively high proportion of families receiving some money from unemployment compensation or welfare, and a below average level of formal education mark this situation. The comparatively moderate figures for children and adults per household belies the frequently crowded conditions. The averages in this instance conceal a wide variation in family size across respondents sampled and, of course, take no account of the often substandard living quarters which would crowd even an "average" sized household. A similar conceal— ment occurs for income. While figures here are only slightly below average, this amount of pay represents the sum of family income. Fre- quently, three to four family members' work was required to bring home an "average'' paycheck. Basic demographic data are presented in Table 2. Field Personnel and Interview Procedure Actual interviewing of respondents took place during the last two weeks of July and the first week of August, 1970. Some 14 black women who had prior census and city directory survey work were located, hired and trained as interviewers. These individuals were all residents of the survey site and were familiar with interviewing in low income black neighborhoods, a research situation having many unique pitfalls. Prior to actual survey work, interviewers attended a two day training session to familiarize them with certain sampling procedures and admin- istration of the questionnaire. In practice, questionnaire administra- tion took approximately 55 minutes. Respondents were paid $2 for their time; a procedure which past experience showed necessary in many instances to secure cooperation. Interviewers were instructed not to mention the inducement prior to 35 completion of the interview unless it became necessary as a last measure to gain the respondent's cooperation. Interviewers also gave the respondent letters indicating the organization (Michigan State Univer- sity) directing the study. These were designed to ease respondent suspicions and hostility. A copy of the letter used is contained in Appendix A. Completed questionnaires were checked with the interviewers as they were turned in to resolve illegible coding, missing answers or clarify ambiguously worded responses. A subsample of completed interview schedules were selected for validation. Of the total of 366 completed schedules, portions of 78 of them (21.3%) were selected for this purpose. Interviewer performance was individually checked in this way an average of four to six times depending on the number of days worked by each. Validation consisted basically of a telephone check on (a) whether the designated respondent had been interviewed, including a check on the respondent's age and sex, and (b) a check on the answers of four randomly selected items from the questionnaire drawn each day. On both criteria, the extent of confirmation on the validation check was satisfactory (93.6% agreement in terms of name, sex, etc., and 94.7% agreement within 1 1 point on a response scale in terms of the four randomly chosen items). Considering the nature of the setting and the length of the questionnaire, refusal rate was relatively low. The principal diffi— culty, as indicated in Table 3, was the level of ”no answer” respondents or ”vacant" addresses selected. The exact procedure by which inter— viewers replaced ”no contact” homes is detailed in Appendix B. 36 Table 3. Disposition of sampled households. a. Households included in both sample frames 750 (100%) b. Households screened out due to burnt out blocks, urban renewal, etc. 44 (6%) c. Interviews completed at origianl sample households 170 (22%) d. Interviews completed at replacement households 196 (26%) e. Not—at-home respondent households, vacant households 258 (34%) f. No eligible adult (too old, young, white, deaf, can't speak English) 42 (6%) g. Refusals 40 (6%) Questionnaire Development Questionnaire design evolved over a period of four months prior to the survey. During this period, draft copies were pretested twice on an urban black sample in Lansing, Michigan, using both black and white student interviewers. The goal during development was, of course, to minimize the difficulty and threat of the interview process for the respondent and to modify questions which resulted in ambiguous responses. Careful attention was paid to the necessity of minimizing order effects and loaded wording of items. The comparative lack of difficulty encountered by interviewers during the actual survey was in good part a result of these efforts. The questionnaire also included many items which were not central to this particular study. For the most part, they concerned financial and consumer knowledge in an effort to tap other aSpects of the respond- ent's ability to cope with his environment. A copy of the complete questionnaire is in Appendix C. 37 Operationalization of Variables In the last chapter, political effectiveness was used as a common dependent variable across all of the ten hypotheses being tested in this study. Four basic classes of independent variables were formulated in terms of three basic forms of information input for all but one case, credibility (see Figure 2, page 13). Based then on the hypotheses listed in the previous chapter, the variables of primary interest are: 1. Political effectiveness of interpersonal contacts 2. Diversity of organizational affiliation of media inputs regularly used of interpersonal information contacts 3. Selectivity of organizational membership of exposure to mass media content of interpersonal contact (gregariousness) 4. Amount of organizational membership of mass media exposure 5. Credibility of mass media content Political Effectiveness Political effectiveness is defined as a complex of attitudes and behaviors indicative of belief that one is effective in gaining services from government to answer his needs. In order to assess this concept, three scales were used. Respectively, they concerned: political efficacy, political knowledge, political participation. Political efficacy represents an attitudinal component that reflects the feeling that one is politically effective, even if in reality he is not. Numerous attitudinal measures have been developed in past years to measure phenomena of this kind, ranging from the specific attitudes of ”political efficacy" to the broader implications 38 of "alienation" or ”fatalism." Our problem in assessing the suitabil- ity of these scales for inclusion into the study centered on which scale items would be (a) specific to politics, (b) particularly relevant to low income ghetto respondents, (c) unidimensional with other scale items assessing this attitude. Two basic screening procedures were used. Items drawn from a variety of political and general scales of effectiveness were included on a face validity basis into the pretest questionnaires. Through the pretesting procedure, items which were clearly inappropriate, confusing, or showed no variation were excluded. Following actual execution of the study, the remaining six scales were factor analyzed to check the item's unidimensional structure.1 A clearly "political” factor resulted, and only one scale item was dropped as a result of this procedure. Scale items ultimately used to compute political efficacy, their origin, and the kind of data yielded are given in Figure 5. The items, compared to their original form, were simplified somewhat in their wording to make them more appropriate in a lengthy interview with low income respondents. The second political effectiveness component, political activity, is determined by the number of common political behaviors a respondent evidences. Campbell's (1954) political activity scale was the basis used for construction of the activity index. Several items from the original scale were replaced or modified when pretest results indicated their inappropriateness to the present situation. Items finally used are listed in Figure 6. The purpose, of course, in creation of 1Insufficient subjects in the pretests did not permit factor analysis at that time. Data Yielded Likert.Item: 4 1‘wa 1‘wa 1‘wa HMO) HNWb agree a lot agree a little disagree a little disagree a lot agree a lot agree a little disagree a little disagree a lot agree a lot = agree a little disagree a little disagree a lot agree a lot agree a little disagree a little disagree a lot agree a lot agree a little disagree a little disagree a lot 39 Origin Srole's (1965) Anomie Scale Seeman's (1959) Powerlessness Index Campbell's (1954) Political Efficacy Scale Troldahl & Powell's (1965) Short Form Dogmatism Scale Haer's (1956) Attitudes towards Source of Power Index Item dropped in factor analysis was Q: so complicated people can't really decide what's going on.” Items Used to Obtain Data Q: City officials are not interested in most people's problems. Q: It really doesn't matter whether you vote or not. Q: People in this neighbor- hood don't have much of a chance to say how things should be run. Q: Most of the ideas which get printed aren't worth the paper they're printed on. Q: The police courts are unfair to people in this neighborhood. "Sometimes government seems This item loaded highly on a general "control” factor (after Rotter, 1966), indicative of one's feelings of fatalism. Figure 5. Items included as an index of political efficacy, their source, and the type of data yielded. Data Yielded' Dichotomous Response 0 = No 1 Yes 0 No 1 .Yes 0 No 1 - Yes 0 - No 1 - Yes 0 No 1 — Yes 0 = No 1 - Yes Figure 6. 40 Origin of Item Campbell's (1954) Political Activity Index Campbell's (1954) Political Activity Index Campbell's (1954) Political Activity Index New Replacement New Replacement New Replacement Item Used to Obtain Data Q: Did you vote in the last Presidential election? Q: Have you ever written or gone to see the mayor, governor, or other public officials? Q: Have you ever campaigned for a politician or talked to people to get support for a political candidate? Q: During the Presidential election last November, did you watch the election returns on TV? Q: Have you ever worked with anybody to get the city to improve conditions in your neighborhood? Q: Have you ever participated in marches, sit-ins, or other kinds of demonstrations for civil rights? included to form an index of political activity, their source and the type of data yielded. 41 "original" items was to make the scale more relevant to political activity in the ghetto and more sensitive to minimal political activi- ties (e.g., watching election returns on TV). Suggestions for these items came as a result of pretesting and through discussions with low income black people. The third component of political effectiveness, political knowl— edge, was also based on a pre-existing scale which was modified to suit the circumstances at hand. The changes involved increased emphasis on local political personalities, which seemed appropriate in light of Cleveland's black mayor and congressman from one of the ghetto area districts. As finally constituted, the knowledge index developed into a person identification test. Questions which dealt with knowledge of civil rights actions, racial disturbances, and partisan political stands showed little variation across pretest respondents or, as was more commonly the case, significantly raised the level of respondent antagonism and suspicion toward the interview procedure. The items finally included for the political knowledge index are shown in Figure 7. Several difficulties arise in the use of the above three scales as component measures of political effectiveness. Though in certain respects different, the three scales should show at least moderate and consistent intercorrelations if they are to be considered as homo- geneous components of political effectiveness. As Table 4 indicates, these criteria were not met. Consequently, analyses were performed individually upon the three component scales as well as the combined "summary” index described. Another problem is that the three scales when combined as an Data Yielded Correct Incorrect Correct Incorrect = Correct = Incorrect = Correct = Incorrect = [\D N 42 Origin of Item Matthews & Prothro (1966) Political Information Scale Matthews & Prothro (1966) Political Information Scale New Replacement New Replacement Item Used to Obtain Data Q: Can you tell me who the governor of this state is? Q: To what political party did President Franklin Roosevelt belong? Q: Can you tell me the name of the Congressman in Washington elected by people in your home area? Q: Can you tell me the name of one of your Senators in Washington? Figure 7. Items included to form an index of political knowledge, their source and the type of data yielded. Table 4. Intercorrelations of 3 component scales comprising political effectiveness and the summated political effectiveness index. 1. Political Knowledge 2. Political Activity 3. Political Efficacy 4. Political Effectiveness (l) (2) 1.0 .17 1.0 .18 .04 .69 .62 (r12 of .095 is equivalent to p$.05) (3) (4) 1.0 .63 1.0 43 index of political effectiveness differ in their units of measurement, the range of values they encompass and the number of items in each. Thus, it did not seem appropriate to "add” the scores to arithmetically construct a summary index. Instead, each respondent's total raw scores were determined separately for each of the three indicies by adding item values. Summed values from the political efficacy (attitude) scale were reflected (inverted), since high values indicated on this scale were indicative of less rather than more political effectiveness. These raw scores were then transformed to standardized "z-scores"1 which allowed their addition to form an overall index of political effectiveness. Political effectiveness then represents an equal, balanced weighting of political attitudes, behaviors and knowledge conducive to one's productivity in dealing with government. Measures of Diversity Diversity measures basically determine, for each of three cate— gories of informational input, the number of dissimilar or varied inputs one is exposed to. They are discussed individually below in terms of the kind of information input applicable. a. Diversipy of Interpersonal Contacts: Diversity of interpersonal contacts was assessed from two basic groups of individuals. The questionnaire provided a detailed sampling of up to six contacts which the respondent had made l'yesterday," as well as three ”experts" he might regularly contact for advice and help in (respectively) political matters, financial matters and personal 1The standardized score is determined by: scale raw score total - mean standard deviation of scale raw score total 44 problems. If a respondent was able to specify more than six contacts for the previous day, the interviewer sampled these contacts as equally as possible from morning, afternoon and evening periods. Since these two groups of contacts were listed for different time bases and pur- poses, their initial analysis was handled separately. Diversity of these two groups was considered along three dimen- sions. Deviation in the occupational prestige of contacts from the respondent's own was considered partly indicative of his ability to cross the confines of homogeneous peer and family contacts. For the most part, the deviation measure showed the respondent's contact with those of higher social status, since the occupational prestige of the respondent was typically low. Prestige in each case was measured along a twelve point occupa- tional prestige continuum developed by Troldahl (1964) for field use (see Appendix D for a description of the scale). The difference between the respondent's prestige and a contact's were summated across all contacts sampled, then divided by the number of contacts.1 This proce— dure yielded an average diversity of contact occupational prestige scores for each respondent. Diversity was also assessed for interpersonal contacts in terms of their average distance from the respondent. In a manner parallel to measures of ”cosmopoliteness” employed in many diffusion studies (see Rogers, 1966), it was reasoned that geographically dispersed contacts would be indicative of a respondent's ability to go beyond lIn formula form: Diversity occupational sum occupational prestige differences prestige number of contacts 45 neighborhood and immediate family for personal contact. Thus, the distance of each contact in the "social" and "expert" groups from the respondent was also estimated in city blocks, largely limited to the city of Cleveland and its suburbs. Distances for each contact were summed and divided by the number of contacts yielding an average distance for the "social" and ”expert" groups. Finally, interpersonal contact diversity was assessed in terms of race, considered indicative of, again, an ability to break with homogeneous familym-peer interpersonal networks. Race was recorded in each of the nine possible contacts discussed above. In each case, contact with a black person was assigned a zero and contact with another race (white, predominantly) was assigned a one. Separately for ”social" and "expert" contacts, these values were summed and divided by the number of contacts in each group. Consequently, an index of mean racial diversity for interpersonal contacts was developed. The several indices used to indicate interpersonal contact diver- sity are indicated in Figure 8. Because each dimension of diversity was expressed in different units comprising a different range of values, the six measures of diversity (2 subsamples X 3 dimensions) were con— verted to standardized z-scores and then summed to provide a composite interpersonal contact diversity index for each respondent. b. Diversity of Organizational Membership: Diversity of organizational membership was determined by summing the number of different types of organizations the respondent belonged to according to the categories given below. 1. sports teams or leagues 2. social groups or fraternities like Masons, Eastern Star 46 3. school organizations like the PTA 4. church groups 5. political parties, organizations or street clubs 6. civil rights or community action groups like the NAACP 7. unions or work organizations 8. home owner's or neighborhood groups Data Yielded Description of Item Used to Obtain Data Mean Diversity: Diversity Occupational Prestige: occupational (for each contact) "What kind of job does this prestige for (a) person have and what kind of place does he work social and (b) at?" (from this information, coders assigned expert contacts each contact and the respondent to one of twelve occupational prestige levels). Mean Diversity: Diversity Distance: distance for (a) (for each contact) "About how many miles is it social and (b) from your home to where this person lives?” (if expert contacts the respondent indicated miles, they were converted to blocks in terms of 1 mile = 15 blocks). Mean Diversity: Diversity Race: race for (a) social (for each contact) "Is this person (contact) and (b) expert black, Mexican, white, or what?" contacts Figure 8. Indices used to compile an overall index of interpersonal contact diversity. c. Diversity of Mass Media Exposure: Diversity of mass media exposure centered on the number of differ— ent media a respondent had available. Basically, this measure consisted of estimating the respondent's access to the hardware of electronic media, to the press and to black publications. The specific items used are indicated in Figure 9. An index of media diversity was formed by summing the affirmative responses ("1") to each item. This total constituted the index for each respondent. 47 Data Yielded Description of Item Used to Obtain Data can't watch, no working sets = O Q: "What channels are you able to watch on your working sets = 1 television set?" never listen = 0 Q: "Now let me ask you about radio. On an ordinary listen = 1 day, about how many hours do you listen to the radio?" never read = O Q: "Some people read a newspaper every day; others some reading = 1 don't because they get their news from radio or TV. About how often do you read a newspaper?" (Question applied to general circulation daily newspapers.) never read = 0 (Same question as above, but data here applies some reading = 1 to general circulation weekly newspapers.) never read = 0 Q: "Do you read any newspapers or magazines that [—1 are mostly about blacks?" (Question applies to ethnic publications.) some reading = Figure 9. Questions used as a basis for assessing diversity of mass media exposure. Measures of Selectivity Selectivity measures basically assess the respondent's ability to select media programs, groups and individuals which will maximize the political information and activity for him. The three basic mea- sures of concern here are discussed below: a. Selectivity of Interpersonal Contacts: Two approaches were used to assess an individual's selectivity in making interpersonal contacts. The first dealt with a general characteristic of seeking out specialized experts or agencies for help as opposed to using probably more familiar, but less expert, contacts like friends, family.and co-workers. The second approach was concerned specifically with political contacts by asking the respon- dent to estimate the number of people he talked to about political and civic matters. 