Lit-s: «I m". LI.” I u gig; is WW" .III “I -. . W" “n" W I; L“; I! gnaw“! M 'II ‘ g 1"!) '1' .1 LL: 1mm, '1‘,” g IE“ gfi' \l‘r‘h ,... .g 4 _. I244»: mm“ LIL g’l :gggN WL I H mf an» f‘ L H u gua'. l | L” . \fl: .nL 1, (ML?! L n a.‘. ::::: n . I u..-— A - - , ‘- ,1 _ -‘- n .J:”"_ -.-.’-...L. ‘< - ' T I — ‘ .5- KM g I ~— W <$... WWI; 3:”. . L ’ .- ...; ‘n ' :. lg II: '-:,'.;‘l g ‘3 11‘ h . i»!-? a 1'3. ”.2. (T A 7 1‘ '1’3‘!";.;‘4;h ' . .“a‘ ‘ "2*“ ~ v‘ . '. ‘Jf‘ 8". u ,2 “1.31%.“;th :fifi‘f . fl iu’Jgg‘ 1);“ ‘7!) iy'flm‘ jlhflnflf , ; l“:‘—;~ Jingx’gfi .‘g‘h: iii“, £11..“ x3:Lml::5;»..IM.mm‘mm minim: ‘. u ‘ 3.15 “1.16:: ’v '- “in L ‘1'” .1 ~ -1 & LAGJLLJWLLNLLah-Wfiba9m" 'Ic‘t: “#13": « adaduag’glLIL T I}. l' " 2)..)13'EL“ than -' uminWWmmmmmmnmmn1mm « 01730 3698 l ‘ LIBRARY Michigan State University This is to certify that the dissertation entitled The Relationship of News Media Agenda Setting and the Production of U.S. Produced Motion Pictures Featuring Black Actors presented by Walter Thomas Mathews has been accepted towards fulfillment of the requirements for Ph.D. degreein Mass Media ' / . ; . Major professor Date August 13, 1987 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 MSU RETURNING MATERIALS: Place in book drop to LJBRARJES remove this checkout from -_ your record. FINES WT” be charged if book is returned after the date stamped below. . lHAY p 0-: r u. . ., . a: has 4.3. _. . x f ' p 9 I ~ ’ Ii.“ 53% 3' ”J- Inn y .33 f g! ‘f' ‘33: THE RELATIONSHIP OF NEWS MEDIA AGENDA SETTING AND THE PRODUCTION OF U.S. PRODUCED MOTION PICTURES FEATURING BLACK ACTORS BY Walter Thomas Mathews A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Mass Media Ph.D. Program 1987 Copyright by WALTER THOMAS MATHEWS 1987 ABSTRACT THE RELATIONSHIP OF NEWS MEDIA AGENDA SETTING AND THE PRODUCTION OF U.S. PRODUCED MOTION PICTURES FEATURING BLACK ACTORS By Walter Tho-as Mathews This study is an exanination of the relationship between the agenda setting function of the news nedia and notion picture production. The relationship is operationalized using the number of articles indexed on black Americans in the Reader’s Guide to Periogiggl Literature and the nunber of films featuring black actors. Fills featuring black actors are defined as films ”in which blacks have either starred, played feature roles (e.g. supporting characters), or otherwise nade significant contributions" (Ploski and Willie-s, 1983, p. 1189). A conceptual model of the decision naking process in the motion picture industry is developed based on a review of research on the agenda setting function of the news nedia, the notion picture industry, and the literature on blacks in notion pictures. Fron the nodel two hypotheses are derived and tested. The results of the study confirm that there is a strong positive relationship between the news media agenda and the casting of blacks in featured roles in motion pictures. As the number of articles on blacks increases the number of films featuring black actors increases four years latter. Eight film types were also regressed against eight news story types (Economic conditions of blacks, Education of blacks, etc.) Strong relationships were found between particular news story types and the production of particular film types featuring black actors. Action/Adventure, Comedies, and Race Relation films had particularly strong relations with different news story types. The overall results strongly indicate that the media agenda exerts a large degree of influence in the motion picture industry. The implications of the results are discussed. ACKNOWLEDGMENTS One hardly knows where to begin. So many people have assisted me in completing this dissertation through their intellectual insights, their willingness to listen to Iy complaints and problens, and most importantly their continued encouragement and support of me. I must thank each one of my committee members, Dr.s Rain, Baldwin, Muth and Redd for their guidance and assistance in completing this dissertation and for their assistance as I progressed through the program. All of them were always nore than willing to listen to how things were going and offer encouragement. Their help is much appreciated. A special thanks lust also go out to the staff of the Telecommunications Department. They were always ready to assist me in anyway possible. I must single out Ann Alchin, the program secretary. Ann, for listening to all of my little problems, helping me clear up small problems over the phone, and in general being a good friend, thanks. I must also thank Dr. Richard Houang for being an outstanding professor and taking the time to spark an interest in quantitative aethods. My family has been aost supportive of this effort. My iv mother and sister have offered encouragement and belief in me when I really needed it. To both of them, this dissertation is a result of your efforts. My wife, Christine, has not only supported me fully in completing this dissertation, but allowed me to spend the best part of the summer in East Lansing. Without your constant encouragement, belief in me, and your care for me, I could not have completed. Thank you always. Finally, there are numerous people that have offered ae support and encouragement throughout my life. Teachers, who took an interest in me, friends, relatives, and fellow students have all played an important part in my development. The only thing I can say is that this dissertation would not have been possible without you. TABLE OF CONTENTS CHAPTER I INTRODUCTION....... ....... . ........... . 