48 As was stated in Chapter I, low contact with and trust in experts was characteristic of inner city blacks and probably contributed to the void of skilled information in this setting. The point of the first approach was to estimate the extent to which each respondent violated this situation. Our basic procedure was to ask for a variety of possible sources whether each was used frequently, infrequently or not at all. The basic format of this question and the sources represented is illustrated on pages 48-49. Information types which are asterisked (**) indicate those ppp considered to be specialized, expert or institutionalized sources. The selectivity index was thus basically a proportion of expert (non- asterisked) source types to the total of sources indicated. However, as indicated in the example, this process automatically weighted source types according to their approximate frequency of use (since greater use of a source would result in a higher value, e.g., a "3" instead of a "2"). Consequently, the index of contact selectivity is a weighted proportion of experts to total contacts used for help or information. "Sometimes when we want help or information about something we need, we go to people we know. Are there any (information source type) around here that you've gone to for help or infor- mation about something you need? Have you ever gone to..... **...neighbors?" (If yes) "How often have you **...friends?" gone to (informa- tion source) **...relatives not living with you?” for help or infor- mation? ...preacher or pastor?" 1For example, the individual who had sought out friends, neighbors and co-workers "a lot” and teachers and lawyers ”a little” would have a total source type sum of 3+3+3+2+2 = 13. The sum of values for expert contacts here would be 2+2 = 4 and the weighted proportion of expert contacts would be 4/13 or .31. 49 ...teacher?” ...a lot?" = 3 ...a little?" = 2 ...civil rights leader or black leader?” (haven't gone = 0) ...lawyer or legal aid society?" ...doctor?" ...public housing (authority)?" ...social worker, case worker or welfare department worker?" **...someone you've worked with on a job?” ...public health or dental clinic?” The second approach, more exploratory in nature, investigated the amount of political discussion contacts the respondent had as a proportion of the number of conversation contacts he made the previous week. We reasoned that this measure would provide some indication of the importance political and civic matters held in talking with others. The question used was: Q: ”In the past week, about how many people did you talk with about politics, neighborhood problems, civil rights...things like that?" The total from this item was divided by the amount of interpersonal contacts the respondent indicated ”last week" (the items used to esti— mate "amount" are shown in Figure 10). The resulting figure is a proportion estimate of political conversation contacts. b. Selectivity of Organizational Membership: Selectivity of organizational membership is defined as the propor- tion of political organizations belonged to based on total organiza- tional membership. To form this index, the roster of organizational types discussed previously with diversity of organizations was used as a basis for estimation. Categories of organizations considered politically relevant were (a) school organizations (like the PTA), 50 (b) political parties, (c) street clubs, (d) civil rights or community action groups and (e) home owner's or neighborhood groups. For each category of organization and individual belongs to, he was assigned ' otherwise, a "zero." Values were summed for politically a "one,' relevant organizations and again across all organizational types. These components formed the proportion serving as an index of selectiv- ity of organizational membership. c. Selectivity of Mass Media Exposure: Selectivity of mass media exposure is defined as the proportion of media time or content devoted to news-editorial content to which the respondent is exposed. Two media, television and the general circulation (daily) press, were used as a base for this estimate. For television, respondents indicated on a TV log the programs they had watched the previous day. For each respondent, the time devoted to viewing regularly scheduled local and national news and educational programs with political bearing was estimated from the completed viewing log. This figure was then divided by the total time spent viewing during the sample day to yield the proportion of time spent viewing news-editorial content. Much the same method was used to estimate the proportion of news-editorial content regularly consumed in the daily, general circula- tion press. Respondents were asked to list in their own terms sections they read "every time they read a newspaper." Coders later content analyzed each subject's responses, separating out those indicative of news-editorial readership. The following responses were considered to indicate this kind of readership: 1. news in general 51 2. headlines 3. front page 4. local news 5. community page, black community page, black news 6. state news 7. national news 8. international news 9. financial news, stock market, tax news, business news 10. editorials and editorial columnists, (e.g., George Condon, Cleveland editorial columnist) 11. political news and columnists The number of sections from a given respondent's list which matched these categories was divided by the total number of sections he read to yield the proportion of news-editorial readership. Because these two measures of mass media selectivity were uncorre- lated (r12 = -.01), and therefore could more likely represent distinctly different relationships with the dependent variable, they were not "averaged together" to provide a summary index of mass media selectiv— ity. Measures of Amount Amount of information input is defined as the quantity or volume of information an individual is exposed to. It does not, of itself, imply the presence of other information control characteristics. Amount was calculated for the three basic information inputs, which are speci- fied individually below: a. Amount of Interpersonal Contact: Amount of interpersonal contact is operationally defined as the respondent's estimate of the number of individuals he talked to 52 in the past week. This information was gathered across four questions specified in Figure 10. The estimate of amount consisted of the total contacts across these four questions. Data Yielded Description of Item Used to Obtain Data number of people Q: "How many people living on this block or across the street have you talked with in the last week?” number of people Q: "How many close friends do you have that you have talked with in the last week?” number of people Q: "How many relatives who don't live with you do you see or talk to almost every week?" number of people Q: "On the usual working day, about how many people do you talk to on the job more than to just say 'hello'?”* *On the job contacts, since they frequently were the same day-to-day were estimated only for a ”usual day” period. Figure 10. Questions used as a basis for assessing amount of inter- personal contact. b. Amount of Organizational Participation: Amount in this instance consists simply of the sum of individual (not categorial) organizations to which the respondent belongs. These data were gathered from a listing of the respondent's membership ties. c. Amount of Mass Media Input: Amount of mass media input is defined as extent of exposure by the respondent to television, radio, general circulation newspapers and black publications. For each of these media, a separate item was needed to make up a total exposure estimate. They are in Figure 11. Because each item of the index represents a different unit of measure- ment or a different range of values, the raw values were transformed to standardized ”z-scores.” This procedure permitted the results of each item to be added, forming a cumulative index of mass media 53 Data Yielded Description of Items Used to Obtain Data count of different Q: "Do you read any newspapers or magazines that magazines listed are mostly about blacks?" frequency of news- Q: "Some people read a newspaper every day; others paper reading don't because they get their news from radio or TV. About how often do you read a newspaper?" (six categories ranging from "never" to "every day") number of hours Q: ”Did you watch television at all yesterday?" viewing TV (Summation of viewing period durations indicated by respondent on TV log for ”yesterday.") number of hours Q: "Now let me ask you about radio. On an ordinary listening to radio day, about how many hours do you listen to the radio?" Figure 11. Questions used as a basis for assessing amount of mass media exposure. input amount.l A problem with amount does occur in analysis. Essentially, at zero values of amount, there is also a zero value of diversity and selectivity. One simply does not avail himself selectively of diverse information inputs if he has none at all. In terms of the first seven hypotheses which posit zero order correlations between various indicants of diversity and selectivity and dependent measures of political effectiveness, a necessary question becomes to what extent does a zero level of amount affect the relationship hypothesized? Since operationally, we have attempted to keep selectivity and diversity distinct from "amount" measures, this potential confounding had to be controlled. A control was used to statistically hold the dichotomous condition versus some amount constant in computation of the correlation In computing "amount" for use with television or press selectiv— ity, only television or press amount was calculated. The cumulative index was not used. 54 coefficients.l This same difficulty also troubles the analysis of hypotheses 8 and 9, since the two predictive variables involved with each (diver— sity or selectivity and amount) would to some extent be non-independent. An analysis method was used which compensated for (partialled out) the potential correlation between the independent variables.2 These several control procedures are roughly analogous to testing hypotheses involving diversity and selectivity only for those showing a non-zero level of amount. In practice, such controls affect about 5 percent of subjects in hypotheses involving the mass media and inter— personal communication and 23 percent in hypotheses dealing with organi— zations. Credibility of the Mass Media Mass media credibility is operationally defined as a subjective judgment of fairness in television and the general circulation press toward racial issues. Two questions were used (Figure 12). An index of media credibility was formed by summing the two responses. Control Variables Earlier in this study it was pointed out that a prime disadvantage of the field survey is an inability to control extraneous variables which might influence hypothesized relationships under test. In order to minimize this defect, we controlled statistically several potential sources of extraneous influence. The method in which control variables were selected and the hypotheses to which they were applied are outlined 1This procedure involved removing from the variance of diversity or selecting that portion which could be attributed to some versus no amount. This method is described on pages 57-58. 2This method is described more fully on page 58- 55 in Chapter III. Data Yielded Description of Item Used to Obtain Data all of the time = 5 Q: "Some people feel that blacks aren't treated most of the time = 4 fairly on television; others think they are. some of the time = 3 What do you think —— are blacks treated fairly rarely = 2 on TV?” never 1 all of the time = 5 Q: "Some people say that blacks aren't fairly most of the time = 4 treated in the regular daily newspapers; others some of the time = 3 think they are. What do you think —- are blacks rarely = 2 treated fairly in the regular daily newspaper?" never 1 Figure 12. Questions used as a basis for assessing the relative credi— bility of television and the general circulation press. Reliability of Measurement Any study, particularly field research conducted under poorly controlled environmental conditions, must be concerned with the relia- bility of its measures. The validation procedure described earlier in this chapter was one check on the consistency of field interviewers. It was reported then that some 94.7 percent of the items checked by a second interview matched the original within plus or minus one atti- tude scale point. Some 86.6 percent of the items checked exactly. However, a second aspect of reliability involves the coding process once the questionnaires have been completed. Basically, this procedure is necessary to interpret the respondent's replies and assign them to categories which may be addressed numerically for machine data processing. Several generalities can be made regarding this procedure in the present study: 1. Coding of all free responses (open-ended) questions was based on a category scheme developed from the responses from all 56 of the 366 respondents involved in this study. This procedure is in contrast to the more common practice of basing a category or content analystic scheme on only a sample of respondent replies. This more thorough procedure was necessary due to the complexity of the responses made. 2. Coding was completed by a team of 13 coders and two supervi- sors, all with prior experience in working with quantitative data. This process was completed question by question to insure consistency, with each coder working on about 1/13th of the questionnaires at a given time. Training was given prior to the coding of each question. 3. The reliability of coding was checked by randomly selecting a 16 percent sample (n = 60) of coded questionnaires and having them coded a second time by different individuals. If, on a particular item, the intercoder reliability1 dropped below 90 percent, this item on all 366 questionnaires was recoded after coders had received additional training. Follow— ing this, a second reliability check was performed on 11 percent of the questionnaires. 4. The reliability criterion was a "point agreement" or exact code agreement measure for all variables. However, this rule was relaxed for several variables where a good deal lReliability measures were computed with the Percentage Agreement Index: number of times 2 coders agreed 1 number of reliability subsamples 57 of calculation or coder judgment was involved. Usually this "relaxation" meant that agreement within plus or minus one point on a lengthy continuum was acceptable. 5. The reliability of coding for each item used in this analysis was at a time greater than 85 percent intercoder agreement and for the most part better than 90 percent, a relatively high figure. Statistical Analysis Responses on questionnaires were translated by coders into cate- gories which could be numerically addressed by machine. These data were then encoded onto punch cards for computer analysis. All coded and punched data were verified, checked for inconsistency and other common encoding errors. For examination of the ten hypotheses in the present study, the following statistical procedures were used. 1. Hypotheses 1 through 7: a. We computed zero order product moment correlations between the independent variable and the dependent variable in the hypotheses to determine the nature and extent of their rela— tionship. b. In order to control for certain extraneous variables which could affect the hypothesized relationships, we used Wards (1962 and 1963) method for isolating the contribution of a given predictor variable. Briefly, it tests the difference in variance accounted for by a restricted and an incremented multiple regression equation. The difference between the 58 two lies solely in the presence or absence of the independent variable; control variables are common to both. The effect is equivalent to a part correlation which partitions from the independent variable that variation induced by confounding factors. c. Zero order and controlled relationships were tested for their statistical significance from zero at a p .05 or better level.. In instances where controlled and zero order relationships differed, the hypotheses were accepted or rejected on the basis of the controlled condition. 2. Hypotheses 8 through 10: Since these hypotheses posited an interaction relationship between amount and other information control variables, a modification of the method just discussed was used. These modifications, basically the addition of a term to the regression to express interaction and the use of additional restricted regression equations, allowed separa— tion of the necessary components of variance. This method is applicable to continuous independent variables and adjusts for their possible intercorrelation. Further, it allows covariables to be used to adjust for extraneous influences. The method, by analogy, is equivalent to analysis of covariance with "least squares" correction for dispro— portionality (Halldorson, 1969; Cohen, 1968; Kelly, 1969). The results of these analyses and a discussion of their signifi— cance is presented in the next chapter. CHAPTER III FINDINGS AND INTERPRETATION The purpose of this chapter is to present the findings and discuss their importance in light of the ten hypotheses tested. However, there are several necessary preliminary notes to be made before discussing the results. It will be recalled that the component scales of political effectiveness——political knowledge, activity and efficacy-~could not be legitimately combined and examined solely as a summary index of effectiveness. Consequently, analyses were performed separately for each component. Verification of hypothe- ses must then be in terms of each. For example, mass media diversity may show--as it does--a significant correlation with one component, political knowledge (and the summary index of effectiveness), but not with the other two, political activity and efficacy. The hypothesis in question, then, would be substantiated for political knowledge, and the effectiveness index but not for activity and efficacy. In several instances (mass media and interpersonal contact selec— tivity), the independent variable has also been partitioned into compo- nent scales because its components intercorrelated weakly and inconsist- ently. Mass media selectivity, for example, is examined separately for the press and for television. Consequently, in these cases, the hypothesis is operationalized and tested in alternative ways. Finally, secondary analyses were performed, holding constant several control variables which past research has shown to strongly 59 60 relate to both independent and dependent measures used in the present study. Further, other similar control procedures were necessary in certain tests to factor out zero levels of organizational and mass media amount (see page 53). Thus, in some instances, two ”control" conditions are presented together with the basic (zero order) analysis.l Control variables (other than zero mass media and organizational amount) chosen for the present study were age, sex, income and years of formal education.2 Their Operationalization and the data each yielded are shown in Figure 13 below. Data Yielded Description of Item or Method Used to Obtain Data a. years of formal Q: "What was the last grade you finished in grade education school or high school?...Have you had any college? How many years?" (Years totalled across both questions.) b. income Q: "Each week, about how much income does your family get from all sources?" c. age in ten year Q: ”What is your age?" intervals d. respondent's sex (At the conclusion of the interview, the interviewer indicated on the questionnaire the respondent's sex.) Figure 13. Summary of control variables used, their method of determi— nation and the data yielded. Control variables, if they are to be meaningful (and used) in the present analysis, should bear a statistically significant lWhere controls are present only for zero amount, acceptance of an hypothesis is based on these results. When results controlled for other covariables, acceptance of an hypothesis is based on these results since they subsume the controls for zero amount as well. 2Lane (1959) and Robinson (1968) discuss the relationship of these control variables to both information control and political attitude and activity measures. 61 relationship (pS.05) with PEER the independent and dependent variables under test. The correlation of age, sex, education and income with the independent and dependent variables of this study are shown in Appendices E and F. There is also a list of the control variables involved for each hypothesis. Specific findings of this study are presented and discussed under each of the primary independent variables outlined in Chapter I: (A) diversity of information input, (B) selectivity of information inputs, (C) credibility of mass media and (D) the interrelationship of amount of information input with the other three information control variables. Diversity of Information Inputs and Political Effectiveness A. Diversity of Interpersonal Contacts: In Chapter I, we hypothesized that: H1: Interpersonal contact diversity of low income urban blacks is positively related to their level of political effective— ness. To test this hypothesis, we computed a zero order correlation between the diversity of a respondent's contacts and component and summary measures of political effectiveness. As Table 5 shows, the correlations do not differ from zero at significance of p = .05 or better for any of the components of political effectiveness. No control variables correlated significantly with both the independent and dependent vari— ables in this test. Consequently, they were excluded. There are few comparable studies with which to contrast these null findings. Studies of organizational communication patterns have found relationships between an individual's productivity or peer evalu- ated job competence and his exposure to diverse audiences and peers 62 Table 5. Zero order product—moment correlations between interpersonal contact diversity and the component and summary measures of political effectiveness. Dependent ' Correlation Variable Coefficient (r) Political Knowledge .0306 Political Activity .0355 Political Efficacy .0294 Political Effectiveness .0491 (see Jain, 1970; Hawkins, 1964). But diverse contacts in this setting seem to be quite job or task specific (e.g., related largely to the variety or range of professional colleagues contacted). The dependent measures, too, (competence) are also closely job related. The inter- personal diversity measures employed in the present study, on the other hand, are not restricted to "political" contacts, but assess diversity across all individuals a respondent sees in a specified period. Perhaps this operational difference is one cause for the dissimilarity in findings. The present results also contrast with the findings of "cosmopo— liteness" measures commonly used in diffusion-modernization studies (cf. Rogers, 1966, 1968; Keith 35 El, 1968), a meaningful difference in that the present measure of interpersonal diversity is similar in operational terms to many cosmopoliteness measures. Cosmopoliteness studies, typically done in less developed lands, usually find a pattern of increasingly dispersed contacts related to a personal syndrome of modernization (which includes, often, increasing political effective- ness). But some sharp differences between the present study and diffu— sion research may account for the disparity of findings. For example, 63 it is likely that the change in personal attitudes and behavior under— lying the modernizing peasant's increased diversity of contacts is far greater than that which motivates the U.S. ghetto resident to visit another part of the city, or to converse with those of a different race or social status. The U.S. urban black ghetto is technologically and economically dependent on contact with the larger community, if not psychologically and emotionally. Residents depend on it for jobs and commodities. The traditional rural peasant village, however, is usually a self—sufficient unit distant from more affluent or simply different people. There is usually little pressure on a day-to-day basis of basic survival to extend beyond its limits. Thus, the inter— personal diversity measure employed in the present setting may well have been too insensitive to the subtle distinction between necessary or forced contact by ghetto residents with the larger community and willing contact. Consequently, the first hypothesis, relating interpersonal contact diversity to political effectivenessl is not confirmed. B. Diversity of Mass Media Use: In Chapter I, we hypothesized that: H2: Diversity of mass media used by low income urban blacks is positively related to their level of political effective- ness. In order to test this hypothesis, we computed zero order product moment correlations between the diversity of the respondent's mass media inputs with dependent measures of political effectiveness. We also computed controlled correlations, adjusting in one case for the correla- tion of diversity and amount at zero levels of amount, and in the second case for other confounding variables specified in Table 6. 64 Table 6. Zero order and controlled product-moment correlation coeffi— cients between diversity of mass media inputs and dependent measures of political effectiveness. Dependent Zero Order Controlled Covariables Variable Correlation for No Amount Controlled (r) Covariables Political Education, Knowledge .2930*** .2722*** .2003*** Amount mass media Political Amount Activity .0193 .0428 .0630 mass media Political Efficacy .0223 .0287 -— None Political Education, Effectiveness .1720** .1766** .1068* Amount mass media p .05 = *; p .01 = **; p .001 = *** As Table 6 shows, one component, political knowledge, and the summary index of political effectiveness were significantly related to mass media diversity. Though lowered somewhat by partialling out the covari— ables of education and mass media amount used, the relationships still were maintained at an acceptable level of statistical significance. Thus, H9 is supported for political knowledge (and the summary effec— tiveness measure) and not sgpported for political activity and efficacy, The lowered significance of the summary measure reflects the averaging of the component indices. These results do have a few parallels in the literature. Lane (1959) in summarizing the effects of mass media on political understand- ing and knowledge, points to the importance of diverse mass media inputs, particularly magazines. Rogers (1968) and others found a strong positive relationship between media exposure (measured largely in terms of different media used) and the political knowledge evidenced 65 by South American peasants. In quite a different context, Parker (1968) found that the channel diversity of information inputs to a researcher positively relate to his research productivity. It should be noted, however, that Jain (1970) failed to substantiate Parker's finding. The restriction of significant findings to political knowledge seems reasonable. Characteristically, politicians and the general circulation media which relay and analyze them rarely take strong partisan stands or openly advocate concerted political action (Lane, 1959). Thus, there seems little content to motivate political action even from a diversified perusal of common media inputs. There is also little evidence to suggest that simply exposure to a more diverse array of mass media improves one's feelings of political efficacy. C. Organizational Membership Diversity: In Chapter I, we hypothesized that: H3: The organizational membership diversity of low income urban blacks is positively related to their level of political effectiveness. To test this hypothesis, we computed zero order correlation coefficients between the diversity of the respondent's organizational ties and measures of political effectiveness. We also computed controlled correlations, in one case to adjust for the correlation of diversity and amount at zero levels of amount, and in the second case for other confounding variables specified in Table 7. Table 7 shows relatively strong correlations, initially, between political knowledge and activ— ity, and one's diversity of organizational membership. However, when particularly amount (and to a far lesser extent, age) were held con— stant, these two relationships were essentially reduced to null. The 66 Table 7. Zero order and controlled product—moment correlation coeffi— cients between diversity of organizational membership and dependent measures of political effectiveness. Dependent Zero Order Controlled Covariables Variable Correlation for No Amount Controlled (r) Covariables Political Age, Amount Knowledge .1496** .1751** .0243 organizational participation Political Age, Amount Activity .4498*** .3338*** .0000 organizational participation Political Efficacy —.0265 -.0095 -— None Political Age, Amount Effectiveness .2949*** .2674*** .1043* organizational participation p .05 = *; p .01 = **; p .001 = *** effect of age as a control was slight since it had only marginal corre- lations with organizational diversity (r12 = .169) and dependent effec— tiveness measures (r12 of .142, .153 and .032 for political knowledge, activity and efficacy respectively). What seems at the base of the problem is amount, reflected in the manner by which people seem to choose organizational ties. To start with, the zero order correlation between organizational diversity and amount is r12 = .97, indicating that the two variables, while conceptually different, are in practice the same. Diversity was assess- ed by the number of categorically different organizations (e.g., church groups vs. street clubs, etc.) the respondent belonged to and organiza- tional amount by the simple total of groups or clubs. The findings show that respondents rarely indicated belonging to more than one organization in a given category. Indeed, a review of the raw data showed that an average of only 1.3% of the respondents indicated 67 belonging to more than one organization per category. Understandably, organizational amount shows a near mirror image of the results for diversity shown in Table 7 (amount had a zero order correlation of r12 = .47 with political activity and an r12 = .16 correlation with political knowledge). In light of this virtual identity of organizational amount and diversity, we must conclude that Hg is supported strongly for political activity and moderately for political knowledgg. The summary index, political effectiveness, reflecting the averaging of its component scales was also significantly related to diversity (or amount). We must also conclude that the distinction between organizational amount and diversity cannot be supported in this study; that in terms of the results, amount is in effect diversity. These findings are in essential agreement with Orum's (1965) contention that political activity (voting behavior in this case) and organizational membership among low income blacks are increasing and related. In summary, Orum states: "If the early experience of other ethnic groups in America can be considered a useful guide, the participation of Negroes in associations represents a significant step toward integration. In this respect, the most important consequence of activity in associations is a kind of civic education. Ideally, if not always practically, voluntary associations are models of coopera- tive effort; decision making follows discussion, debate, and the reaching of a consensus among the members. Participation in associations thus offers Negroes an opportunity to acquire an understanding of the processes of cooperation and compromise that are the foundations of democratic living. "In the area of pure political activity, the increase in voting turnout of Negroes may be due to a greater awareness among Negroes of the effectiveness of organized political efforts....." The present findings and Orum's conclusions contradict traditional "compensatory" and "isolationist" theories of black organizations 68 which allocated these groups politically powerless functions of enter- tainment and replacement for associations denied in the white community (Myrdal, 1944; Frazier, 1965; Orum, 1965). Selectivity of Information Inputs and Political Effectiveness A. Selectivity of Mass Media Use: In Chapter I, we hypothesized that: H4: Selectivity by low income blacks in the use of politically relevant mass media content is positively related to their level of political effectiveness. This relationship was tested separately for television and the general circulation press (major dailies). Due to the low intercorrelation of these two aspects of media selectivity, they were not combined to form a summary index. To test the hypothesis, we computed zero order product moment correlations with the dependent political effec- tiveness variables for both television and press selectivity. These analyses were repeated, controlling for the correlation of selectivity and "amount" at zero levels of amount. The analysis was performed a third time for press selectivity, holding constant amount of press use and respondent education for all dependent measures except political effectiveness. As Table 8 indicates, significant relationships were found between TV selectivity and political knowledge and activity which were lowered when controls were imposed. A slightly higher relationship was found between political knowledge and press selectivity when controlled for zero "amount." By controlling for amount and education, the correlation improved slightly to r .1617. 12' Again, as with media diversity, we find selective media users have a slightly greater level of political knowledge and activity. There is some corroborating evidence from other investigations. Table TV PRESS p .05 69 8. Zero order and controlled product-moment correlation coeffi— cients between press and television news content viewing/ reading selectivity and dependent measures of political = *; p .01 = **; p .001 = *7??? effectiveness. Dependent Zero Order Controlled Covariables Variable Correlation for No Amt. Controlled (r) Covariables Political Knowledge .1465** .1196* -- None Political Activity .1396** .1263* —— None Political Efficacy .0273 .0171 —- None Political Effectiveness .1611** .1175* -— None Political Education, Knowledge .2180*** .1427** .1617** Amount mass media Political Education, Activity -.O475 -.0491 .0811 Amount mass media Political Efficacy .0634 .0466 —- None Political Education, Effectiveness .1203* .0721 .0660 Amount mass media 70 Chaffee's (1970) study of teenage political socialization found causal evidence linking television public affairs content use (as opposed to entertainment content) to political knowledge. This effect was only partly apparent for newspapers since entertainment use was also related, though less strongly, to political knowledge. However, Chaffee's findings did not find consistent parallel findings for polit— ical activity. As he concluded: ".....our data point to the inference that mass communication plays a role in political socialization insofar as political knowledge is concerned, but its influence does not extend to overt behavior such as campaigning activity." Jennings and Numi (1968) see public affairs use of the mass media as part of a developing pattern of political socialization which continues to grow well into adult years. It is interesting in this light that among low income blacks in the present study where such development has been considered invariably low, that variation in public affairs media use is present and appears to contribute to polit— ical knowledge and possibly to political activity. In conclusion, our findings show statistically significant support for H4 in terms of political knowledge for both press and television and for political activity in the case of selective television use. The summary political effectiveness measure also showed a low but statistically significant correlation with selective television use. B. Selectivity of Organizational Membership: In Chapter I, we hypothesized that: H5: Selectivity evidenced by low income blacks by their member- ship in politically active organizations is positively related to the individual's level of political effectiveness. To test the hypothesis, we computed zero order and controlled product moment correlations between the dependent political effectiveness 71 variables and the respondent's proportionate membership in politically active or civic organizations. In computing this relationship a second time, the confounding of organizational selectivity at a zero level of amount of organizations was controlled. A third analysis was per— formed, also holding constant the covariables indicated in Table 9 below. None of the control variables correlated significantly with both political efficacy and organizational selectivity. Consequently, this calculation was omitted. As Table 9 indicates, controlling for Table 9. Zero order and controlled product-moment correlations between selectivity of political organization membership and dependent measures of political effectiveness. Dependent Zero Order Controlled Covariables Variable Correlation for No Amount Controlled (r) Covariables Political Sex, Amount Knowledge .0250 .0141 .000 organizations Political Amount Activity .2531*** .1177* .0793 organizations Political Efficacy .0022 .0274 —- None Political Sex, Amount Effectiveness .1421** .0820 .0137 organizations p .05 = *; p .01 = **; p .001 = *** zero organizational amount and subsequently for the full range of amount and sex as covariables obliterated the significant correlation of organizational selectivity with political activity. The problem here is reminiscent of those which occurred earlier with organizational diversity and the intervening effects of amount. Simply put, it appears that as one joins more organizations (amount) the chance also seems to increase that they will be political or civic action oriented (e.g., selectivity). This supposition is reflected in the rather high 72 correlation between selectivity and organizational amount (r12 = .39). However, since this correlation is far from unity, we can only state that there is pomp tendency for selectivity to increase as amount increases. Consequently, HS cannot be confirmed, even indirectly. This hypothesis is also not confirmed for the other component political effectiveness measures. The tendency for increasing amount of black organizational parti— cipation to also imply selective participation in civic and political groups is partly substantiated by Orum (1965). Comparing low income blacks to low income whites, he found that the black group was far more likely to join political organizations, and about equally likely to join civic groups, as the white controls. Possibly, the failure of organizational selectivity to account for much variation in political effectiveness can be traced to diffi— culty in determining the political activity of a group through categor— ical labels. Church groups in the present study, for example, were consigned to a non-political, non—civic action status. Yet in recent times, some church groups have served as a locus of black political activity and civic improvement. Conversely, street clubs and neighbor— hood groups, considered in the present study to be political or civic action oriented, may in many instances serve a merely social function. In sum, the form of classification used may well have been inappropriate to variation in the political activity of organizations. Finally, organizational participation, regardless of kind of group considered, may be conducive to greater levels of political effectiveness. In this perspective, the group would serve as a social— izing agent, one which would open the respondent to the fundamentals 73 of issue formation and the collective power of group activity. C. Selectivity of Interpersonal Contacts: In Chapter I, we hypothesized that: H6: Selectivity evidenced by low income urban blacks in the use of expert advice or contact with political information sources is positively related to their level of political effectiveness. This relationship was tested in two somewhat divergent ways. First by computing correlation coefficients between the use of expert infor- mation sources relative to total sources used for "help and information" and the dependent measures of political effectiveness, we tested the political consequences of a general pattern of seeking expert help. The second method attempted to test the hypothesis in a specifically political context by estimating the proportion of average weekly con— tacts used for political discussion and information gathering. In neither case was it necessary to control for the "built—in” correlation between zero interpersonal contact amount and (therefore) zero contact selectivity since every respondent reported having some interpersonal contact "in the past week.” However, the relationships were controlled for the covariables shown in Table 10. As can be seen here (Table 10), interpersonal selectivity measures in terms of proportion contact with experts showed no significant relationship to dependent political effectiveness measures. Contrary, those who reported a high proportion of political conversation contacts did evidence higher levels of polit- ical knowledge and particularly activity. Controlling for education however, left only the relationship with political activity viable. Thus, we must conclude that H5 is supported only for political activity as predicted by proportion of political conversation contacts. Politi— cal effectiveness, the summary index, is also significant under the Table 10. 