1 CHAPTER II LITERATURE REVIEW. .......... . .......... 6 Decision Making in the Motion Picture Industry.. .................. 6 Uncertainty in the Motion Picture Industry...... ..................... 7 Uncertainty and the Audience Image ........... 9 Major Players in the Production of Motion Pictures ..... . ....... 11 The Casting of Blacks in Featured Roles........ .................... 13 The Motion Picture Industry’s Uncertainty and the News Media ............ 17 The Develop-eat of Agenda Setting Research .................. 18 Agenda Setting Theory........... ........... 19 Results of Agenda Setting Research ......... 22 Explanation of the Results of Agenda Setting Studies.. ................ 28 Model of Motion Picture Decision Making... 30 CHAPTER III METHODOLOGY ...................... 37 Rationale for Time Frame Selected (1950 to 1980) ................ 37 The Variables............................ 38 Operationalization of Variables ........... 39 The Data Sources and Intervening Variables .................. 40 vi Coding and Classification of Data ......... Classification of News Media Articles. ................... Classification of Films Featuring Black Actors ........... The Hypotheses and Statistical Testing ......... ...... ................. CHAPTER IV RESULTS............ ..... .. ......... Descriptive Analysis of the Number Films Featuring Black Actors.... ....... Descriptive Analysis of the Number Articles indexed on Black Americans.... Comparing the Distributions of the Variables........ ..... ...... ..... ...... Comparing the Distributions over Tine ..... Statistical Tests of the Hypotheses ....... Hypothesis 1: ........................ Regression of Total Number of Fills Featuring Black Actors against Number of Articles at Lagged Time Periods.. .............. Regression of Percentage of Films Featuring Black Actors against Number of Articles at Lagged Time Periods ........ . ...... Comparison of the two Regression Equations ...... ... ................ Hypotheai. 2:00.00000000000000000.... Regression Analysis of Film Type and Media Story Type.......... Results of Regression of Total Number of Films Featuring Black Actors ....... .. .............. vii 43 43 45 46 52 52 54 56 57 60 60 61 63 64 66 69 69 Results of Regression of Action/Adventure Film Type......... Results of Regression of Biography... Results of Regression of Comedies.... Results of Regression on Documentaries. ..... ..... .......... Results of Regression of Drama ....... Results of Regression of Historical Themes ..................... ........ Regression of Musical Films.......... Regression of Race Relation Films.... Overall Results of Hypothesis 2: ..... Results of Both Hypotheses.. ......... CHAPTER V SUMMARY AND CONCLUSIONS ............ Review of Results ......................... Interpretation of Results ................. Trigger Variables and Contextual Variables ................... Interpretation of Hypothesis 1: ........... Interpretation of Hypothesis 2: ........... Interpretation of Results of Total Number of Films Featuring Black Actors.. ....... .... ....... .. Interpretation Results of Action/Adventure Film Type ........ Interpretation of Results of Biography Film Type.......... ..... Interpretation of Results of 0°.edi°.sssssssssssssssss ......... Interpretation of Results of Documentary Film Type... .......... viii 70 70 71 71 72 72 73 73 74 74 75 75 76 76 78 80 80 81 83 83 84 Interpretation of Results of Drama Film Type ................ Interpretation of Results of Historical Themes ................. Interpretation of Results of Musical Film Types ................. Interpretation of Results of Race Relations Film Types ......... General Interpretation of Results.. ....... Limitations of the Study .................. Directions for Further Research...... ..... APPENDIX A Number of Films Featuring Black Actors and Articles indexed on Blacks in the United States by Year ....... . ...... APPENDIX B Titles of Films Featuring Black Actors 1950-1980 .................. ... REFERENCES. ..... O O O O O O O ........ O OOOOOOOOOOOOOO ix 85 86 87 87 88 92 93 97 98 LIST OF TABLES Table 1 Data Sources for Variables ............. . .......... 40 Table 2 Regression of Films on Media Agenda ............... 49 Table 3 Regression of Film Type against News story type... 50 Table 4 Frequency Distribution of Films Featuring Black Actors......... ............... . 53 Table 5 Mean, Standard Deviation, and Range of Films ..... 