74 Zero order and controlled product—moment correlation coeffi— cients between selectivity of interpersonal contacts and dependent measures of political effectiveness. PROPORTION CONTACT WITH EXPERTS PROPORTION POLITICAL CONVERSATION CONTACTS Dependent Zero Order Covariables Variable Correlation Controlled (r) Covariables Political Education, Knowledge -.0880 -.0368 Age, Sex Political Activity .0054 .0205 Age Political Efficacy —.0206 —.0205 Sex Political Effectiveness -.0533 -.0531 Sex p .05 = *; p .01 = **; p .001 = *** Political Knowledge .1113* .0887 Education Political Activity .2373*** .2290*** Education Political Efficacy .0102 -- None Political Effectiveness .0848 .1632** Education p .05 = *; p .01 = **; p .001 = *** 75 controlled condition, and may represent largely the influence of the political activity component. The contrast between the results of the two selectivity measures is interesting in that higher contact with experts, implied by the first measure, seemed to have little effect, while political conversa- tion contacts (who need not be expert) implied in the second measure did. Thus it appears, to some degree, that talking about politics with others appears to be more important to political activity than whether contacts show one's willingness to seek out expert help. Mass Media Credibility In Chapter I, we hypothesized that: H7: Perceived media credibility is positively related to the level of political effectiveness for urban low income blacks. To test this relationship we computed zero order correlation coeffi— cients between the respondent's judgment of television and press freedom from racial bias and the dependent measures of political effectiveness. Also, we computed these same basic relationships but held constant several covariables which are indicated in Table 11. The results of the analysis show a significant relationship between feeling of political efficacy and a positive appraisal of mass media racial fair- ness. This result is somewhat unique in that it constitutes the only information control variable which relates to political efficacy. Since our appraisal of efficacy was set in terms of feeling able to deal with the political status qpp_and the measure of credibility in terms of perceived racial fairness of the media, blacks who generally sense less racial oppression may respond to both these measures in the same way. In other words, the factor supporting the correlation in this instance may be a general positive feeling of racial fairness 76 Table 11. Zero order and controlled productudoment correlation coeffi— cients between perceived credibility of the mass media (press and television) and dependent measures of political effectiveness. Dependent Zero Order Covariables Variable Correlation Controlled (r) Covariables Political Age, Sex, Amount Knowledge —.O305 .0116 mass media Political Age, Amount mass Activity —.0910 -.0939 media Political Efficacy .2604*** .2414*** Sex Political Sex, Amount mass Effectiveness .0717 .0939 media p .05 = *; p .01 = **; p .001 = as* in the social system. Having a sense of political effectiveness, relatively speaking, may imply less perceived hostility from white dominated media. However, it is clear from the data, that this positive outlook did not, if present, extend to the other political effectiveness compo- nents. Indeed, a somewhat surprising deviation is the lack of any media credibility relationship with political knowledge, the prime correlate of selective media use discussed earlier. One would expect that under conditions of low credibility, the information presented by the daily press and television would have little effect upon politi- cal knowledge and an enhanced positive effect under high credible conditions. Our data, of course, did not support this supposition. Political activity, though not significantly related to media credibil- ity, almost attained this status in a negative direction. On a specula- tive level, this negative correlation could suggest that perception of racial bias from the community and the media may act as a spur 77 to black political activity. Finally, the summary index of political effectiveness showed a nearly significant relationship with media credibility, probably reflecting an averaging of component effectiveness measures. On the basis of these data, H7 is pppported for the efficacy component of political effectiveness. Information Ipput Amount and Political Effectiveness The purpose of these remaining hypotheses is to test the possible mediating effect the information control variables have upon information input amount. In Chapter I we discussed the seemingly high but unpro- ductive (in a political or self—help sense) volume of communication in the inner city. For example, several studies discussed earlier had found that while interpersonal communication is high, it tended to be non-achievement oriented. Also, it was engaged in with primarily family and close friends who pass on little new advice or expertise to address to basic living problems. Equally constraining conditions had been discussed for the mass media and organizational participation. Given these circumstances, it was reasoned that high information input to an individual would be of little benefit, possibly harmful, in addressing basic living and political needs. However, if this input came selectively, from a diverse array of sources or was perceived as credible (free of racial prejudice), higher amounts of information input could well encourage political effectiveness. Basically, we have advanced interaction hypotheses which posit that high information input level under conditions of high information control is conducive to higher levels of political effectiveness. The independent contribution of this interaction free of main effects (and sometimes other control variables) was tested for diversity, 78 selectivity and mass media credibility. For diversity and selectivity the interaction with amount was tested for mass media, interpersonal contacts and organizational participation. Credibility was tested in interaction only for mass media amount. For each of these instances, we have also computed the information input amount main effect for purposes of comparison with the interaction analysis. Our earlier discussion suggests that amount of information input alone has a null or negative effect on political effectiveness. The results of these analysis are given below. A. Diversity and Input Amount: In Chapter I, we hypothesized that: H8: For low income urban blacks, information input amount inter— acts with input diversity to explain significantly more variation in political effectiveness than both conditions alone. We tested this hypothesis by means of multiple regression, a procedure which allowed us to isolate the amount of dependent variable variance accounted for by the interaction of diversity and amount beyond that accounted for by both variables alone. The results are presented in Table 12 under the ”no control” condition. The analyses were repeat- ed, holding constant covariables which showed significant correlation (p$.05) with either of the two independent and the dependent variables. Covariables used in each instance are indicated in Table 12. An exami— nation of the table shows that in one instance the interaction of diversity and amount attained statistical significance under either the no control or controlled situation. Consequently, H8_is supported for the summary political effectiveness index by ptesent data. However, we find this one result difficult to interpret since no component efficacy scale attained a statistically significant level. Possibly, 79 Table 12. F ratio1 of diversity and input amount interaction for dependent political effectiveness measures under condition of (a) no controls and (b) covariable control. All signifi— cant F ratios indicate a positive correlation. Amount X Interaction F Ratio with: Diversity Political Political Political Political For Condition Knowledge Activity Efficacy Effectiveness Interpersonal No Controls 0.0004 0.0882 1.1021 0.1370 Contacts Controls 0.2430 0.0789 0.9424 0.0262 (Covariables) Education Age Sex Education Age Sex Organizational No Controls 1.6879 0.6602 3.4491 1.7329 Membership Controls 2.0208 0.5538 -- 2.0356 (Covariables) Age Age None Age Mass Media No Controls 0.3153 1.8698 2.9497 3.5396 U59 Controls 0.5112 1.6767 —— 4.5473* (Covariables) Age Age None Age Education Education P .05 = *; p .01 = **; p .001 = *** 1F ratio represents the test for significance of the interval between squared incremented (with interaction term) and restricted (interaction term drOpped) multiple regression coefficients. The difference between the two squared coefficients constitutes the variance in political effectiveness accounted for by interaction. The value for F was yielded from the squared correlation ratio (eta) for the interval with df = 1/361 (approximately). 80 the finding is the result of error reduction due to the presence of covariables. In most instances the null findings above can be explained in View of our earlier results. Interpersonal contact diversity, for example, showed no significant relationship with any of the dependent political effectiveness measures, nor did interpersonal contact amount. Thus two predictors, essentially uncorrelated with the criterion and with each other (r12 = .03) offer little hope of yielding a significant interaction. Organizational membership diversity was earlier found to be positively and nearly perfectly correlated with amount of organi- zational membership. Two variables that are so highly correlated in this way are not likely to interact significantly. Finally, media diversity earlier did show a significant relationship with political knowledge and the summary effectiveness index. However, the interaction of amount and media diversity showed little relationship to political activity or any of the other political effectiveness measures. Contrary to our initial expectations, the main effects of informa- tion input amount showed several positive relationships with political effectiveness measures. Organizational amount demonstrated significant positive correlations with dependent effectiveness measures shown in Table 13. To a great degree, these findings reflect the strong relationships observed for organizations in our test of H2, somewhat reduced by the partialling out of diversity. However, in conformity with expectation, mass media amount showed a significantly negative relationship with political efficacy. Remaining media amount relation- ships and interpersonal contact amount showed no significant relation with political effectiveness measures. These findings, for 81 interpersonal contacts and especially media use, are in contrast to the positive associations typically found for middle class populations. Table 13. F ratio of information input amount main effect, controlled for diversity of input, the interaction of diversity and amount, and other covariables (where applicablel). Amount Main Effect F Ratio for: Political Political Political Political Amount of: Knowledge Activity Efficacy Effectiveness Interpersonal Contacts 1.8481 3.3276 2.4567 0.2610 Organizational Membership 0.2351 5.8882* 6.4824* 7.7462** Mass Media Use 0.1809 0.2075 4.7959* (-) 3.4789 p .05 = *; p .01 = **; p .001 = *** lCovariables correspond to those indicated in Table 12. (-) indicates a negative relationship of amount to dependent measures. All remaining significant F ratios test positive relationships. B. Selectivity and Input Amount: In Chapter I, we hypothesized that: H9: For low income urban blacks, information input amount inter— acts with input selectivity to explain significantly more variation in political effectiveness than both conditions alone. We tested this hypothesis by means of the same multiple regression method used above. As before, it allowed us to test the interaction of selectivity and input amount, after partialling out their main effects. The results are presented in Table 14 under the ”no control" condition. Analysis was repeated in each case holding constant covari- ables which showed significant correlation with either of the two independent variables and the dependent variable. Covariables used in the "control" condition are also indicated in the table. It will Table 14. 82 F ratio of selectivity and input amount interaction for dependent measures of political effectiveness under condi- tions of (a) no controls and (b) covariable controls. All significant F ratios indicate a positive correlation. Amount X Selectivity For (Interpersonal Contacts (Experts) Interpersonal Contacts (Political Conversation) Organizational Membership Mass Media (Television) Mass Media (Press) p .05 = Interaction F Ratio with: Political Political Political Political *; p .01= M; p .001 = #3101: Condition Knowledge Activity Efficacy Effectiveness No Controls 0.5857 0.0382 2.7158 1.2919 Controls 1.3492 0.0015 2.2675 2.3088 (Covariables) Education Age Sex Education Age Sex No Controls 1.3021 6.9369** 0.5967 2.2969 Controls 1.3269 6.9209** 0.6108 2.3421 (Covariables) Education Education Education Education No Controls 2.7676 0.2909 5.4529* 5.9420* Controls 3.0569 0.3628 -- 6.4194** (Covariables) Age Age None Age No Controls 0.6373 1.2539 0.4315 1.8191 Controls -- -- -- -- (Covariables) None None None None No Controls 0.8727 0.0000 0.0276 0.3166 Controls 0.4423 -- -- 0.0798 (Covariables) Education None None Education 83 be recalled that interpersonal and mass media selectivity were assessed in two variant ways. Interpersonal selectivity was measured both in terms of general contact with experts and proportion of political conversation contacts. Media selectivity was assessed separately for television and the press. Selectivity's influence on the political importance of information input amount is restricted to two instances. First, one's selectivity in seeking political conversation contacts is apparently important to political activity as his circle of contacts increases. Secondly, selectivity shown in joining political or civic action oriented groups bears a relationship to political efficacy as one joins more organiza- tions. Thus, selectivity does have a meaningful effect in combination with amount, but apparently not alone. However, one important question remains unanswered in any correlational analysis, principally whether selective.membership in a widening circle of organizations is cause or product of political efficacy. The trend for organizational membership is also reflected in the summary index of political effectiveness. Controlling for covari- ables affected these relationships only slightly. Consequently, we must conclude that HQ receives limited support in specific instances of organizational membership and interpersonal contact selectivity. Perhaps the greatest rebuke to expectation occurred in the case of the television consumption. Indeed, it was expected that selectivity and high viewing amount would interact strongly in terms of political knowledge. Present and earlier analyses in this study of television selectivity and amount with political knowledge show positive and significant correlations. Their interaction, however, seems to offer 84 little additional explanatory power. Selectivity shown in seeking interpersonal expert contacts here, as earlier in this study, failed to account for any appreciable varia— tion in the dependent variables. Press selectivity, though signifi— cantly related to political activity in earlier analyses, also failed to gain appreciably in predictive power in interaction with amount of press use. As before, the amount main effect did not conform in all cases to our expectation of null or negative relationship to the dependent measures. Table 15 shows that amount of organizations is related to political activity to a highly significant extent. This result reflects confounding of organizational amount with diversity and selec- tivity discussed earlier, but particularly diversity, since the contri— bution of selectivity to this association has been controlled. Amount of television use, too, shows a significant relationship with political knowledge, activity, and the summary effectiveness measure. Together, these findings clearly demonstrate that amount, free of mediating selectivity, is important to political effectiveness. However, for political conversation contacts, selectivity seems to be an essential adjunct of amount. While Table 14 shows a substantial relationship of the amount X selectivity interaction with political activity, Table 15 indicates that amount alone is unrelated to political activity. Thus, our findings show support for the importance of selec— tivity to amount in this limited circumstance. C. Credibility of Mass Media and Input Amount: In Chapter I, we hypothesized: H10: For low income urban blacks, mass media information input amount interacts with media credibility to explain 85 Table 15. F ratio of information input amount main effect controlled for selectivity of input, the interaction of selectivity and amount, and other covariables (where applicablel). All significant F ratios indicate a positive correlation. Amount of: Interpersonal Contacts (Experts) Interpersonal Contacts (Political Conversation) Organizational Membership Mass Media Use (Television) Mass Media Use (Press) Amount Main Effect F Ratio for: Political Political Knowledge Activity 1.0232 0.0232 1.0501 0.0234 0.0231 17.7968*** 18.1619*** 6.2299* 0.5169 0.5118 P -05 = *; p .01 = **; p .001 = *** Political Efficacy 2.1054 2.7419 2.7753 1.1897 0.1386 Political Effectiveness 2.1187 1.7204 1.3858 16.6157*** 0.6807 lAdditional covariables correspond to those indicated in Table 14. 86 significantly more variation in political effectiveness than both conditions alone. This hypothesis was tested in a manner identical to the previous two. Again, we isolated from main effects the additional predictive power offered by the interaction of credibility and amount. The results of this analysis are shown in Table 16. Clearly, from these data, E10115 not sppported. No condition showed a significant relationship to the dependent variable. The presence of covariables made little difference to the outcome of the test. Table 16. F ratio of mass media credibility and mass media amount for dependent measures of political effectiveness under conditions of (a) no controls and (b) covariables control. Significant F ratios indicate a positive correlation. Amount X Interaction F Ratio with: Credibility Political Political Political Political For Condition Knowledge Activity Efficaty Effectiveness Mass Media No Controls 0.2575 2.3036 0.1008 0.1316 Controls 0.2372 2.5361 -— 0.1815 (Covariables) Age Age None Age Education Education p '05 = *; p .01 = **; p .001 = **k A number of possible explanations for these null findings can be found. Credibility, for example, is viewed here only in a very specific light--perceived freedom from racial bias. While an individual could rate a particular media source racially fair, it is also quite possible that the source is regarded as incompetent, uninteresting or irrelevant and of little consequence for political effectiveness. In short, one problem may be with the present credibility measure. Another possibility is that credibility may be important only if media content is selectively consumed or gathered from diverse 87 sources. The effect on political effectiveness of perceiving Ell content as credible could well be null or negative. Essentially, information control variables for maximum political effectiveness may have to follow a sequence. For example, one might well have expo— sure to diverse inputs which are then viewed selectively. Finally, their content would have to be appraised as credible for the content to have any positive effects on political efficacy. The main effect of mass media amount was significant for political activity in this analysis (F = 5.3138, p5.05), which does not conform to our earlier expectation. The remaining dependent measures were in conformity and did not relate significantly to mass media amount. A Final Note on Amount: In our analysis of these last hypotheses, amount has frequently shown predictive merit for political effectiveness independently of information control variables. Appendix F, showing the zero order correlations of amount with effectiveness measures, reflects much the same pattern noted in the controlled amount analyses above. The majority of these amount-effectiveness relations have occurred for organizations and the mass media, and, with few exceptions, have been positive, or contributing to effectiveness. Typically, for interper— sonal contacts, there has been no positive amount—effectiveness rela— tionship. Only one negative relationship was apparent, with covariables controlled, between media use amount and political efficacy (Table 13). In light of the generally strong positive correlations of television amount with political activity, knowledge and effectiveness shown in Table 15, this finding is somewhat difficult to interpret. While the result conforms to initial expectation of null or negative amount 88 effect in the absence of information control variables, the relation occurred for only one effectiveness component in opposition to the positive correlations of the remaining components with television alone. In part, this discrepancy may reflect the use of different covariables in these analyses and indexing differences between TV and mass media amount, but it also points out the general lack of consistency of efficacy with political knowledge and activity seen throughout this chapter. In conclusion, the ability of information control variables to account for political effectiveness seems limited largely to organi- zational and interpersonal selectivity. Amount main effects, however, have shown a greater than expected ability to account for the dependent variables, particularly for the mass media and organizations. CHAPTER IV SUMMARY AND CONCLUSIONS The main objectives of this study were to (a) present a conceptual framework for analyzing certain patterns of communication behavior and their relationship to political effectiveness among low income minorities, and (b) to empirically test these conceptualizations. A Summary of the Conceptual Framework The purpose of the framework was to provide a rationale for politically effective information use. This rationale was structured to detect variations in politically effective use among black inner ‘ city residents.