53 Table 6 Frequency Distribution of Films Featuring Black Actors.... ..................... 55 Table 7 Mean, Standard Deviation, and Range of articles. 56 Table 8 Coefficient of Variation ....................... 57 Table 9 Analysis of Variance of Regression of Total Number of Films Featuring Black Actors. ....... 62 Table 10 Regression Equation for Number of films ......... 62 Table 11 Analysis of Variance of Regression of Percent of Films Featuring Black Actors.............. 63 Table 12 Regression Equation for Percent of Number of Films .............. ..... . ........... 64 Table 13 Variables used in Regression of Film Type against Article Type........... ..... .... ....... 66 Table 14 Descriptive Analysis of Film Types ................ 67 Table 15 Descriptive Analysis of News Story Type .......... 68 Table 16 Film Types and Trigger News Story Types .......... 89 Table 17 Number of Films Featuring Black Actors and Articles indexed on Blacks in the United States by Year .................. . .............. 97 Table 18 Title of Films Featuring Black Actors 1950-1980... 98 xi List of Figures Graph 1 Films featuring black actors ..................... 58a Graph 2 Articles indexed on black Americans ............. 58a xii CHAPTER I INTRODUCTION (At its most general level ,this study is an examination air-.014 ."I,v1C (“xi/6 ’ /{fl"! 4//6/ ’rC/‘I (I (If) of the relationship>between the news media<find the entertainment media?) It is therefore subsumed under the general study of media and media influences.(/On a more \ specific level, this study will examine the relationship / 7 “--_..~._..___ between the media agenda setting function of the news media I- Rmflggfl fifl,11“_, and motion picture production. At its lowest level of /"""‘\c "firm-”H“ “““““ 1 -- .. inquiry, this study poses a simple question; Nhat is the [-..- relatio_n of the news media agenda and the number of motion pictures with black actors as cast members. iIt is different 11”” V ..“”_ gandcimportant_ in that_ it will be the first to quantify the relationship between agenda setting and the production of motion pictures. This study will examine only one of the various motion picture elements, the selection of the motion picture cast and relate casting to the media agenda. The casting of a notion picture may also reflect theme and plot elements. {(Up to this point, the agenda Overviewfof Study: setting research has been limited largely to the political arena./)On the other hand, the research on blacks)in motion pictures has largely been historical in nature. (L‘his study proposes to extend agenda setting research to the l /"" Aztr dl ’llvdf$- . \ entertainsent arena by determining the relationship between> the media agenda and the production of films featuring black actors. (;The study is unique in perspective. It hypothesizes J J {I \‘J A ' {If/‘1’"CI, ' t’,.t‘rl ‘t.’ (a: that the Jew; media set the agend_a for t_he film media. ‘It also proposes that the news media may set the agenda for the entertainment media. This perspective alone separates this study from previous ones. The measurement of the media agenda is relatively easy since there are a number of studies using different methodologies to determine the media agenda. Ordinarily, the categorization of a particular motion picture would present a difficult conceptual problem. Each film contains a variety of themes, plots, cast characteristics and other variables that make categorizing them a difficult if not impossible task. The use of films with black actors as cast members eases this problem. The use of black actors in a motion picture is readily observable and subject to verification in a number of ways. /The first step in this study, then is model construction. This model arises from a careful review of a number of factors related to motion picture production. First, all motion picturendecisions are made under fl 1) {Pix/“v” ‘r [J (“P11 conditions of extreme uncertainty. The operators of the m—...~.,_ film industry are not exactly sure what their competition will be when the motion picture is released (Jowett, p. 19). Additionally, they are not sure of what audience (if any) the motion picture will attract.<:As a result of this uncertainty, the industry tends to rely on successful ”formulas" for notion pictures. This tendency to imitate is a sign of the uncertainty of the environment. Because of this imitation, the industry is very slow to change or pick up new themes or develop new genres€> /’ ‘xSecondly, under uncertain conditions, motion picture operators tend to conceptualize the desired audience for !(."'h their films in certain ways. Jowett (p: 29) feels that producers are forced to develop a mental image of the desired or anticipated audience to ease their uncertainty as they make decisions. Gans (1957, p. 316) earlier labeled J. l " I' ' r f , , f. SPBCifiCOIIY. thi' study Will quantifygthe relationship :9 ” between the agenda setting function of the news media and the production of mass entertainment motion pictures featuring black actors as cast members. This agentification, however, cannot be done outside of sose ',‘{’,L V,‘/{{"{,"l"‘c tanding of ./Each public is composed of a heterogeneous collection of individuals that are attracted to certain film elements. Throughout the production of the film, the producer makes decisions based on what will appeal to the various publics of his audience image. Communication theorists have called this type of behavior anticipatory feedback (Jowett, p. 29) and it is particularly necessary under conditions of high uncertaintyi} This study will test the influence of the agenda setting function of the news media on the production of motion pictures by measuring the relation of the media agenda and the number of motion pictures released. The Research Probleg: The research problem, then is to determine the relationship between the media agenda and motion picture production. The operationalization of the media agenda will be simply the number of news articles about blacks as derived from the Readers Guide to Periodical Literature (H. N. Nilson Company). Readers Guide indexes major periodicals and releases indexes both quarterly and yearly. For this study, the yearly indexes were used. Motion Picture production elements will be operationalized as the number of films produced with black actors as major cast members. Implications of the Study: The results of this study will have implications in a number of areas. First of all ‘as an agenda setting study, it will add to the knowledge of the nature and effect of the agenda setting function of the news media. The research on the agenda setting function of the news media is in a period of defining the limits of the function. This study will further define those limits. Secondly, if the test of the hypotheses are confirmed, the study will confirm the conceptual model from which they are derived. If the conceptual model is confirmed, the system of decision-making in the motion picture industry will be better understood. Since the decisions in the industry are made under uncertain conditions, the results of the study also will have implications on how decisions are made under uncertain conditions. Finally, the study will add to the knowledge base of the relationship between different media. Although others have observed that other media affect the motion picture industry, very little has been written about the relationship between the news media and the motion picture industry. This study will attempt to quantify that relationship. In sum, the results of this study will have iaplications in at least three different areas. The ’knowledge in agenda setting, media decision-making, and the influence of one media on another will all be increased. CHAPTER II LITERATURE REVIEW It would be impossible to gain any understanding of the relationship between the news media, motion pictures, and the casting of black actors without a background review of the literature. Although, there are a variety of studies and critiques in each area, none specifically links all three together. In order to create a contextual framework for analysis of the relationship between the news media agenda, motion picture production, and the casting of black actors, a review of the literature in each of the three areas will be undertaken. Decigion Making in thgiMotiog Picture Industry The production of a motion picture is the consequence of a series of decisions. In order to gain some understanding of the decision process in the industry, one must examine the environment in which these decisions are made. In sum, then, the motion picture producers are faced with uncertainty on two levels. First of all there is no reliable manner to predict a motion picture’s success. Secondly, the producers do not have an idea of what the competition to their film will be when it is finally released. Uncertainty and the Audience Igggg The uncertainty that surrounds the decision making process in the motion picture industry results primarily from the nature of the industry. Motion picture producers must base decisions on imperfect information about the outcomes of their decisions.\:Gans (p. 316), suggest that under these uncertain conditions, motion picture producers develop an "image” of their desired audience. In order to evaluate the decisions that they make, motion picture producers constantly measure the consequences of their decisions against this imagined audiencéy 10 : Since the industry is filled with a great deal of uncertainty, producers try to combine motion picture elements in a way that will attract the largest possible audience.:>The producers realize, however, that although the audience may look like a mass from the box office rentals, it is actually made up of audience members that came to that chose that movie with different predispositions. Additionally, the producer must also be aware that at least initially, his film will compete with others on the elements of the