\ It was reasoned that the mass media, interpersonal 1 contacts and organizations provide the prime channels through which political information from the general society enters into the low income black community. Yet prior evidence showed that these inputs, while they seem to function effectively for the middle class individual, are unpredictable carriers of political and other specialized informa- tion to the inner city. Communication in the low income neighborhood seems to center on two non—complementary systems. Mass media, the primary conveyor of "outside" information, is oriented toward a white, middle class consumer society. In contrast, the low income black's network of friends and family is typically restricted to a tight circle who disdain or disregard information that is unfamiliar or comes from white, "outside" sources. The role of black organizations seems more ambiguous-—from an accommodationist, entertainment function of black 89 90 groups to a view of black organizations as a base of increased political discussion and action. However, one point was clear from past research evidence, that these channels do provide a high volume of amount of information input. Ghetto media use was found to be comparable to middle class samples and in many cases surpassed them. Interpersonal communication--gregari- ousness--was also high. Finally, evidence showed that low income blacks showed a greater propensity to participate in formal organiza- tions than did low income whites. Despite this high information channel use, numerous studies characterized inner city residents as apathetic and very low in politi- cal activity. The prime research question became one of isolating methods through which inner city residents maximize the political effectiveness of the high quantity of information that does reach them. The communication characteristics chosen were intended to show the extent to which low income blacks deviate from circumstances which make for low effective use of the information inputs they have. Basic characteristics were: I. Diversity: The variety of information inputs used by an individual. A high level of diversity was considered necessary to maximize the coverage of political events and to avoid the bias implicit in the use of one or few information sources. II. Selectivity: The extent to which the individual selects out and focuses on politically relevant information in a particular channel. An ability to sort out politically useful content from entertainment was considered necessary to 91 political decisions and action based on accurate information. III. Credibility: The extent to which a particular input is be- lieved to be free of racial bias. In the present study, use of this concept was limited to the mass media. Essen- tially, it was reasoned that political information must be perceived as free of bias to be of impact in the black commu- nity. These three characteristics were reasoned to have direct positive consequences on politically effective use of information inputs. A fourth characteristic, amount or quantity of input, was reasoned to contribute to effectiveness with the presence of the other three in an additive manner. If information control characteristics of diver- sity, selectivity and credibility were not present, amount would have less and possibly even a negative effect upon political effectiveness. Political effectiveness summarized three component dimensions: politi— cal knowledge, political activity and political efficacy (political self-confidence). Each component was predicted to be affected positive— ly and in the same manner by the information control characteristics. There is, of course, the possibility of complex interactions among information control characteristics variables beyond those exam- ined in this study. Selectivity, diversity and credibility, for exam- ple, may together affect political effectiveness. The empirical aims of the present study, however, were limited to testing the lower order relationships of each component. 92 Summary of the Empirical Study The empirical study was part of a larger field study designed to examine the communication behavior of inner city low income blacks, particularly in relation to their consumer behaviors, political effec— tiveness and general ability to cope with their surroundings. Data were gathered from a probability sample of low income black residents from high density black neighborhoods in Cleveland, Ohio, using a pretested and structured interview schedule. The performance of inter- viewers and the reliability of the measurement instruments were checked by reinterviewing a portion of the original respondents. Coding and data processing phases of this study were also checked for encoding reliability. Ten hypotheses, relating information control characteristics to political effectiveness (and its components), were tested. Two basic types of analysis were used: (3) Relation of information control variables individually to dependent measures of political effectiveness. The concern here was to demonstrate the relationship of diversity and selectivity in terms of interpersonal contacts, organizational participation, and the mass media to the dependent variables. Further, we wanted to test the relationship of mass media credibility to the dependent variable. (b) Relation of diversity, selectivity and media credibility in interaction with information input amount and their conse- quent effect on the dependent measures of political effective- ness. Interaction hypotheses involving diversity and 93 selectivity were tested across the three basic types of information inputs (persons, organizations, media). The interaction of credibility and amount was tested for the mass media only. The summary results of these tests are shown in Figure 14. Where necessary, tests were controlled for (a) an implicit correlation between diversity or selectivity and amount (there being no diversity or selec- tivity if there is no "amount" of an information input) and (b) covari- ables of sex, age, income, education and amount (except when used as an independent variable). In order for covariables to be used in a given test, they had to correlate significantly (p5.05) with both independent and dependent variables involved. Thus, our analysis provides partial support for seven of the ten hypotheses. However, in no instance was an hypothesis confirmed across all dependent political effectiveness components. Mass media and organizational diversity, mass media selectivity, selectivity for interpersonal political conversation contacts, mass media credibil— ity and the interaction of interpersonal and organizational selectivity with amount showed significant relationship to at least one component of political effectiveness. Dependent measures of political knowledge and political activity were most frequently related to information control variables. Limitations of the Study Amount of information input complicated several hypotheses, especially those involving selectivity and diversity for organizational and mass media inputs. Diversity of organizations so completely inter— correlated with total number of groups belonged to (amount) that the €94 mmusmmoa ucmvcmamp Ham co Hanz mousmmufl ucmncwawv Hfid pom Haaz monumema unopcmmmv Hflm now Hfiaz swapnucou umatnmcoo Haaz H422 Haaz Hanz umnufimcou Hanz mousmwoa ucovcwnwv add no Hasz MOHDMMME unmvnmamu Had you HHaz Haaz vuahwmaou Haaz Hafiz «Boouao .eo-m6 .a mmocm>auoowwo Hmuwuwaom mo mouamnoa humnfism new ucocoaaoo mmoco>wuuomwo Hmuwuwaon mo mousmmoe Ahmaasm new unocoaeoo mmmco>fiuuuwwm HGUfiuHHom uo mouamnofi humauzm pom ucmcoaaou mmmcm>fiuoowwm HmUfiufiaom sumuumum Hmuauuaoa s0w>fiuu< Hmuwuuaom ownofiaocz Hooaufiaom mmonm>wuoomwm Hmofluwaom mum0fiwmm Hmuaufiaom sun>auua seasonaoa aweoasoce amusuwaom mmmco>fiuuowwo HQUHuHHom mo monsmmofi mumaeam mam neoconsou "How oucmHum> mo mHmAHNQ< musmcw Hmcomuoanouau new Hmcowumuwcmmuo .mfiwoe mmma mmouum mmoco>fiuuuwuo Hmoauwaoa mo mounnmme humufism cam acmcomEoo “How muanHm> wo mammamc< mmocopwuomwmm Hmuauwflom suaoamtm snouufiaoa suapnuu< amouuaflom owpoaaocx HMUfiuflHom “new cowumaouHOU uumm ummH no vonwuumop mm voeuCcoofl .suaafipweouu memos Lua3 muumuouca om: memos mums we unsoe¢ Ammoumv wane: mmmz AGOfimfi>oHoHv memos mmmx nanmquEmz Hmnowumuficmwuo Anuomucow definmmuo>coo Hmoflowaomv muomncOo HQCOmuoauoucH Amoummmmv muomucoo HmcomummumucH "no“ unaaw mo >uw>fiuooaom cows muomuoocfi usacfl mo unnoed .usmafi mo Auflmuo>fiw cues muomuoucw uddCH mo unsoa< .mmofim>wuommwo Hmowuqaoq cu nonmaou hao>fiufimom ma sufiflananmuu «name 6662 6 can .n m . momonuomwm 66stflusoo Hfiaz sentences Hasz mmuzmmoe uCupcoaow Had no“ Hanz monswuofi neawdmomw Haw now Hasz HHJZ Hfinz Hajz vofinfiwcou eueuuucoo Hasz vofiuflwcou sentences Avoaqucoo Hasz Heuetfimuoo «vmfiuflucoo sentences Hasz Hasz emctwucoo wouammma udmncoaov Adm “Ow.HHJz esouuao mmoco>wuuowmm HMUfiueaom suauamwm unsouaaoe snw>anua a: finance omruHSocu Homemade; ”HOW COqunuonxo) out; mnccm>nSnummu Hmonnflaoa we mnnsmtas zunE32m poo neocoaeou ”new cowsr~onno. “cam mnmcu>nuuoawu Maoauamoa Lo :zcnmxun womeasu new acutcnuoc "new conscionHOu one; mmoCu>fluummwm HUuHuHHom mumuiweu Housmafiom s1s>wu6< lmtnnaane assessors Emtuuumsa "MOW :OHUmHoMACu mum; wmocm>fluoowwu Hmuwufiaom evenness Hmuauusoa s0a>auu< mmuiuuloa mwnoasgus Ha..suaom ”tow :oHucHunnc» nnmm mmocm>auovwwu «moflnaaom suauamua Hacauuaos scu>u06< Hmuqusaca wmvvaaocx Hmuwnqfltm “no“ newuoflmupoU whom mmocm>wnoomww amounwaon evacuees quunwaon Aua>wuu< Homeopao; owvo43cnm Hmuwuaaom "new coflumfiounou upmm mmmco>fiuuumwo Hmownwflo mo mwusmmoe mnmzn: was nCoCOndou "no sowumaonuou nufiuo ouuN a w u owes numwu ofiu mo oEoouso new some anon on new: posses .mommnuomze mo >umeanm "hpnum mo mwcwvaflw can: .ea musmfle Amouducou cofiummuv>noo Annauqfiom newv AmuumaA Low: Homecou moms .mninu>fiuuowwo Hno«u_.an to runaway m_u>ndn.3a .H nonnsoz coflusaLowcq Hirn.nua“ozcw we was Hts :« mnfi>auc;unm ”.m .nw9:o>nu;moim Hwodnafitm 3“ women;n Aao>nu.mon 1“ m, was» cut»: «mono» ;>«3;n >._n ”Lafioa CH mastuuDCJG u>qaw ..m ”mm v Armoua emu COM» Aconw>oHoH unwV .mmoco>wuomwwm Hecwuflaoa cu pwuaflou hHw>fiuqroa ma unwucou maven noun Hmuwueaoa e: um: e>fiuuomum ” m .uwmco>finouwwo aaofluafinn Co noumfiun xao>wuflroe ma hmwmum>ac unsatunfiaa AeneaumNHCGmnc “mm .mmmcr>wnpowwo Hmuauwfiom ch wuumflaw hamsnufimon ma xuwunmrm: cwwir wmmx ”Nm .mmocm>wuommmo Hummuafloo o“ mousse» m~o>auemoa ma Annmp0>wo nuanced decenhvnnoHCH .~= 1141411 mouacuoc>m 95 two variables were essentially synonymous. Similar but less severe problems occurred with selectivity of organizational membership and media diversity and selectivity. Partialling out "amount" reduced some initially substantial relationships to a marginal status. In all these cases, problems occurred mostly with measures of political activity and knowledge. Though we were able to control the effects of amount in testing hypotheses, its persistent positive correlation with political knowledge and activity does question a basic contention of this study. While we have advanced that organizational and mass media amount would show null or negative relations with political effectiveness measures, much the opposite seems confirmed. Amount, with few exceptions in these instances, appears to have significant positive relationships, even with other factors controlled. However, in two instances, (amount x selectivity of organizations and amount x interpersonal conversation contacts), the original contention was affirmed, namely that the interaction of the information control vari- ables with amount was significant, while the main effect relationship of amount was not. Also, quite apparent in this study was the inconsistency of political effectiveness measures. While our hypotheses were framed in terms of equivalent effects for the three political effectiveness components, this hypothesized uniformity did not result. For example, while political knowledge showed a relationship to media selectivity and diversity, political activity and especially political efficacy did not correlate strongly, or, in many instances, at all. Conversely, mass media credibility showed a significant relationship with political efficacy, but not with knowledge or activity. 96 These difficulties can be traced to several problems. First, the conceptual basis for political effectiveness is far from established or clear in the measurement literature (see Robinson, 1968; Neal & Rettig, 1967). While there is some rationale available to expect the diverse behavioral, attitudinal and knowledge components of politi- cal effectiveness to relate uniformly to the hypotheses, it is also possible that they could be contradictory in a relationship. For example, while one could feel politically effective (attitude) he might objectively be ineffective (in terms of his behavior). There is, however, little theory on which to base these non-uniform expecta- tions. A second problem can be traced to the difficulties of measuring political effectiveness in the inner city environment. Most of the effectiveness indices used in this investigation were adapted from those used on predominantly white, middle class populations. Little content in either the present activity or efficacy component measures strongly dealt with extreme militant civil rights action or movements for black separatism. One can reason that individuals of this persua— sion likely would feel ineffective or evidence little activity in terms of the more traditional items used. Indeed, they are not well accounted for by our measure. Finally, our measure of political knowl- edge concerned only the identification of political figures. Though potential respondent hostility and suspicion restricted our ability to include items dealing with knowledge of current civil rights events and local racial controversy, items might have been developed which would have assessed the respondent's knowledge in dealing with hypothet- ical race and low income related political situations. 97 There are other principal limitations to the present study which Vvere better known at the onset. Perhaps the most obvious and constrain- ing is the lack of a white low income control group. While we have created a conceptual and in many instances an operational basis for isolating information control characteristics of particular applicabil- ity to the low income black community, we have no white comparison group to substantiate their unique properties. Indeed, some of the findings do not appear to be unique from the standpoint of past re— search. For example, the "amount" variables in the present analysis, except those for interpersonal contact, showed significant positive correlations with the political effectiveness measures (see Appendix F), 1nuch as they do with a middle class white population. Mass media seem.to have approximately the same positive effect on political knowl— edge for black residents as it did in Chaffee's (1970) general popula- tion teenage sample. The positive political consequences of organiza- tional membership also seem about the same for whites as for blacks. The likely occurrence is that though the direction of the relationships is consistent with white groups, there well may be a "levels” or magni- tude difference. For example, though there is a positive relationship for both whites and blacks of organizational membership to political effectiveness, the level of membership for one group may be greater. Inie problem, of course, is lacking a white comparison group, we can only speculate on probably differences. Another difficulty is that while we have posited a time order theoretical model, the measurement model (correlation) and the type of survey method (one time only) do not permit testing in a causal or time relation sense. The problem of time order is seen in the 98 contrast of political activity measures to attitudinal political effi— cacy. On an intuitive level, political activity seems the overt result of some motivating force such as mass media content or organizational social pressures. 