film that have promotional value. The producer fashions his Audience image as being composed of a number of different "publics.” Each "public’s" wants and needs is what the producer tries to satisfy in the film (Gans, p. 316). In turn, the satisfaction of these wants and needs can be promoted to these various audiences. The Audience Image, then, is an artificial construct that is used to measure the potential success of the motion picture in the uncertainty of the production stage of the motion picture industry. As might be expected when only imperfect information is available on the outcome of a decision is available, motion picture makers tend to use whatever information is available (Jowett, 1980, p. 30). It is proposed in this study that ,the Audience Image is based largely on the information available from the news nedia. 11 Major Players in the Production of Motion Pigtgres In addition to being characterized by a high degree of uncertainty in the production of a film, the industry is also characterized by a high degree of administrative diversity. From the very beginning, film has been described as a collective endeavor. David Picker an official at United Artists flatly states that there is no ready formula explaining how a film is produced (Picker, p. 25).(/Monaco (1981) also observes that there are a variety of ways that a motion picture can advance from idea to production. Monaco’s major players are the producer, the director, and the star of the film. These major players have different degrees of power depending on their financial or market value to the film}; Although the idea for a film can arise from a number of sources (including a screenwriter), the Motion Picture Producer is generally the person that makes major decisions about all areas of the film including the initial evaluation of the screenplay and casting. The producer represents the financial interests of the company producing the film. Ultimately, it is he who structures the film elements that will serve to attract the desired audience. The Producer may be employed either by a studio or be an independent producer. A studio producer works for a major studio and is responsible for creating films within the prior agreements that the studio has made. For example, major film stars now 12 have contracts that extend for a number of films. The studio producer is therefore mainly an administrator and delegates to others major decisions in the creative area (Lewis, p. 13). An independent producer, on the other hand, has the freedom to mix and match film elements at will. The independent producer however, must sell his idea for a film to financial backers including the major studios. In both cases, the producer, is the first major player in the creation of a film. He decides what screenplays are likely to produce successful films. It is he who contracts the talent and the financial resources to complete the film. In order to complete the film the producer may delegate his authority to other players since they may be critical in creating a successful film. In attempting to create a financially successful film the producer tries to interest major players with know box office appeal. Two other categories of film workers, film stars and film directors may have this box office appeal. Depending on the popularity of the film star, the star may have considerable power in deciding many aspects of the film. Information => Attitudes => Behavior (Shaw and McCombs, p. 4) The result of this orientation of the research was the Moderate Effects model of mass communication (Servin, and Tankard, p. 249). The focus of this model was not on what the mass media does to the audience, but rather on how the audience members use the media. This model views the media as informing rather than persuading the audience. As a result, this model directs the researcher’s attention to the awareness and information steps in the mass communication flow. Two main sets of theories are derived from this model, the Uses and Gratifications model, and the Agenda Setting Model of mass communications (Chaffee and Hochliemer, p. 78). The most relevant of the theories for this study is the Agenda Setting Theory. McConbs and Shaw were most responsible for the current research on Agenda Setting Theory (Servin and Tankard, p. 253). Agenda setting theory simply states that while the media does not tell people what to think but rather what to think about (Servin and Tankard, p. 253, Shaw and McCombs, 21 p. 7). Beyond this simple definition, Shaw and McCombs (p. 7) further refine the concept by saying that Agenda Setting is not only the selection of events, but the importance assigned to them in the pseudoenvironment in which decisions are made. The theoretical relationship is simple. The news media, by its selection of issues sets an agenda of important topics. This ordering of topics in the media consequently affects the ordering of important topics in the public’s mind. Agenda Setting studies are concerned