0n the other hand, the basic predispositions implied in political efficacy would seem to be the cause of various behavioral states. In our analysis, however, we cannot test this distinction. Finally, for the sake of testing primary relationships and the lack of appropriate theoretical bases, we have ignored several higher order interactions of information control variables. This, of course, means that the conceptual basis of this study is not completely exploit- ed by its operations. Contributions of the Study The conceptual framework advanced by the present study suggests a closer examination of media organizational and interpersonal informa— tion sources in isolating the determinants of politically effective behavior. Traditionally, the contribution of these information sources to political effectiveness has been viewed in light of their magnitude or amount. The present study suggests an alternative stance; that information sources should be examined in light of how they are viewed or used. This position was underscored by our examination of diversity, selectivity and credibility as mediators of input amount. Secondly, the selection of information control variables and the manner of their Operationalization have been adapted to black inner city residents. The dependent variables of political effective— ness have also been adapted as measures to inner city conditions. Though parallel concepts have been tested and interrelated for general popula— tion samples, the present study represents an attempt to direct them 99 toward variance within the inner city. In doing so, the study has demonstrated that there is significant variation in black political activity and communication behavior and that these two conceptual domains are related. All too often, traditional, middle class biased operations have characterized this setting as homogeneously alienated, apathetic and devoid of politically productive communication. The Operationalization of political effectiveness has rarely been elaborated among low income populations in terms of the multi— dimensional approach used in the present study. Studies of these groups have typically dwelt upon alienation or anomia attitude measures with little attention to behavioral or knowledge components. The present study suggests that the prime effect of information inputs may be upon behavioral and knowledge components, such as awareness of government representatives, petitioning local government for aid, and voting. Finally, at a conceptual level, we have posited a simple model emphasizing time order importance of information control variables. It allows for other information control practices, such as the use of institutionalized information sources (public or private) versus personal sources, or the centrality the respondent occupies in an interpersonal communication network, to be added and tested as further mediators of information input. Moreover, the model can be expanded to test the effect of higher order (complex) interactions among infor- mation control components. At an operational level, the present study suggests information use or control criteria which can be used to identify politically effective low income blacks. Too, it suggests methods to detect how 100 the mass media might be used to disseminate political and civic self— help information. For example, one effectiveness measure of media programs which promote civic and self-help knowledge would be the extent to which they displaced strictly entertainment content and increased their own audience (selectivity). For organizations, the study suggests that by increasing the group participation opportunities of the ghetto resident (diversity) political activity will likely also be enhanced. Also for organizations, we have found evidence which contradicts the ”compensatory" theory that black voluntary organizations serve a politically dysfunctional use. However, the importance of media credibility and political conversation contacts to political effectiveness remains questionable, since there is little intuitive reason to see these factors as more causal than a product of political effectiveness. Suggestions for Future Work Given the restraints of the present study and the general lack of communication—poverty research, it is a near understatement to suggest that much additional work remains. However, in terms of what this study suggests, there are several areas worthy of particular attention. Previously, we outlined a causal or time order process model, but were unable to test it within the scope of one-time-only sampling and a correlational analysis. This restriction left us with only the ability to support causal relationships on an intuitive or theoreti- cal basis, not an operational one. A study design to make operational conclusions of a causal nature must measure at least at two time points in an ongoing process to show that "cause" preceded "result” in time 101 'order. Chaffee (1970), for example, employed a "cross—lagged" correla— tional method comparing data taken six months apart to demonstrate the effect of selective media use upon political activity and knowledge at election time a half—year later. This method could be used in much the same manner to test the causal sequence of other information control characteristics. The list of information control variables considered in the present study could well be expanded to include others. For example, the ghetto resident's centrality in the interpersonal communication network which surrounds him may well influence how well in touch he is with prevailing political opinion and activity. The kinds of indi- viduals selected by the inner city black community as opinion leaders and how they affect political activity in this setting also needs investigation. Do residents tend to heed highly vocal militants, are they attracted by the opinions of successful middle class blacks, or do they prefer individuals like themselves? What aspects of polit— ical life do the opinion leaders easily effect and what matters are relatively immune to their advice? More, too, needs to be known about how inner city residents verify rumors and other questionable political information. Are friends and neighbors called upon, is verification sought from the mass media, or is inaction the most probably result? Higher order interactions of information control variables also need more attention. In the framework of the present study, for exam- ple, credibility of the information source may have a positive conse— quence for political effectiveness only when it is preceded by diverse and selective selection of the information input. Attaching high credibility to information poorly chosen and of dubious political 102 value could well harm one's ability to cope politically. The version of credibility employed in this study——perceived racial fairness—— since it does not consider the expertise of the source would be partic- ularly vulnerable to this situation. With little doubt, political effectiveness measures need further refinement, particularly in the construction of instruments suitable for low income, non—white populations. In the present study, for example, black militancy was not directly tapped. Existing scales also need clarification in terms of their dimensional structure. Too often, indices like political efficacy bear nearly total operational overlap with conceptually different indices such as dogmatism, control, fatalism, self—esteem, alienation and anomia. Different political effectiveness measures, other than traditional attitudinal Likert techniques should be considered as well, such as hypothetical problem solving situations in which the respondent is asked to find solutions to common, troublesome political problems (see Mathiason, 1970). Behavioral inventories (like the political activity scale used in the present study) could be expanded to look intensively at involvement in a greater range of inner city activities, such as the nature of civil rights activities one engages in (as opposed to simply some kind of political activity). Steps such as this, while they run the risk of making scales less useful among diverse comparison populations, offer promise of better sensitivity to ghetto political behavior differ- ences. Also, part of any effort to increase scale sensitivity would have to critically review the middle class political biases and assump— tions which may be present in existing scales, and undertake an effort to better find out the characteristics of ghetto political savoir—faire. 103 Behaviors and methods generally shunned in middle class political circles might well epitomize effectiveness in the inner city political environment. Research also might direct attention toward points of discrepancy or breakdown between outside information sources which serve the black inner city and the inner city resident. One possible method to accom— plish this is the coorientation paradigm used by Grunig (1970) to show discrepancies between black low income residents and a government housing agency. In this instance the method showed differences in outlook towards common problems and assessed how well each party in- volved understood the feelings and outlook of the other. Finally, little attention has been directed toward message strat— egies which will increase political effectiveness. Indeed, the present study as well as the bulk of poverty—communication research is concerned with field descriptions of the status quo. Far greater research effort must be directed toward experimental designs which will isolate methods capable of holding interest and motivating greater levels of effective inner city political behavior. Which agencies, persons or media hold most promise for this kind of task, and what are the most visible and effective kinds of programs to promote political interest--for what kinds of problems? How expensive are they? How will effective methods be guarded from abusive use? All are questions of growing importance as one moves from description to experimentation with alter— native communication strategies. Indeed, potential questions easily outweigh answers and will do so increasingly as more is known of inner city communication and political activity. BIBLIOGRAPHY BIBLIOGRAPHY Ackoff, R. K., (1958). ”Toward a Behavioral Theory of Communication." Manpgement Science, Vol. 4, pp. 218-234. Allen, Thomas H., (1968). "Mass Media Use Patterns in a Negro Ghetto." Journalism Quarterly, Vol. 45, No. 3, pp. 525—531. Almond, G. A. and S. Verba, (1963). The Civic Culture. Princeton, N.J.: Princeton University Press. Ascroft, Joseph R., (1969). "Modernization and Communication: Control— ling Environmental Change.” Doctoral Dissertation. East Lansing: Michigan State University. Axelrod, Morris, (1956). "Urban Structure and Social Participation.” American Sociological Review, Vol. 21, No. 1, pp. 13-18. Berelson, Bernard R., Paul Lazarsfeld, and William McPhee, (1954). Voting. Chicago, Ill.: The University of Chicago Press. Berlyne, D. E., (1966). ”Curiosity and Exploration." Science, Vol. 153, No. 1, pp. 25—33. Block, Carl E., (1970). "Communicating with the Urban Poor: An Explor- atory Inquiry.” Journalism Quarterly, Vol. 47, No. 1, pp. 3-11. Campbell, Angus, Gerald Gurin, and Warren E. Miller, (1954). The Voter Decides. Evanston, 111.: Row Peterson. pp. 187-89. Chaffee, Steven, Scott Ward, and Leonard Lipton, (1970). "Mass Communi- cation and Political Socialization.” Journalism Quarterl , V01. 47, NO. 4, pp. 647—6660 .Chilman, Catherine S., (1965). "Child Rearing and Family Relationship Patterns of the Very Poor." Welfare ip Review, Vol. 3, No. 1, pp. 9-19. , (1966). Growing up Poor. U.S. Department of Health, Education and Welfare. Welfare Administration Publica— tion No. 13, 117 pages. Washington, D.C.: GPO. Clark, Kenneth B., (1967). Dark Ghetto: Dilemmas pt Social Power. New York: Harper. Cofer, Charles N. and Mortimer H. Appley, (1964). Motivation: Theory and Research. New York: Wiley. 104 105 Cohen, Jacob, (1968). "Multiple Regression As a General Data-Analytic System." Psychological Bulletin, Vol. 70, No. 6, pp. 426—443. Columbia Broadcasting System, (1968). White and Negro Attitudes Toward Race Related Issues and Activities, 37 pages. Princeton, N.J.: Public Opinion Research Corporation. Dervin, Brenda, (1970). "The Communication Behaviors of the American Urban Poor: A Summary of Research and Suggestions for Future Research." Project CUP, Report No. 13. East Lansing, Michigan: Department of Communication, Michigan State University. Donohew, Lewis and B. K. Singh, (1968). Modernization pt Life Styles. Ag Appraisal pt the ”War pp Poverty" ipflp Rural Setting pt South— eastern Kentucky. Lexington, Ky.: University of Kentucky. Douvan, Elizabeth and Alan M. Walker, (1956). "The Sense of Effective- ness in Public Affairs." Psychological Monographs, Vol. 70, No. 2, (Whole No. 429), pp. 1-19. Easton, D. and Jack Dennis, (1965). "The Child's Image of Government." Vol. 361, pp. 40-57. , (1967). ”The Child's Acquisition of Regime Norms: Political Efficacy." American Political Science Review, Vol. 61, pp. 25-38. Epstein, Lenore A., (1961). "Some Effects of Low Income on Children and Their Families." Social Security Bulletin, Vol. 24, No. 2, pp. 12-17. Frazier, Franklin E., (1965). Black Bourgeoisie, 222 pages. New York: Collier—-original ed. in French publication, 1955. Gieber, Walter, (1960). "How the 'Gatekeepers' View Local Civil Liber- ties as News." Journalism Quarterly, Vol. 37, No. 2, pp. 119-205. Greenberg, Bradley S. and Brenda Dervin, (1969). "Mass Communication Among the Urban Poor.” Project CUP, Report No. 5, Michigan State University. (Mimeo) - and Brenda Dervin with Joseph Dominick and John Bowes, (1970). Use of the Mass Media py_the Urban Poor. New York: Praeger Publishers. and Joseph R. Dominick, (1968). Television Teenagers. Project CUP: Communication Among the Urban Poor, Department of Communication, Michigan State University, Report No. 4, 36 pages. (Mimeo) ‘ 106 Greenstein, Fred, (1965). "Personality and Political Socialization: The Theories of Authoritarian and Democratic Character." The Annals pt the American Academy pt Political Science, Vol. 361, pp. 81—95. Grunig, James E., (1970). "Communication, Community Decision, and Low Income Housing in a Maryland Suburb of Washington, D.C." Research Report, Urban Research Group, Bureau of Governmental Research, University of Maryland, College Park, Md. Haer, John L., (1956). "Social Stratification in Relation to Attitude toward Sources of Power in a Community." Social Forces, Vol. 35, No. 2, pp. 137-142. Halldorson, Marvin Harry, (1969). "An Iterative Multiple-Regression Technique for the Factorial Analysis of Variance with Unequal and Disproportionate Cell Frequencies.” Doctoral Dissertation, Colorado State College. Hawkins, Stuart H., (1964). ”Influence of Role Commitments, Perceived Role Conflict and Gregariousness on the Publication Behavior of Agricultural Scientists.” Unpublished Master's Thesis, East Lansing: Michigan State University. Henry, Jules, (1965). "White People's Time, Colored People's Time.” Trans-action, Vol. 2, No. 3. Ingram, Jim, (1969). ”What Blacks Read--and Who They Believe." Detroit Scope Magazine, Vol. 1, No. 39, pp. 6—8. Jain, Nemi Chand, (1970). "Communication Patterns and Effectiveness of Professionals Performing Linking Roles in a Research Dissemina— tion Organization." Doctoral Dissertation. East Lansing: Michigan State University. Jennings, M. Kent and Richard G. Numi, (1968). "Transmission of Politi— cal Values from Parent to Child." American Political Science Review, Vol. 64, pp. 169—184. Keith, Robert, Dharam Yadav, and Joseph Ascroft, (1968). "Mass media Exposure and Modernization Among Villagers in Three Developing Countries: Towards Cross—national Generalizations.” Mass Commu— nication 33g 3E2 Development pt Nations, Ch. IV, David K. Berlo (ed.). International Communication Institute: Michigan State University, East Lansing, Michigan. Kelly, Francis, Donald Beggs, and Keith A. McNeil with Tony Eichelberger and Judy Lyon, (1969). Multiple Regression Approach. Carbondale, 111.: Southern Illinois University Press. Komarovsky, Mirra, (1946). "The Voluntary Associations of Urban Dwell— ers." American Sociological Review, Vol. 11, No. 6, pp. 686-98. Lane, Robert E., (1959). Political Life: Why and How People Get Involved it Politics. New York: Free Press. 107 Lerner, Daniel, (1963). "Toward a Communication Theory of Moderniza— tion," L. W. Pyle (ed.). Communication and Political Development. Princeton, N.J.: Princeton University Press. pp. 327-350. Liebow, Elliot, (1967). Tally's Corner: .5 Study pt Negro Street Corner Men, 260 pages. Boston: Little, Brown and Company. Litt, E., (1963). "Civic Education, Norms, and Political Indoctrina— tion." American Sociological Review, Vol. 28, pp. 69—75. Lyle, Jack, (1967). "The Negro and the News Media," in Jack Lyle. The News 22 Megalopolis. San Francisco, Calif.: Chandler Pubf lishing Company. Ch. 9, pp. 163-182. Marvick, Dwaine, (1965). "The Political Socialization of the American Negro." The Annals pt the American Academy_ of Political and Social Science, Vol. 361, pp. 112— 17. Mathiason, John R., (1970). "Communication Patterns and Powerlessness among Urban Poor: Toward the Use of Mass Communication for Rapid Social Change." Paper presented to the International Communication Division, Association for Education in Journalism, Washington, D.C. Matthews, D. R. and Prothro, (1966). Negroes and the New Southern Politics. New York: Harcort. ' McCombs, Maxwell E., (1968). "Negro Use of Television and Newspapers for Political Information, 1952-64." Journal pt Broadcasting, Vol. 13, No. 3, pp. 261—66. McIssac, Hugh and Harold Wilkinson, (1965). ”Clients Talk About Their Caseworkers." Public Welfare, Vol. 23, No. 3, pp. 147-154. Menzel, Herbert and Elihu Katz, (1955). ”Social Relations and Innova- tion in the Medical Profession: The Epidemiology of a New Drug." Public Opinion Quarterly, Vol. 19, pp. 337—352. Milbrath, Lester W., (1965). Political Participation. Chicago: Rand McNally and Company. Miller, James G., (1965). "Living Systems: Basic Concepts." Behav- ioral Science, Vol. 10, No. 3, pp. 193-237. Minuchin, Salvador, Braulio Montalvo, Bernard Guerny, Bernice Rosman, and Florence Schumer, (1967). Families pt the Slums: Ap Explora— tion of Their Structure and Treatment. New York: Basic Books. Mussen, Paul H. and Anne B. Wyszynski, (1952). "Personality and Politi— cal Participation." Human Relations, Vol. 5, pp. 65-82“ Myrdal, Gunnar, Richard Sterner, and Arnold Rose, (1944). 52 American Dilemma. New York: Harper and Brothers. 108 Neal, Arthur and Soloman Rettig, (1967). "On the Multi—dimensionality of Alienation." American Sociological Review, Vol. 32, pp. 54—64. Orum, Anthony, (1965). "A Reappraisal of the Social and Political 1 Participation of Negroes." The American Journal pt_Sociology, Vol. 70, pp. 32-46. Parker, E. B. and others, (1968). Communication and Research Produc— tivity ip_§p Interdisciplinary Behavioral Science Research Area. Stanford: Institute for Communication Research, Stanford Univer— sity. Riesman, D. and N. Glazer, (1950). "Criteria for Political Apathy," Gouldner, Alvin Ward (ed.). Studies ip Leadership: Leadership 23g Democratic Action. New York: Harper and Brothers, Publishers. pp. 531-559. Robinson, John P., Jerrold G. Rusk, and Kendra B. Head, (1965). Mea— sures ot Political Attitudes, Chs. 9—13. Ann Arbor, Michigan: Survey Research Center, Institute for Social Research. Rogers, Everett M., (1966). "Media Exposure and Modernization Among Columbian Peasants." Public Opinion Quarterly, Vol. 29, pp. 614-625. , (1968). "Mass Media Exposure and Modernization Among Columbian Peasants.” Mass Communication and the Develop- ment pt Nations, David K. Berlo (ed.), Ch. III. International Communication Institute: Michigan State University, East Lansing, Michigan. Roling, Niels, (1970). "The Evolution of Civilization: A Theoretic Approach to the Diffusion of Innovations with Special Reference to Modernization." Doctoral Dissertation, Michigan State Univer— sity. Rosen, Bernard, (1964). ”Attitude Change within the Negro Press Toward Segregation and Discrimination." Journal pt Social Psy- chology, Vol. 62, No. 1, pp. 77-83. Rotter, Julian, B., (1966). "Generalized Expectancies for Internal versus External Control of Reinforcement." Psychological Mono— graphs, Whole No. 609, Vol, 80, No. 1. Seeman, Melvin, (1959). ”On the Meaning of Alienation." American Sociological Review, Vol. 24, pp. 783—791. Sigel, R., (1965). ”Assumptions about the Learning of Political 1 Values." The Annals pt the American Academy pt Political and Social Science, Vol. 361, pp. 1—9. 109 Singer, B. D., (1967). .A Psychological, Social, 33d Economic Profile of 500 Arrestees. A report prepared for the U. S. Department BflLaBor, Manpower Administration, Research Contract No. 81-24— 68—03, 72 pages. Department of Sociology, University of Western Ontario, London, Ontario, Canada. Srole, Leo, (1956). "Social Integration and Certain Correlaries: An Exploratory Study. American Sociological Review, Vol. 21, pp. 709—716. Troldahl, Verling C., (1964). "Occupational Prestige Scale." (Un- published mimeo). Department of Communication, Michigan State University, East Lansing, Michigan. , and Frederic A. Powell, (1965). "Short Form Dogmatism Scale for Use in Field Studies." Social Forces, Vol. 44, pp. 211—214. U.S. Government, (1968). Report pt the National Advisory Commission pp Civil Disorders. New York: Bantam. U.S. Government Bureau of the Census, (1960). Census of Housing for the Cleveland, Ohio, Standard Metropolitan Statistical Area. U.S. Government Printing Office, Washington, D.C. Wade, Serena and Wilbur Schramm, (1969). "The Mass Media as Sources of Public Affairs, Science, and Health Knowledge." Public Opinion Quarterly, Vol. 33, No. 2, pp. 197—209. Ward, Joe H., (1962). "Comments on 'The Paramorphic Representation of Clinical Judgment.'" Psychological Bulletin, Vol. 59, No. 1, pp. 74—76. , (1963). "Note on the Independent Contribution of a Predictor." Psychological Report, Vol. 12, pp. 197—198. Watzlawick, Paul, Janet H. Beavin, and Don D. Jackson, (1967). Prag— matics pt Human Communication. New York: W. W. Norton. APPENDICES 110 APPENDIX A Letter of introduction handed to respondents prior to onset of the interview. MICHIGAN STATE UNIVERSITY EAST LANSING - MICHIGAN 13321 COLLEGE OF COMMUNICATION ARTS - DEPARTMENT OF COMMUNICATION July 1969 This is to introduce , an official interviewer for Michigan State University. We are conducting an opinion study in Cleveland. If you have any questions, please call the Michigan State University Cleveland headquarters at Holiday Inn, 3614 Euclid Ave. (phone 432—1130). Mack Clemmons 8 Associates, Inc. M.S.U. Survey Project 111 APPENDIX B Sample selection and not—at—home household replacement procedures. 1. An initial sample frame was compiled according to the criteria outlined in Chapter II. 2. Within this frame, initially 40 blocks were randomly selected. 3. Household selection on a given block proceeded in the following manner: a. A team of interviewers listed all households on a given block. b. This total less 15 homes was divided by 5 to yield the number of households which should be skipped between each of the five household clusters of three homes. c. A corner of a given block was randomly selected as a starting point and a value within the range of 1 to the skip interval was selected as the starting house of the first cluster to be interviewed. Interviewers proceeded around the block after the travelling direction (clockwise or counterclockwise) was randomly determined, counting off homes sequentially until the start household of the first cluster was reached. Interviewers were instructed to secure respondents at the next two adjacent households in the cluster. Using the skip interval to count off homes between clusters, the interviewers proceeded around the block until they had completed interviews at the 15 listed addresses. 0. Ideally, had every selected household in each cluster responded, only ten interviews or 3 1/3 clusters per block would have been necessary. However, the high rate of vacant homes or ”not—at- home" respondents made it necessary for interviewers to utilize the full 15 households listed for each block. e. These 5 extra replacement respondents per block were not adequate to meet the non-response rate. Consequently, five extra blocks (PSU's) which had been randomly drawn into the original sample were used to secure additional subjects. f. Both of these replacement procedures were not enough to compensate for non-responses. Further, it was necessary to screen out 1 several blocks from the original sample due to urban renewal and the presence of clearly upper income families. Consequently, 112 a second sample, lying entirely within the frame of the first was selected. The frame for the second sample constituted a much more confined area limited to the most crowded, low income blocks. A map of the first and second sample frames and sample blocks selected is contained in Appendix I. Procedures in securing households were the same in the second sample. 113 APPENDIX C Copy of the complete questionnaire used in the present study. The television log shown on the second page was changed each day of inter- viewing to correspond to programs aired the day before, the period for which television viewing was sampled. Person contact sheets found near the middle of the questionnaire were used for up to six contacts the respondent had indicated talking to the previous day. m ... ..m UJ< .(n -. u . nus—E Suvnm an :2. ......Z n. ...-m 3.3 1.52.1.1an “Hum“. :62 Sex .Hho1< . T.- HE: 2...... ... Hung-5:50. n n31: rig—.0 :25 £3» .. x I 11 Dv—HcO .. Hi; 1,-2.2:m $1030 .. ..E-HH..:Hm .. ... ...-’23.} .. .52....) d .. I... .. :5 H2 1 5m ...23 o .... . 1 . .o ......._..._ mmawOF—OZVI fl: (:3 u. | .. 9:15 a: ...... hxjfil 1.532.} n. .14. ...a 2:47.... : .5 (3:4 . ...I-u — 3!. i . a :31 n .5} >2... EnoHo .wn I van )otuE: gin—am p102. .. :mxn 3.62 .....w. gm uH. . ...: . ....IurJ “SIM? 1.1.116. I. IIII..1......¢... I8. ISM}: I ...-£3 0:5 ...> 52.3 1:. ....< ...a Bel—...: eH. .. ...—Etw m_. .. Suc< tier—Hon I.on I.mv 0.1.1.330 H: Inez 2:3... .. ...... . a ...“...u ... ...uaoo 1| 1 Ill. Ida: I 8 — :52 in m ..u...........u.v......uq......u...u. 1...... -III‘II'I‘IIII-"I|IIII I—v‘pIa “F ’O'IOII‘II I WfiflvflII GOIWI-le. 11— —|-m. .ncs —. | 1 DOI—uhunVI 1A Flu—10‘ . 5.3. 5 u: o.< 0.2 3:33 0* .133 mN.N._._m>3. ”Nixd3.“ “.353 .n3w3 .uhlk ...FNHHF £2 7H,}... a; > 3.1.. can: m u: BL...” Hf H .\._1— w m _ _ H. 89:32.08 mm; 39.03 0: / Q0332: .Hm ..Hm HP to .6... 2K. Smgmumm; :onSuHou co snows... 20> mHHocm so WI IEQHo put: 2...? on . . HEwHHuc L65 3 ~22: m... CH; NHHH... mm H98 :5; m 8.5; vie. m 8:3 0:. m . r. by IR. Ft 4.1.1320 B3... A . H 815 Bow 65.1.0... .3 u Hr .H. 30> flrHH... >515». ......vfi... 253 S? C; 9:; 3.2: 50> ..Ho couf. Op 2: HR...“ ALE: L:C> 2.. .52. :C> 3. ALE; H53 nix... :ch...._H.:H..L 55...... HQ; .h . 8 Hm 3 t. . s A: 3 [HOLw .HIEHH “(0.3 .H... ”Mauro (.... LEA; :23. .srh kc ML12< ..E H_:. HngH UEFUEV EC .2<: . E19? Edam PH. 3193 50> NH umna .cofimgmaou “been mmcflfi 9.8m 30> game 3 Hum on ma p.m.... ..H . firm ......I . >3 m. 0.6. 5.5. 8.9.... w >3 a $393 wtwc 9b m >3 m. 5&9”. mam: 9.0 .. 2?). m vuuo 50rd m xmw... .n. 3:0 5...? mme H ,II. .7..an mownmp co mFm....o.Hn mH.MH.H ...u ImeH no; at 35.7. 20:. 9.2.2.6... ....H H.) ..InkJU .5 «ram 2 _ ....xb Ho uwnh. H. nfu no 95.... m 5.4? .....H )6 ....muu «o 20: .H _ III II- no.) .4. o: .H .uw MEI mo n.3,... o» EumIH 30> puny man. Incmtuu CHE: 5.3.9:... ... 9.1.x. ...c.H. .d 115 —> . I 26.03 0.33 89:73.53 2.6 ... misc; 0» cmwmn Lo>oz H050... ...H ......zm...< .. 0H «062 ”5 cu Spud =2. 8 J): HEP. ...dc. #330 {and 5520 cm. no .039» 30%.... 30> an. ME #3 .302 . \tYVKMIG . mcommgmafi. no 3.50.2.5 Cayman... (tr. 5.. ...o> .6: 2. 23.. H... > 50H \7 (5?... 3.022 9L...» 05. mo Bow 0.5.. v.3 06 “was. 01H m J 5.5 2.. Hr, :m m 2...? no 2.58% Huang gmfln mhmIIchfi 30> ow v9.5 "rm :wfi. MVHIHHHLu WE‘VO .COdmw>0HwP CO >N§H flwPflaH U. :23 91014.n— IHUJH 4.1.4— .1234; 2.5.) .oH .IOHMO 8N DQL SH» “say we "HF—Cu I, w owns mat mo uman m 9....3 "Eu Mo HHm .. 3:0... 3:2 8. urn... 50> a.m..» >... ....0 van.” I\ WI 0.3%? agiligpé 05 :3 3. co n10 ......H xi... :0\ Q... .-. ......n. ...H .H HRIIH. ....w... ..u 9D 5.5 vHQpI I. \....Q n. ....F.I...I.0I..I. .....t... G LIIIm \Vmé ... rrIHoLn. v.33 Tho .. meU :3 >3)... ... *0”... K 8:0 ughmiI I». Act} a. CEO :5? ..an IH \MImFth NHL. ....0 99.5.8.“ ... .35 :4 . 0t COCO SOL .Hmkoccu C... .chrmlan—CIH ...C. rhrLrCL... I....~.L +..n\x.r : x. a... (5.. Us .0 .wcsmrd nflm. .H gm... 7. . . . . .DQ\. \... .. 3G\. .570. . ...».Haxuu Pam/“mtg ow trig...» cflxm no... on. Lo .9... EN...» uHUcm PM. .50... may ...H. 3 2. no.» 0). . ‘TXu MUHIII 1.: u H....T \L C>0 ...IC... H ... 'J \‘J t“ 01.9”: 1 Cum 860 \y-l 9v)... 1. C‘ H955. $3... $5: .35.- E. SB.v 28x. .350 5H... :2... 3H 8 3rd :9... ... POH m m 942 m. 337 2 2.51m m .Zm um 32m... N S... 2.. Eur. H. I]. Linn... ..mnlqy 9H7: .\ )H3 30> 0: Lose 30: my mug Hugwo «5 Wlw ..ngr UOO:LOH. .93 .0m wamfloa usonm 04.0.1 3. .3539, mm. 6: Emit :2. 2. I z. .5? rpmfu m2: mcuagw.....m4£udfi HMEB .m...L... 0.: L2 ...”. .Wfiuhoa “50.3 $.94 L . p v 30> Us m1aomc. >22: .7. F c TILE: CUCO EOE .3 ...figfip/o :8? x09... m mmEJ wanna a mxou... guuou m .320 :mfi. my; .9.../\ magafl 35 meow c331 2.14 L: (.2 » no.» mum”? ow mfiufl, H62 9 quu mmL/flwdo... .8 .mnn/L. £233 ’1 mod“. «5 3.63 If PM...) .. (...a. , . ”flu on ....n; .30.» J L... L S v. ...:Loz 30> 95:3 yoga H. I 5... Lo mvfioz >Sfi 20A 30/? 3 undo V 07.9.... mmA N 0: when .m pm xnos 20> on $28 mfiomm I]! . . . REAP... mum“. < .3. x33 3 lmotlrm B v.43 to mmw 20> on gum 5w: warm “...—Low 0;... mUL/wmmamh 5......an mm“ ..o; puufi. .xum ; umma 93.5 \H \ .2” .3” m H L238 39$ ..meEE mmfiu N50» 9.... EU .63.. Em ...euHE 8 J ; .r: L.,! any 5.“ F33 mead. mfimr. no: LL“: 2.3.2: 398. :3 #334. C3. .5.“ no 9.1.3 ....cdwn. \LM... _. 90H: knacLF mom OH v3.19 30> Q1503 01.: mo 00 30> .150; LPIJ... . «pun R. n. «Lucian ..erfi. coflificucw L 3.203 0.66.... >mm mL. \.. TL. .uww a m mrdxbn uuonm c043. .8. Vamp 50> 333 05 “501d .55. :0 Hex...\ L}: t L C: _u\. Op vmurdx 30> m 86. B 9...: who... 3; :4 Luguo dmurfiano “5039.5 l l HNF‘JLOU) :39 09.3)... so c .. L9. mung: no oven: .56? .XUMHL .LCULRL at...» .....H U....U “at dunl» mo £83 . \ may}: "3059 0E0: :50.) ..5nL t. 0.." mad: LO $304.; 5. m3 N o: a NLFfiLU mfiu Fm .4503 COMLUOL meP a): mxnos Jun “as?“ .n our?“ we T.me new no 9.5L 3m vie: m; moon woman Lo UH} umfiz ...cm 5...; 93/...an wmnu $8 20... we we? pWrLLL L la_l II mv> m 9.. H _/ Trim/Luaéc ... .77.... 7..., 2L. 1.. L,.4.s.mH)hn)L 72 .10.» n... Luumnfl. ...L m,. J “.0 MVP... 1 Q LLoLHmMLO 3. 2 Lunch m. “Sm: :5 +3: COPE.“ mwfip ano 0p «:33 30> conmecmEo cw Lo LwnEflLL m L 9.5.3....» m, Lav: .m .533 o :0: ..«d....... 0.3.: (50> c« 7.5: ):.LL 1 .w r (L, rytn \\. LEE»: . . . .coLLv . mwf ... w m z m \otLVJM anafileqlunq .MJOL m\ NCO m v23 >23 3 ..L S ox «131.5. ...o 849% meL o. \du. 20> compmé \om ..aoHHMS harm: H I'LL O: CH ox «.... u 233...: \hbrvmwi :3er ... . LunC nuw. ..oxmmlx‘l 9&5? .ll nhtdmamavuc m nag 30> on 550 10:. 506°. .3. L0 035. FDRL m3: 0: H fly...» umu $5 omedNI u .56 €050 {€390 quuamzw: m n39 01.62.. 26m .8 . muLAufiu ......SU ho 9639.6 805L095? .8333 “scan. on yam» S. 93” 30> «5de ...»...me 3 .mn \h 118 . .2“.an ..flnnmm me we «.5vame m ..r Hm». n. mum .FHJ ARE “.02 mg 1%...“me U4......00 ...HL. 9.....an 57.1w \QL 3b». m.LJ_I|..... 3. 4. amp 3...». ....1... . \7 . Ill $.50: H ...N Ldeh . 1.1.4.4-... r k r th‘sll.( 0:53 93 mo upon m on: .O\ Lflilull.‘.|1l. 95w 0.: we “no... 8.5 9t we 3.. w . amp? 0:.L\....Lwn 8.....me 396 LmHamd H5. 5 >4. Juan. nvumg 90mg mam: .....cE... :9. on BE... .98 >93 x55 mhmrflo .LvnmnmLfi... >35 931E?» 05. 5 8590 >31...“ 5:88 39.4; MEL. >rn wflrflh o. 32.83; .71 734.3. 30> c0932. . 453 ......K. .-. ... B.,]... a.m..“ 2.7.1... .... . r2. Cu ....3 .3... 2:92. o 2 H L... m > m z o o .... m ”L .H. r L _ < _ _ .J o . 2 H LL. [I m .. o ..L L... 99 L18 BPELZF. ..t. Mmomm 6....» H160 .mcicLo h>unh0pmw> 8 nmvndu 10> 38mm. ”.53.“...8 mu.» Sm Ho um: r. myi .....H.. .m: pun mm“... .1».wa 30H.“ 0w yum... Yuan. 0.... perFHuL >w2mwmm> 8 MILE. no.) {.3 can... .>m.v..mymm> Cowpmo rum Lo mama.” my.» no on norm.“ no: warden... «....t 3m 50.? .455 0w 30> 045 Pu: .sm ..‘w ..L... . \ n ... 3.3.4.4.: 0:1. wfiwcmnmLEc omcIMu. mo mmflmg EL... 9... .33. 20> ado A» 3-38.. 2: 8 «.9635 ”SQ/d 350: gm. HEP mmfingmfl ...o Ewammmfic >5. Umvn 30> om .Nm >33. H 2.3 .5 mo 98 N .95 a...» .8 “.8... m v.5“ my.» mo 19m .. 14235 no .05» 93 “.0 30m Jpn» ecu u wmdn .whp mu.“ .8 SH. 5.25. 9.? 28. pouncmwm: §o> 5 mum. we.“ x541. sex. 8 .dmumfim 5 :8 mm»; 1 €0an WW1” WHJL? 01141.0”... (1.1.0: Cry.“ 5.0.2 020?. .5 910% :2 T.| mums: Lo .3123 .rdomyqr $3an c095; mmrfi 3 m2? .5 9.603 . momma 3.3 m5 pm Yugo: new“? 35 $0: 9 1 1 Urd mt... numb?“ m1. d run/11 1a..” .7) _ ”00.1.? 11.9 . "Lo... .10 15v. Cw Jun/.... v.1. mo“, 09.4.1. \ C: H mnOw 1.. m>mfi Lo .4153 COmhmu war? moon OH UCOHIAVO 30> COM! «1111.24. 1.6.0 («M 1H0 LULTUE m. wanna?» m vcdwé m Lovfimcoo 3C», ocoaCOm x 51.21: “......ccuccc. 1:46. 3 WWW Vivmrom m kcwaroxm Sol 3 :15: uni In 9.10: m .5969 Lmyuo Hm .N 0.6: Eh “n. H mfihmxv FHC. Du nuufinu :0; ums: 30> 9.9;. 018:: \wmmfiu . . . wunonm 53 30> 30 4mg... J/mvhmpwmx, Ecru Fax ow $6.5m); 30> can. GEE?“ a mag: m modndumm Lo Mam: comhwa mwfiv mun £0er mmnzon we #3 .h .m x) .m .m .H gm 191.58 10mg 3815.11,: Ar. «$91.49 “11.1114 I .an‘Ju‘h) Bang; :5 69.2... :Boflxfl. .xuflc. gown...“ 3.3 mm .h moaninl Lo 9001.» $91.2 comhoa much. 20;: ow 9.5; 930: EEC 1% mm node. Lo .0 \ mg m c... 1, . . $203 30> woman 95mm m5.um x13: conned 35. 30c I "ova? u.c 21.; "now. .3 triv... Ed v20: uh 1 van: Cam. MSW: .16ng ad: want. non m m>£ «6 x18... c011»: 1A....u 301 .n OP hCOHUO 30> COHPnfiACMiLO EM. no LUZ. ...: 3 05.4% .c novflardo my. unorrbm .2 mmAm szaurUm 52.5 :0 use 633 B:§.m.LwEon um IUD. PM 3?. 2:35 F]. or. .0053 .m nuzonm v2.1 30> EU vmn: ,>wv»mumm._., .u g N13 11? £11.. m “.3956 .5 SE comcmc «.1211 mw xumzu 9100?qu .H a b Evy.“ Hm m“ my.» a oc H 9.33 30> 83.11. firm.» 2» pm. to: c898 3:1» muoc "8.39 mo 21.2 ”new. 10 $52. 3.0 x15: m: moov 00ch mo 23 “mg: can uses nomuva v45 mmoo new .6 mix #55. 4 mw> o n mnom m. «$0; .8 xhos :omhma mflfi $8 on .130de 30> samursficvao cm mo Loafing m N AcmoEoEV ,8qu . c6322 ‘ L050 Haucmflpo LO 3L1) 33: (‘1 r1N xocfi; no 9.003 E xv... OE r-I @2538 m UCGNSM m LwaDCOO JO> UCOMSM umow ggm m owalm :0 use 5 .1643 39MB 9.“??an um w HZ um m y. Vang: mEoL m.com,5e .5510 um wEEQQLm use 4.5: 8.. J 0.15.02 C30 Hm. 2."er EH: 0.... V325 50>. FE: 30> 918 3 9.6L: .2022 cemgoa mdfi mH , r , 513% < Box 335 .8 9602 :cmé 5; upon? m. . . .comEa mflzw mH HWM: ...waflfim no came coax mmfi mun £020 2328 we um: nlll ‘mufika \mmme URHVFH? “mm r wunonm vamp 50> can #3,: 132331» 205 FM; on swim» 50> 55.. UHMFUM 35: N 5% H3580 28me H .h .m .3 .m ..n $91.31. cowMea mwfi. 2oz: ou who; (50> 5pm H." mw m3? :8 mm; N Luxuo acuuho ACMOALQEV fdg...‘ 031...? :85. 0:05 Lo :43wi xvcfln mp2: .5 .823 .rdoflxmr mto M: «lab r—4N x935 c0111.: 11.n— nm .5 meE 15 mydouu WHEN...“ < Mk o: H Ego: 30> woman 23 m5. Hm v20: c028 mdt. mwoc. "00%?” .WO Dcflx mum V78: “new. .wo Tray. Ame.) .310.ch Mo U5 0 Han: mam m>mfi cowuma 35. no 8 :0 ...0 E5 par? mg N Tharp 3.8.61.1" p51... 6 \\\... \ omam gmcgbm Hum/Em :0 8.6 :onm Eamon. mlmnhmn um awn pm. c: a 50.... m 9E Lo vie: com“; .1511 mwon. .m 0» 553.15 30> noflwsgwfio cm mo L933: m, 93mg?“ m 893m m Lovflmcoo 30> 38.6w \x 1.5.1: m....coan m3 3 5 m madacomw p30 : h o m ...paonm v23 50> 3U #33 1>mu2mpmm> 29C 5:. 8 » 845,.8 )14. .. \.11.1 .1 «Hwy—Mm 31w... «magnum .8 mg cowhwa mwfi. mun .xomcu 03.15% mp 93 .H u HEEBO :g .350 m 3.5.16 m AcmowLmE<.V 5%....” 2 3.4.1:... m 51.05. CHLOE 1110 Cd (fir/or. llw xumd.’ 1.” $1,... .8 .32.: .56“ £9112 2.020%. 35 wH 1N. $35 13 9.003! mwwnwu .10th. mu...“ 95...... on 9b... .90: 59C .1.." m. mmflc. no 930213.111... .9... “1.09. .m T»). m c: a «7.15... “1.0.x. woman. 91Mm 05. um v10... combo.“ mac... mwoc .um x15... 0... wnfzu one... 1.0 0:1 .1111". 91.1.1. now1ua mun? mmov no.” ..0 «:3 E. . m9, .1. o: H l n 20..“ m 96... 18 .418... comhwn. m2... moon 1m 3 £161.13 30> aoaufi.cm.1.26 :u “.0 1.9.28. m 153.32 m Bani m 53300 30> @1858 .>. v. mag 9.22.083 muggy mH .... mfim wBEmfium m Piano“? “.16 .. “.01 111110 CO U30 5 1.17.4... um; m .1111u.11m ..h1.1n1ma.m.nm1.11.1 um m. 1,8... m.:0m.u.1ccno .uo pm. m 131 um m 01.9.“ E6 um M d 2.55 .11: ow 34%.» 30> :1 : 20> 23: Ems: .m I!!! \nmam 2.. “Macmn I; musnm 53 50> V1.8 .113... .>mv»mumm> Cup. .11.... 3 tmfimw 30> can: 1~ 355.. H 39.. m «33.3 no mam... comp”. mwcv 3 £005 35:8 2. pm... ..n .1 '1", n fish’s . Em 85:8 7.0% 3mm . Arduasofié. m1.u 59.... 09.0.5 .5 511. 1451 ("I $23.3 it .3 531.5: .x maafl... 1.0 who” «.334 conga v2.2» 01.10:: 3. 95: 81.9. 51G n... ma mad... so «1.1.3.... M... mwxlw calla \\. 0x903 30> 8qu 3.3 mf um vie: new; 1c 1.11“...» mmoc \. .9...er 1.6. “EC. ...a... 1.0.1.... mum x15... 11.; meow 330. .0 may pm...» Ucm w>m£ conga must. meow new 1.0 “EC. €11.15 . I I ci> N 0... a ...nom m. w>m£ no x163 :01)..an. 3...... wwxvn .m 0» ucodcn 30> GOMHnSdLnfi...“ .13 mo LuXPwkt m. "1335...... 11 «1.1.0 H .u. m Lmvmmnoo 30> 0.1.8.1bm. .... wflm 93$)».er w womflm co uuollln coamm 533.111.181.13 “F. m. 0.10.1. 1.c.1.w1.1.11.1 1...... Hi pd m .m \nun11.1..1. . opzonm vamp 30> 3v HE... $.uvnmvuwx. 7.1.0 1.. .14.... o1 to...) $5.3 A “1:11... N wogm .5 01nd: cowhwa 25 3" 71.006 31130.1. er. N3 1” § 54.5.8 N W m 0.43 mo 38.» m5 03.33 vwfllahu.... .. 1 .. . I _ I|| . . r, ,v. 3\ smug.” nan _ m .5 33 80.» m mw wr‘wu co mrnflfi . xnflzw no.» on .2 cob; CHASE“); £3 a: 3 own: ~ .5 7ka (.050 CH vuméibrfi... . M m m All H Ufihflo H3ch (8 grown 32a... m m AI. A new a cc 5 .3 @9275: U>:.o> wcuufiuo. . . m m H Lox; ..V. Luciana \U ml: 9pm.: 3 Lo Can/“Lo: $5 59203 .388. .. . ) . .» .\ w......:o> 95p ”.an n...» $3 co 3. .co Om vcm £05m 3i» P75... K1. N 0 AI. H ...;2. use; 04., Ear... 30> 9.55 no mtfinxdfi vaogw mcoflnqé :3 m 30> Mm“, ow FHA Fox .5: mo m m AI. a LouuoU. . . n m All A zuMCOm 3m. 2 ammo.“ Lo Lo . 2 1 m m H Lovnua ion TM Lo AI p283 mp5? 330... N m AI H raring. . . n m T H 93me Lo Lozomga... n m .. .1, 30>. 5.2.. uuxm; fluogn ufl : w. 2. .I .. .. _ 1 Shim you 83.?me . . llll‘l'l‘il‘l‘ll - n - - . .. H N . o 1 T r 353.)“ 9.0305 cram . Ru, 9:qu .5 ( m All A . mLOMEuM 0: . . . m uoH m. mo.» 0: ..Huqxrdnfll, no _ w 3. Gun: H N. w muom . M.,.mn Hmu ho>u 30> 92mm \ w H. tram: 20> \L: u: .H , . \ ace 01; mm , $01,an 9:30., ....cwnumF m 59% £35.58er ,3 93.1. Low ow ucom Lo 2 :5 T } 9:50.) “flu 93L nunflfi. again...“ >5. 0.35 93 .592 m3 030mm ou om +Pn$xo you .mcommwn ucmfiwwmau no“ E. r053 .55. ban a. «2. .33 m3 mrdzumrbw usonm 539503 Lo 3m: 2.83 m: :05: mgpmfim .5 r | . - D., \m . 3 cu. ducodra uwoz on 0.3.30: r022 $33 03. m5 no .30., 8 ucmppoora ummmfi or 3303 353.33. umufi .30 £02: .\ r30.» ou ~53?“an umnc 2 339...: €ch mmwzu “3 £035.. 33.9: “3359.8 nor. on 33.5 .c. -..prnqfi .33 mmca awz333c|u~mom .38.... n. 023 muuufinma 9...? 4.13 30: co: ... 30> :0 >33 Hmflmcou 5.3 :9» cu 333333.313 \. 6.313.333 ,3qu 9.3133: 9.33 0» 83¢ urfiu 35,: 30> E393 okUCOC' 3:93:30: LC H .3395 no $.35”...an >32 Ou 005.0 m 33:38—30 ca 30> Ou ucmuLonEfi 3mg: m ..vmfi: .mm - , 1333.... numowu. Juan W: ......» We 2: .3533“ 3.93.?) d3. {83.3333 NEE bprlxu Hum 0.3. Hwnfirfh 0.“ 0.3302 fiwc? VEM/fipmvwflévwé 7353933333.»: “mo: mgmzmo E Dc.» 8. PESQLHE um E36: 2035.. x332 awn/3E 5.330..” 0.3..35 LL 30 3.39:1. rat/<03 H.153 H.253 30> HM? 30> :C $39.23;. Mm “3.33.302 3.30.353... ~33 €2.33, SE 33.32 9.: 933 30 leFUU 20 {935.2 mo 99.0“ no: m9 tho—Two <2 BF Ema/....Ulkdflc hmuc Err nSHw @303 3.30.3.3... 3g _ 233m r161" ho 3.0.335 .— tnrfni z 3.12.1.3... unwary!) v3.1.9. ...r. .r. :0F8mm wam mo..— hflfiom M315 Mm: :. .1 44.8 E Paw 01:63 5 ...... ..EB 9.33. E.“ .30 /.AV\ walrus HmMC... mwF. BSO 01503 mow. "3......Com OHImMDo $3.3m 50ng m UL... 2.2.00 8624.3 r. £363 «595 m .yddn 0. 5339.6 no mgafi, 353.3 :0 >393 35.533 0389 Eur: movie. .30 um: m 3393 cu madam EL .5 \ 1'. H \ _l mu3> 933.1333 ,8 v3.3 3.0.3» 3. 5.69.6 co +mw >( m3 333 nu 30 305.. 30\ 3240. M.,...u 4. a.3_ taco... afghan. :0 acapmfigfi .53 3mm? .33 3.30293 m3? Ow 0:33.33 9.59 33> $3.3 . , , .. .. ,_____._._._— \3; .1. «33%| gar 3c> on mnfiowm >99... 33h... 3.3 .333 38 o: $8.3m 3cm. pm 3.5803 @935 “Bud. 3303 00. .mm 9..an 3 3.8.53.8 soovuopunoOUl ..wmflw moigfiw: .mmomm :93.» mH H. ..\E .3363 0:9 HS. .3592. B. 4E 30> 3.32.30: IL. 1.. .3 3 2.. .3. ..quuhf. 3 nHEQ 3;. .10 ANV .\U wan“... . mun new 30> 3.38 93233 $333.5 :0 30 05.3. 3‘30 uwn >3. m. >33 3 3. ~53 ...u. Wm 1% En:3u..r0m E L Lon, 33m: :5 up... .3.“ EBQEQU Axwufi .353 auc.......~3 \ .wqum “30.33.49.323. .mmow :mEH m. .Bnm< . flu... 30» 380.. 1:33.: .33,“ .3547 wquCm. mc 33 (0.15.0 3 \nl 1:1. UH. So». 1,.u .83 3973 an 30 .553 303A 3033 new Sal. or»... 49.3.3 on..." goo. 09 OM 30> 3.300 9.313: .003 :0?“an 3 0.33333 _o> \um .om \m3l 31 can m Lo 3?." voom m mm 33.3.36. >35 3 >0 Emits 3.33.3 1.03. on .3 mm. . AI 93». ”23.8.4.1. m2: . 204. 4rd..- HAW» m1... . _ - - t . 30230.» new .. I , . 90908 voom m ufincm ml. rdhvaflulu r z .....b -9.ch m. .. . ..h.fi..o> m ymmH on... ..R mi. o... ...2... N unto”... no ......th 0>oum m mfizsm .25 5.502 1.01.4.2 mwswu. .>.:m.E :0: .592 ...no 3: E1633 «5. E89155? co go....p..60.5 ...6 ...dEttm ..E. .3 LEE :2. .174; .mm nofl 1 wins... Tl” . ...x v.4?» .r .. . Nam Al «33?. . 30> NFdAFQu .>Mm .0: All 554 L0 09.1r H; u... 4. . t. .l. ‘ \l I o I 2 Au| , col: H.359. w 1.8.: am?" :1. wmm 1.0%. 39.23» H N ox; rank by .a 2m v.39? >E ml ......CH ... ....>vv :5 x5... ...... ouwso>nw .50.» co mouucww “UFQA m ...mmH ucw cum ng >959. m9» 9.. .3: Hm?» 3 50% cu wm»6&0L 50> 3:. .. H a nudge: cowwo a. no 3903 0y $5.3. .Emé NSC. V1.2. Erin: pmfiu main”. Emw map. . N .wwomAc eJuwNfizom opp «0 >5. €92 >1..ch @2852 9:... 3....\ EH“. .019... L35 its. $05 5:... $59.00 van uohULn . .Lunz 30> u: 9.2.; 3%. ”$33... .23...» 0331.5. N52 25%. vcoUvmnr .... ‘16 30> ....nu 0.. avg .....wN. .hn 5.3., Hmfi CA pane H N 30> mEomo 33H 09m 95me PH. BHLO>UM (56.» CM 3&0?» 0cm. .vw 339» B." 5...": any 9.3.3 uzonm fl ”Comm pawfltmuev .43.. .r. w .23 9.2..” L wuflnoéflh (Sow 1.54....8 Fm...“ .3 wocflhw mi: "nah. .00...» 2rd. -... sown: 33.0 mi: many... Roz van» m2: mo «.11er MH mme M? H. a» 8 .N. : cco :3» some . «51.958 L8 9th.? ... 330.2. £02.... 10:. (8393 u .mPL. v.0: 90 09.5 «...a. P33 3me x55 Du» .3 mm «Ema .3. Sean >m m waaofl muggy meow UN.“ on mcaou NEH N52 .3 clam NE m>nn 3.5. :0: ... .... US. 9.0 LC 83 Ngon ou 20>. whnmro 3.303 mwomd .cfiNBH. ..E. .... .11.... .8 .oo :8: me .mm Fla x22 22: .6 .82: .5on9: .613 Eric 35 mu mHENmu mHmE (NH ...ng .uo «Hue comuwa must mH ..... xonfio w: um; mode mo 9.82 “a.m./d. «Gum/:2 < .56\ ngMH cam??— mwflu 20:3 cu 2.2 56> Eek. um mun 336 Lo 9303 >55 30; “5002 mi, N on H $195 30> mum Ha Numm 05 um v.50: compom mHfi. moon 82a no 39 new no on; 5 23m $8... 3 moon «3% mo 39 «23 can 1 rdmuwm. ”Eu $00 now no vfix ems... / A III my». N 0: H when 0 ~52 Lo #52 :8.an wwfi. woos ou ucoHB 20> Satanic am no .595: m 9333 m 9.8.2». m .5328 30> 2628 030: 50> CH mm>wH or... war/1E9. \.......COWH¢Q MEP mH . 8qu 95$»? ”umomE 20an 2.5. 5H... SOL... 53 >53? 90> on mac?» mo mung 99.... .EmHnofi Hmcownwa uaonm umnE 3 VHS 50> rdwuoa 20 95 mo x59 hvdk 2. mg... m Eudmm zc N..5.Em NEE/imfiz» «E N 1m< Jtudmm uzo v2.1? umnz muwfifiz ENEzEmH— .: a my. N313.“ .50..“ CH goHnga o>m£ ho cam H08. 30> :95 00m 45> $.2an29 anowuoa 938m. XHuu 0v 0.» 9w 30> 0:023 285 3 Nu. H N m ... .co Sacra 839...» he..- 0.2. ”5.5:. unfit: “32.3.3: awn LUNG». maxi. m5 mu. amok .. m N H .aomrfluum 5%... ”30nd yummy nu murmflu U2: 3 uduwwmflv wLH H N m N. 6.2 mp. 032m mnrfib 30... >3 8 uncut £03.. 92 Eros vnOLLQEnHmc vmfi CM “Aug... 3 m N H “51... mm Nauru amum C... \HHmwn w...\\......v.. H N m N. .2930“; H3093. LHvr... w>H0m o» 39. e9? HHS 3.3.: m8 3:8 .3 pH mxHH 3303 m mos quL H N m 3 .uo... nu 3Q. 30> .3 L3»? Bum- 33....» pm a m N H .09". >02); wow: 20L 8 mummnvu mm: 0,538 wvcwwhm \yrd: 3a.. H N m : mhuHuopa mbHanK 30.... c... uwumw L35 uQL Em maowmmo 5w .cuwu in? {L um "......J: H N m 3 fit? wfi ..NN MAX 5 3.3m. mvcmawv now woo.“ m ufifibmw évnw LON quw Puma... 30...... H N m ... u.c0v 5.31. 30.03 .amo: uOH .m prpHH m UHpuHH m uOH w @9me 6> 8 ...mo mumu< Ho» on £55 a» #23 306* 0» 93H v.9; .mcoHpmw1.v mmwfi 0p ”33me mag} ho RANK 0: 3m 955. .mufifi mo L045: m pnonm nowcwmo 56> xmm 0p 05H 9H :62 $4. :9 mmmmomb. hwfiuo 5w: 2.3o> 5:3 N380 0:3 >Hdm3m3 30> mzozm E. >:m 305 g nwnumcow PH. 5 50%,: ow urdccmaa 2.30> 93:03.an r..." m30£m >cm 0:56. mp< fi|r mason véofionxumm: .5 $5330 who: < .coau ISflmepo x903 Lo :0..E3 < 8.52 mxdn .a3oum newwom BflEfiEOU .5 953» fl>mu < .030 ..Evbm .5 SC Iflnarflwfio Lo 3me H8348 < 339% 2826 < 9...: or? Ova: .cofiaumdmmpo Hooxom .q. 23m Emummm .2802 9a." ..nfiEBwC be 039% $1.38 < .omwsm :03 Lo F63. m4 u< mu» 3. “3...... we.) oz "OH mcoflwn 30c 30> oa . .8 ”4.60.740 30> mgr/B UL». we «3 050m 9306.. 30> v3.0 0p 93% 0.03 .302 .o: mH 31E.“ H mflm: N ““395.“ .3 0.3: cumin mflt fitting... 0.: p3 .m: . mead: mo ”£007. b... 3.59.. ...: ...; 2%. 030:3 0p mtoc. «90> Hui um 10..» xiv: 70... l / mm> N O: a x $753 30> 003a 98m 05 pm vie: 520.“ mfit mwna “coma; uc vcmx I Low 0.0 ENC. . _ mum View. m: meow 8030 we we: HE: . cam 90; comgmh mwzu mmoc new ““0 GE um»? ww> N o: n 30m m gm: .6 v20: 580a mmfi meom ow magma 30> 300331350 Gm Mo 0023 a Cx.«€.A...L 3.. 933% m. pvtwncou 30> 00.00.30,... 32E {50> 5 mwsmfl or”... wcomh0mu ..x Eda 2.888. £5 mH .murumn\ Naomgwm mazu Lua: “Bonn v2.3. >H4ml.,ww_ We. 100.... (tr .t .0 WTMX pm»? .Hho: CH. nvfidsw z .36“... . molm xm< .25.de use 04$; with” 0&0: Hounocmcd 9.:me mtmflnopn Lo. .60me ,0 .ml chmfloc u 4). LHms mnwangauurtwangc. no mcfix utmnp p380 2H0; on S. cm 3.\. m...6\.rd ...: \lmfl. .wPoHaruH {any fl .9 127 E05 0.5: H 5.00. N 505 0030: m NPMLS LO DA €10LP $0 ”1.0 630 new on 5:00:00. .00 u. 5.1.. mmmzommwd 9.2 .82 09 $23 90 pow 03430 0.qu >3 0; um 00x H3 .3: 37¢ (.30.. 00. tauLE. he 1...}... .n7 E05. v.5; 5.4003 0230: mm: ccHuoHL xcHHS 30> 03 :50 ow :H :3 N H ouuruHL HH>H0 Low n...oH..0.~u.£ - .n:H pro .550 .35: 5 Va? .. T. 33...!5 3sz :30 29. How ou 3003 Lou >13 503 05 mH H000 30> 00 03:» mo 5?: N H .31. H:.....«...\ H5 .....HH‘L 0 00.: ”Leona pmm Op “Hank. ow 105... :MHUHuHHoa 0 Low vwzkamrmu Lo>w 3K. .u...m:. .2 >245 omuanp HH>H0 P5: yum 8 90m? .8» :3 SE mlhllefl 35>. 8 $9: mo 63: N H mHmHoHto 0H H» .5 $ 0 so .hn.:\.m>..u .LQFE 0t. mmm 3 mcoz Lo cquHF... .56 30> Npmdm. wcchgumcaumv 0:0 3:805 2003 mcoéw ...oHumfimaooo mzmH 30: wdemmm 0:95.539: N H NUOOLLOnerm: 30> :H mco Hqucoo Q6305. Ow Hubs .05. pom 0p \{onxhw 5.3 filo... Lei... 30> . >4... .. 4 . . . ....qu fix .....mufiwh HH>H0 80.: “00 :00 83002 >00 uHco a m>03 ..E00 0.»... 0.7. : .om £5an N H mrdHuuwHw HmHurdemwfi 53 05. :H 30> 30.. 30.... 00> 0: .\IP/ 143 93:... >72 . .Mn,.omm\ :2: 00.. .0330?“ we u3o 0:"...th 50> $3020 rdHuwaw 3.0. mm: 0.0%. 3... cu 3:450:70 00 Lo. 0:038 00m 30> 0H300 0H u2u....uH 0.3000 33.9: 3( > 45:55.. W63? 3 .335: 2.". :duHHLo 05 0:0 x003 no .039 30> .-. un0H wfi m... :3 30> :3 E6 mu" :0: 05:03 €20 0H0: 00000: mh30> ..Ho mucuHcG 3.0m wH v05.— .9. .mEd. 0%.: 3:00 9050 oHHLS wwHuH>Hp00 “034.79... :0 05v «0 «0H 0 0500 03003 gem .moHuHHoa «380 mvéHummfic 30H m “.504. 03 :5: .3 a D! 8 2 1 €0th v5}: Nuconn qu>0woom :HHVEEE 000.3st.. 3.... .Fflu HSUHHX “fl: 0.... .0» L053 05}: NcouficHme: CH acumcom 56> «Ho 0:0 00 0rd: 0L0 0.: HHS 30> 53 .3 ”E005. 0.30: 0035 mLucoh . ...an 3.0. Lno>r ......Hxfih >2 03030 cowmchmfl... CH Enamuwuwcoo 05 mo 23:. 05 0.; HHS 50>r and .3 ...000Hauoo 50>. 0H0 M5555 0.2» .00 WH00> >56 20: mm E0321 Va}: 20590 9H. 00 omH 003.0 35 we garage“. 0L0 0L... 8. . ..L HHB 30> :00 .030: 05 CH 0300; 0.50 0:80 30> x00 ow 0xHH ....H .30 .. .3 35505 now 0.0 mcHHOOLUm nH0 ucHx 00L: \ 1 m N H 40> .uo ._/ HEEL. «Hauwn. 6.10 ....x: 4.400.. ..hnu... .10.“. 00> o: H N m : .x... .H L. H.‘ 00 0... Hum NmGHchfiu new .8 mzHHoozom 00 005.2 L050 >c0 00: 20> $0: .No 0 «Inf. 0.3., H wurHLu nun... 0.5 3.508 Maw; H N m r -....L .- . mo HHHC 5000 000. HHd H p .30“... 0.. .LchHw 30> 30 £028 .«0 ”89> >HEE 16: H H N m L. ..-. I mw> O: :.. x 0...... ”....-. . . .... ...”. ... ......Hom 0.0%.: 205.. ....m.0 m“ ... .\ ... RC. 0...... «000300 >cc 00; 30> 020: H N m : .roecL‘HLnH 0.. uHLv ...... HONlooV 00080 03000 0» .335 ...H... m..Lb8 00H .HQ. 0...... 500.100 HRH: .8 H0050 009E CH umLchHm :0> 002m 005 05 00: p23 :3 H N m 3 .03...-0 so... ..EHU: m .53 u :00 09:.” 39$ .00). 0090 Hg?“ .8 05:23 0 ...H H £300 225 m CH N H N m a .0 .....v... 9.45:0 LC 5.5 m. CH m o.» 9...: 30g Op Eur. mLH m>... 000W NKHQ (Ho 9585.. Lo :33 HHEB .030 Hr. ...?» .Ho mcderm : m N H 0.de >¢Lu m0 ....n m CH I- EH0 0:0 00 uon 0>HH 30> 0H0 20:3 nHHno m 9.55 50> :05; .8 .0? 05 59C p.300.» mmHn ...Lu ..0Hn mu. 330 Mo 0.5: .H N m 3 .FG ..EHou 0.03: 03000 >300» 0....00 03.09.. «.00.me 00:5 :H: 2500 30> 203 2023 .mm umpMOHHEGo Om 0H ucuhE0>on 0030wa HmmlooV mommwuoum “283:0 H N m 1.0.00.5 u...“ ..x. 0L0 00.0 >330: 0.5.0me gHrommwm N>uH0 0H5 :H um 00>HH 30> g2 000000000 €0mech >56 30: .mm U.oH m OHLLHH 0 0prHH 0 00H H. 3068 £30» Ema 30> 8 2.00 ELF...” no so NbHo 0H5 cH 00>HH 30> 0502 mcoH ...hm .>HH.._..mm §o> HEM 30> 0330 08300.6 300 m xmm ow 05H 0.03 $.30ch .Nm mN 0mg: 765» cu 0XHH 0.03 a .xfiduao .|> .mcoHumgv 0005 cu 0:01.000 map»: :0 HRH,“ on 30 000E. .mucHLp .«o 0095: m 0.330 COHCHno §o> x3 0H 0.53 ....H :02 .Nm 129 a: 326% H wane N m'Hmwxa ONImH H mmmm 50> mm #33 .mm (50> puma m £20... 326 u 935 n \Losb n \ "320mm 2.0 a $58.: . _.\ . ..L‘ . I \ . 2 in) ) . ... ‘ mm>4r$§ 238 U39 asap/H N.JYHJJ; 2H m1»: .. >.> (0 Aoomloaov xmw: #87 $338 Ham ng ku 3E3 .50» moon >30: c.qu so: “Bonn £me ..Huwm .5 9.605 r My?” :0... mums, Fix. bfiéumm Hamuo‘v 5335950 EBSOHRQB mnow 22w maimfimn vcm vim: ...FGC ~3an \Em umbwmut Auumbhdum K33 mwufimfi you undo: wit. 5 magma 2308 menn>Hflhmw Lvox. 2E (.mwx. ymmm way .(I: .3 m \IL All. mlul «Ill 83a no Pd. no.h we 2Q «ppm v1.53 Amzwv w; mun.“ . 003a m0 vfix #3,: 23:92 Nb; muHGEmmow mxfix Elrwnwmuk 3 32.3 0: moon new mo U23 warm... NEcmm wwwt. no 5:.“ anf :H ENE; flmomu. o» ( wmzoz many 5... ~33 0:3 33K wfi Sum .33.?" .3. LOW pmwfl :9 3.40....” .3 $703 EH20 mvcmfihmnm Lo «mac: many 5 g: 3H0 £9» 3 nwnrdv c9635 E ....bm .zm $393 mudfim SEX—m so 330: $5 5 g: 3.33:2" mr‘wvnflofiv mpdfiw mumu 30m .3 a 130 APPENDIX D Troldahl's Occupational Prestige Scale The Troldahl scale is designed for coder use after the raw interview data has been collected. To make a judgment, a coder needs two pieces of information: (a) the kind of job the person has--the tasks he performs, and (b) the kind of place where the person works (e.g., a factory, an office, a restaurant, etc.). With this information, the coder makes a judgment of how much prestige the subject's job has in compariSon with the list of jobs named on the scale. Each scale position has two jobs listed. The first indicates the top range of the scale interval; the second job indicates the bottom of the interval. To evaluate a person's occupational prestige, the coder starts at the bottom of the scale, asking himself for each interval if the person has more or less prestige than each job listed. In this manner, the coder is able to assign an individual a value corre— sponding to the prestige of his occupation in relation to the scale. The occupational prestige scale is illustrated below: 12 = more prestige than a lawyer 11 = lawyer (sociology professor) 10 = assistant public school superintendent (aircraft instructor for the Air Force) 9 = consulting chemical engineer (high school teacher) 8 = sales engineer for electronic_parts compapy (purchasing agent for electric company) 7 = soil conservation worker (automotive cost account manager) 6 = credit manager for an advertising service (cost estimator for a glass company) 5 = rate clerk for a transportation company (offset photographer) 4 = auto mechanic (salesman for thread company) 3 = boiler operator in pickle factory (service station operator) 2 = waitress (newsstand Operator) 131 1 = bean and berry picker (garbage collector) 0 = less pretige than garbage collector Individuals who were unemployed, retired, living on social security, on welfare, or a student were assigned the mean occupational prestige value for respondents in this population. .a dvaaauN a.“ vawwvnuun ..J @393» > A”. nun: mu fiuwka> Amucv>auumwuu nauuuaaom aquacumanu mo auuuaJuuou «ch huflfiu one) mathuafl .mfiiannu unnnuuu>00 5 0953.4. 3 32:63:35 umuuuuaom 3.332.. anuquaaom 3.405.. 43 Zuzana. 33.3433. 1.33.3... huqucuw 433:9. :31: «a3. «0:. an:. uni. aqo... .30.: an“ a vanes...) ...: u unovuw .25; ..E: 3:4,; 350 non... m3nnflq> a $3.333. £3: I m: macro?“ .F. :03. :35 «5 ..m. 3 33 3 1.333 9.5 no": ..md.w.,Ll.. . If... 30.: 32.1.03. «no... 30...: “swear..." . hvuuun no mo. 9 I a >333... Jon ..zocg . >uuua :: 4.0:. wmof :81: 20.: ZAWS. 75 .>u«>3u<. «3m. ..nNN.I n40; Co. one. 2 .03.." um Huuuu «a n 33...: 22:. 3 nunnouz unuunaaom uzmm ...H... . 1 (343:3. ...a... .53.... .1.... "5.6%.. ...nn. 20. $0.: 30.: 13. 713.: :: 20.: So... «n3. .3“. 5,375: 37: 1:23; 4 ..om £35392 1T... "coauauaum mwfl). moo. 30. «c0. 20. So. I mno. «cow. «02. Re. 4955.. mmu 35.7w Eu :umlfiMoL £335... .15 .1.—7” .3.“ 3r_cr.< «mg. mqo. .33. 1:. moo. 33.33323 «nu—HZ)... I 20.1.31: So. 52. z a: .3333... .16. .... ... .1; 7 5.5254 ..wflflwmjtowfl:wfltuoamflml .uuviuccm .7; 3:45.12. I «RN. «mi. mmo. «NE. .2326 «3.335va 133:: «Sn. go. 050.: amo. «22. 12511535 .uxuflaccx Jam ”:03:st "you u:,::.:..4an< 2mm mac. woof 03.: «00. go. ... 0. #333933 N _ : m: #2331. ununbmfig uuuncunuu n... “52:60.. L acne—...... xmm m4}. 3.3L .635qu 991.3,. airways... 5:35.13... .1 ocuvclrucca 53f! ... ......cucu un..~:u._u.._ than :u.) :02. ”(Chou ”no. .5 niacin.» Hoyuuou us... un..ut..:...ur.|:~ any? .; ......” Heron. mi) :3: .CnOu 3c. ...v 3.3.. 7:230 1, ... 1.16:3 {J ...»: unnuuuuum: umu .: .n.4.m mo_ro::> “22.3 fluouuum 3:33“qu... Couuaamuuou :u 34....” 353:5} no??? 333“; aouucou use)“ on .nuczuuuwwuoa no“ u: :55 .HHH Hmummso ca wommsomfiw mwmoSuoa%£ “ohms How moapmfism> Houuaoo bum uaouaomowafi coosuon mGOHumHouuoo umvuo oumN m NHQmem< APPENDIX F Zero order correlation coefficients between dependent political effec- tiveness variables and control variables. Control Political Political Political Political Variables Knowledge Activity Efficacy Effectiveness 1. Years Formal Education .195* .083 .094 .192* 2. Weekly Income .031 .048 -.043 —.013 3. Age in 10 year Intervals .142* .153* .032 .169* 4. Sexa .171* —.024 —.133* .007 5. Amount Organizational Membership .163* .465* .019 .333* 6. Amount Interpersonal Contacts (experts and political conversa- tion contacts) —.013 .072 —.117* —.O3O 7. Amount Mass Media (summary index) .211* .122* —.076 .132* 8. Amount Mass Media (press only) .218* .113* .046 .194* 9. Amount Mass Media (television only) .005 .064 .018 .044 * = p .05 or better POSitive correlation indicates characteristic related to males to a greater degree, less by females. Negative correlation indicates characteristic related to females to a greater degree, less by males. 134 Aouv o.~ mm. Nn. no. go. “a. no. q~.- ma. ma. u : ”0.- mm. mo. ofi. m_. mo.- o.~ 59.- 00. mo. ed. No. mo.- co. #0.- 00.; No. ow. no. no. mo.- 0“,. o.fi mo.- #0.- ~o.- 00.- no. mo. «0.- 0H. oo. «o. wo.u 00., 40. No. a.“ mm. mo. me. No. A“. “N. qo.- mo.- we. OH. “H. mg. no. o.# NH. No. m~.- no. «H. no.- o~.- ow. ac. «a. N“. 00.- o.H no. 6H,- mo.n No.- “0.: no.1 0". No. mo. no.- Na.- 0.“ #0. on. no.. m“. «o. od. mm. NA. mm. No. o.H «0.- Ho.- :o. no.. 00.- no.. mo.- 50. em. . o.H No. No. no. Ho. ~¢. mo. afi. oo. “adv Amgv Andy Aoav AnHV Acflv nmfiv ANHV ‘AA~V Aodv Amy Amv Amy on Amv Aqv Any .mmanmflum> uaovsmmww cam uamwdmmovdfl “ohms mo xwuuma GOHumHmuuooumuaH U NHszmm< ANV o.~ Adv AONV nguaua>u uouuwmucruuu Hau«uwaom uufiuuouokm Amdv nfluax clam unson< N >umvnym Amdv uuqk cavax ucuqu< Anflv ufiuw >h unuonm Aomv mJJhm zaxaus+ oumnuo “u,un< An~v mdnwumu‘angn Jumam unwou< A<_Vuuau:oo anaemhuunmuum « u . c‘:huw “adv mao«umNficmwuo unusul Away uuvmx mnmx xuunanuumhu Afiflv adamuunnmx amcowumwwnmwuo xua>auuu~om Ao_v mauz mnuum Auw>uuuw~mm flow mum“ >H sua>uuuoflum wuvnucou amnnmnuuuuuuu va aufl>fluumfiwm ANV muvmx mum: >u«muu>ia flow magmkmnfiux Hucowuanwcwwuo hu«w~m>un AmV unannnou unnamuwnnuunu >uwmuo>ua va AxmucH annunumv N muaav>«uuuwwm amuuuuaom Amv N xuau«uwm Hauwufiflom ANV N au“\«uu« Hmuauudo~ AHV owuuaaoax Hmuauaaom 135 APPENDIX H Summary of analyses of variance used in H8 through H10. negrees of freedom for each F ratio, other than "controls" is approx1mately df = 1/361. The appropriate degrees of freedom for controls is approx- imately 2/360. ’ F RATIO for: Source of Political Political Political Politica15_——- _flypotheses Variation Knowledge Activity Efficacy Effectiveness H8: Diversity X Amount Interpersonal Contacts no controls Diversity IP Contacts .215295 0.466472 0.040176 0.240188 Amount IP Contact .073214 1.813775 4.976778*(—)0.343023 Diversity X Amount .000362 0.088192 . 1.102100 0.136991 controls Diversity lP Contacts 0.005775 0.591997 0.064983 0.166196 Amount IP Contact 1.848184 3.327599 2.456739 0.260992 Diversity X Amount 0.242986 0.078908 0.942443 0.026220 Controls 17.654702*** 10.461480** 3.631266 20.886244*** Organizational ‘ .yenmership no controls Diversity Orgs. 0.859655 0.004629 14.776768*** 0.000000 Amount Orgs. 0.700598 7.090740** 6.482406* 9.755712** Diversity X Amount 1.687874 0.660185 3.499054 1.732862 controls Diversity Orgs. 0.530056 0.030612 5.578308* zmount Orgs. 0.235160 5.888218* 7.746231** Diversity X Amount 2.020793 0.553803. 2.035594 Controls 4.747826? 1.660482F___P*~_fl_hw~ufi““3;§9}63§;___ Mass Media Use "1 no controls Diversity Mass Media 14.658846*** 0.094860 5.037707” 0.?C?T??" Amount Mass Media 0.083101 0.446216 4.7959333 2.650566 Diversity X Amount 0.315308 1.869807 2.949722 3.539622 controls Diversity Mass Media 13.329026*** 0.191#40 7.3605664k Amount Mass Media 0.180878 0.207489 3.478949 Diversity X Amount 0.511197 1.376729 4.547311* Ccntrols 16.041774*** 11 CL5811*** m 19.936135 if H0: Selectivity X Amount Interpersonal Contacts _§xoerts no controls Selectivity 1P Experts 0.565996 0.000364 0.482605 0.000729 Amount 1? Contacts 0.523583_ 0.003276 1.973164 1.134257 Selectivity X Amount 0.585740 0.038223 2.715766 1.291864 controls Selectivity IP Experts 0.048467 0.066467 0-09:431 0.109305 Amount 1P Contacts 1.023198 0.023151 2.105420 2.116350 Selectivity X Amount 1.349212 0.001493 2.647476 2.300817 _£9ntrols 17.044324*** 10.311122fjfi 4.63:82ff” 21.590816¥** Interpersonal Contacts ' 1 Political Convers. Contacts no controls Selectivity Contacts 0.831125 4.739007* 0.150%67 0.823840 Amount 1? Contacts 0.360191 0.071653 1 1.987500 3.138815 Selectivity X Amount 1.302060 6.9355hh*k 0.596691 2.296869 controls Selectivity Contacts 0.290003 4.306566* 0.0256?i 2.039291 {ntOLuit 11’ Coxlt.n ts 1.1351117]. 0.(l?.l43f‘ 2.‘74‘:}Z.§ 1 .73913538 Selectivity X Amount 1.326871 6.9;)i”w”" 0.610332 2.343135 “ -- - 7 “.':..“"~l',-' - ' I.‘ ."‘. " ('11L7_01:‘ 1L.§.‘/\1 L314...” (_l.(l}11l“ :“ (;._,I\; l- 11 I},l‘;" 136 APPENDIX H (continued) Summary of analyses of variance used in H through H10. Degrees of freedom for each F ratio, other than "controls" is approximately df = 1/361. The appropriate degrees of freedom for controls is approx- imately 2/360. F RATIO for: Source of Political Political Political Political 'flypgglugses Variatiort___ Knowlgdgt: Activiln' F1{EEBEY____2¥EQ%9211iE?SS 933313199115 ’ no controls Selectivity Orgs. 3.147577 0.545920 3.514338 2.639337 Amount Orgs. 0.266992 20.632167*** 2.775367 2.473291 Selectivity X Amount 2.767555 0.290909 5.452941* 5.942028* controls Selectivity Orgs. 2.711414 0.702247 2.219522 Amount Orgs. 0.023132 17.796816*** 1.385839 Selectivity X Amount 3.056882 0.362827 6.419354** Controls __ 4.592718* 2.573033 5.202031 __ Mass Media (TV) no controls Selectivity TV 8.854914** 7.561232** 0.289712 10.574095** Amount TV 18.161856*** 6.229902* 1.189749 16.615713*** Selectivity X Amount 0.637285 1.253944 0.431479 1.819136 controls Selectivity TV Amount TV Selectivity X Amount (control; not Cgplitfitlfi) Control” _ flags Medig_(Pressl , no 263173137 Selectivity Press 18.764481*** 0.723861 1.528192 5.715075% Amount Press 0.832698 0.511839 0.138595 1.04/917 Selectivity X Amount 0.872713 0.000000 0.027646 0.316t23 controls Selectivity Press 16.027843*** 4.147092* Amount Press 0.516862 0.680729 Selectivity X Amount 0.442352 0.079817 Controls 10.877314** 12.237910*** n10: Credibility x‘ Mass Media Amount no controls Credibility MM 0.000000 2.325892 24.866874*** 2.914423 Amount Mass Media 0.471308 4.274925 0.284935 1.339293 Credibility X Amount 0.257489 2.303571 0.1008]; 0.131518 controls Credibility HM 0.077272 3.721153 2.330731 Amount Mass Media 0.414462 5.313846% 1.213640 Credibility X Amount 0.237190 2.536155 0.181594 Controls 17.240032**4_11;:31fl3{232 20.093::Z{:: * - p .05 ** = p .01 *** — 1 — p .00 APPENDIX I Map of Sample Frames, Blocks Sampled, and Census Tracts Involved 137 CLEVELRMD P00511171! $171312! 111111111117. 1.111.) of Lannie 1-‘r ramco, ‘ 1310mm mm; nu." , and Census LEGEND - tmcte inlolved '. BLOCKS m omcmnc Suov saggé a some“ meme IMEQvu‘wCyG—Vo How: to chmcmmur 9n NRICH WERE lNTizu'. 9891:) 0mm Nor «urea-1191.58? BLOCKS IN gar-12* uNUm' Firing (1 Without? flames It-‘I‘o comma. . _ tau-1m muck NERI. liEr‘quD ' Etc-cm 1151;» FOR ‘Tm‘uvmc- m--2 u Imemewecs 139.9 d OVN'DMZY 0 menu“. SMVLE Futf‘ NAP: 0.3.CEfl9JS mo /‘ 1411 "101‘: 1111-1 1.31— 1 r) D .' L I f 1' I ‘ \ 1 ‘ ‘ v . . v . ‘ y. l , . _ , . 1 .- 1 . 1-. a l f 1 . ' | . -‘r' nanny-90"“! III-r ’ 1 Lawn mm m‘ panama:Commuting";L Mr J. .2 2w ..=L$ .s‘Ewm INTENT, # Questionnaires Blookfi L1-3 119-1 L2-5 L3-3 14-5 LfinllR L4*19 Lé-ZOR L4-23R L5—2 L5~7R L5—19 L6—3R L6~9 L8-1R — I'" f‘ . . u “\I .- v n 1 r‘ L-u‘ .Lu 1.11"§) N1—11 113-“3 114—1 M7-6 N1“7 N1*9 IQZ‘]6 N3—19 P1—2 1’4- 11 P4-27 P5‘14 R1~5 112-11 R5‘11/12 116~13 mfl‘ .— 34 b1()C1u; (k)ngw1_01-0(1 ——-—.-__- 10 10 12 10 10 11 10 10 11 10 9 15 13 11 10 1") 10 11 10 10 10 11 10 10 ’s L 10 10 12 11 10 10 360 two digiLs = cvusua ‘ock se1cctod frvfl 138 APPENDIX J Sampling Record by Blocks # Original Sanp1e Houuuholds {\JLnbkfir—‘Omb O \J O‘HO‘C‘WUJOO\JLW‘~JC\‘J i1 Pfq>lacxxncnt_ 11013911101418 HO\L~‘U1CD'JI\JU1\OC\®O\ I.“ .Q HDJLWODKOWDCCNFQDU‘b’II'-k"\l\l"‘ agnon< SJmplo. 6 5 7 11 5 b 1 9 l 9 170 195 (462) (53:) trnct; romuinin; d£g115 : blbuk number BEAEQJSVEQRHRHPLACENENT 1x0 NO L1igd1v10 N521fflflt Rdwfll 10 J O 9 3 1 6 2 l 9 0 5 2 4 4 10 1 O 10 O O 11 2 0 9 0 0 5 1 2 7 0 5 1 O 0 10 4 O 10 0 1 10 0 2 3 i U o 4 1 3 1 4 7 3 0 7 1 0 5 1 1 12 O 0 9 1 2 8 1 O 8 O 3 13 1 0 6 2 1 2 6 0 9 O l 4 1 3 2 1 1 9 O 1 13 1 0 13 1 O 258 42 40 Srbjcct FYEfEIS 001-010 011—020 021~032 033—042 043—052 053-063 Oéfl—O73 074—083 084—090 095~104 105-113 11a~128 129-141 142—152 153-162 103*1/0 17/—18o 187—198 199—208 20?—219 220-229 330—239 2&0—250 251-200 261-271 272—281 282*201 292~303 304-313 314~323 325-335 336-246 3&7-356 J.) 57"30b -————____.—__—_... within tract indEcatcd; 139 xmou\muunu:ou Huuuuqflon cemukonoua mmMOoa N Ham manuummm use; oH\~ no .02 ca om: mussewum uso: oH\~ we .oz cw Mm: umuu muudzmmu mumama .oz mam» maumfi:Muu macaumuansn .oz 10M3\muufiuflou mcomuMNfizuwuo no .02 uwauuz I msan> 30H cu vuwcoamn Aquu mo maowumuwzmwuo «wouuaaoa cowuuououn vmmu Hugo” we uumuzou ”dquoqum Imam: ccmuu2L0pa uawu wcauuw> anon” No mums coauuoaouu ham; uau muuusom mo Muunu=0u unwnxu nowuuomoua vmm: unuus unmumuumu museum uauuuuuqv haaduwuawounu unnmmmz mo uwcp \aoaumwuummn UGMGOQEOU HOW UOUOG omfim wHN NUMV QWmSH 0 OthO 0 «male \0 o .-4 I N o oodlo co eaten one: mmmmm mo~.n~ «mo.~ nmm.nm man.am mno.o Hmn.H un unnu:Muw mqo.mH 0.0 «wo.nm Hmo.aN oom.~ mma.~ oqm.om o~N.H qnm.m ¢o¢.mm nmm.H~ nmm.oH ¢m~.¢w an.m mmn.~ cam: muUMucoo :oHummuu>:ao flauauaaom acadho.ou; va umD mflvmx mmmx unread am: cwnmm ucaoa< um: Goamn>oauk ucsos< um: nonmamzoz zaxwws w AAHmn uczafic Mm: :ofiunuwanam xUMHm uczos< uunucou chomuwauwucH unsoa< nwsmumnamz anneaumnwcawno unnofin mfluuz mum: on zufifiwnfiuiuu magmnwnfimz auaoquNfidmmuo >ufi>uuum« um: uoanamaaz Auwpquuwpou mm: >9 mufl>fiuum.wm muomuCno HmcownmnnmuaH zuw>wuuunmm «Hum: man: Auamhm>ws mazmuonaux Mucouumuwamwno muamum>uc mammammw .vMuumeou mm: onuw wanu awkamqma ANV Akmaazm van wouxamcd annouuafimnuou CHI: .mouoom ANV vmuuvnancnum no woman muuamuwa new vmunaaou uon ovofl “n1 umn1m- wuowuaou Junan OBOE I UDHM> 30H muumucou xuaan whoa a wJHm> 3nd wxu0an >uwu mxuoan >uqu Hu>wa mwAumunn w.ucov:ommwn scuw coaum«>wv muwc: wfluum wwwumuua cvwa H034 out mmnn w.ucmvc0LmuH wmfiummpn :wwa @UNOUM N W0 53m mmuouu N we Dam x«Azumuwwww 30H u wDHm> :wHLV ucmfimmuwn mo ~u>wa WhUlMEN UUUNMOU MMMIWEN u UOLHOU ouzmmwz «0 gas: \COuuawkommo c a N m ave: UmGNm m-.~ mo“.m wmm.~H o«_.-fi Noq.~ ¢~N.H mmm.~ mew.~ am~.m own.” ~_m.— aOHumu>wo vuwvawum .mm~.o mam.m nan.» omh.mo «mn.o qvm.o Nom.N~ w~n.~ cmwm muuaucou unanxm sung .w mmumMCno «macaw muwd .» muunuuou uuumnm wunnumfia mmmuu>< .u muumucoo auaunm wuzn“mqo ammuu>< .u muumuuoo uuunw mmwumuum aq:0uuwmuuu .n 014 wuuaocao Huauom wmuumuum awcowuaaaouo .m Anuu1wnux xuumum>ao ucmunmnouv ANV nuumuuou HmcomnwnuwucH xummum>wo 7‘ av mmucm>wuuvwwm Amouuuacm humouwuu auuwuwfloa mu1>auu< amuwuuaom umvwazocm uaoauaaom mapwfiuaw .huflmuo>ww uomuaoo Hmcomuwawmuafi waHmHMano moofiwcfl .zvsuw wHSu SH wmhoaaaw mmusmmma unaccommv can uamvdomwvafl you uaoamusmmoa mo uflaa wan mace .wwawu .m¢0flumfl>ow unawamum .mcmmz .M xHQ2mmm< BOUND BY BURGME’JFP BOOK BN6“r mum: “I1111111111111111ES