with the rankings of issues from three major sources. The first source is of course the media. The media agenda is generally measured by a content analysis of various media or media indexes. The second source of rankings of issues is the public. The public agenda is measured by polling members of the public on the issues that they rank as important. The final source of rankings of issues is reality. The reality agenda is more difficult to measure. Any ranking of issues implies that a person or people are ranking the issues. The reality agenda is generally measured through counts of events like unemployment levels or crime rates. These three types of agendas are then either analyzed individually, compared with each other, or compared with other variables. The results are then used to make inferences about not only general relationships but causal ‘\_/ 22 relationship between variables. Results of Agenda Setting Research The seminal work in agenda setting research was done by Shaw and McCombs (1972). Their 1968 Charlotte study was centered on a political election, much like Lazerfivld’s Erie County study of 1948. The Charlotte study, however did not concentrate on media effect on voting behavior, but rather on media effect on public awareness of the issues. Graphically, the differences between the two studies can be presented as follows: Awareness----> Information---> Attitude---> Behavior Shaw and McCombs Lazerfield (1968) (1948) While Lazerfield had found no significant effect of media on voting behavior, Shaw and McCombs did find a media effect on awareness<§3dwfeund someignificantxeffectzofmmedia. oa—votingmbehavior,.Shaw.and-McCombsadidwfindmawmediaweffect~- ‘en—ewareness;> The effect of media, then, seemed much further up the mass communication flow model than Lazerfield and most other scholars had analyzed. Specifically, Shaw and McCombs interviewed Charlotte residents in June and October. The residents were asked to outline the key issues as they saw them regardless of what the candidates said (Shaw and McCombs, 1972, p. 178). From these responses two lists of issues for June and October 23 were ranked by public importance. The media agenda of issues was determined by performing a content analysis of five newspapers on 15 categories. In order to establish the direction of causality, e.g. whether the media agenda influence the public agenda, or via versa, cross lagged correlations were performed as part of the analysis. The graph below presents their results. June October .43 Newspa er ------------------------------------------- Newspaper - Ir .46 i : 10 v v Voters ----------------------------------------------- Voters (Shaw and McCombs, 1977) The diagonal correlations between newspaper and voters were used by Shaw and McCombs to support the proposition that the public agenda in October was caused to some extent by the media agenda in June. This conclusion follows since the correlation between newspapers in June and voters in October (.51) is much higher than the diagonal correlation (.19) between the voters in June and newspapers in Octoberff Since the original work of Shaw and McCombs, other researchers have also investigated the agenda setting function of media. As a result, agenda setting theory has been further refined. 24 Shaw and McCombs (1977, p. 24-25), after finding evidence of an agenda setting effect in 1972, began to define media and audience elements that affect agenda setting. Their research pointed to four elements that would affect agenda setting. The elements were: 0 Media Type of Story 0 Media Emphasis given the story 0 Audience interest/knowledge 0 Audience orientation The media type of story refers to whether the story is news, a feature, entertainment, or an advertisement. The Media Emphasis given the story refers to the amount of time or space devoted to the story in the media. The Audience elements revolve around the audience’s need for information about the issue. The audience interest/knowledge is defined as the intrinsic interest of the audience in an issue. The need for orientation refers to the audience’s need for information related to a particular individual’s need. fl Funkhouser (1973), found a strong relationship between press coverage of national issues and public opinion rank. He used the Reader’s Guide to Periodical Literature as a measure of the media agenda, and Gallop poll results to measure public opinion. He used fourteen issues in the 25 study, and eight of the fourteen were ranked closely in media coverage and in public opinion. The issues ranked by both the media and public were Vietnam, Race Relations, Campus Unrest, Inflation, Crime, Drugs, Poverty, and Sex (declining morality) The issues were ranked for each year according to the number of articles in the gggderflg Guide and the percentage of the public responding that the issue was important. The two resulting agendas were remarkably similar. Funkhouser (p. 66) found a Rank-order correlation of .78 for the two rankings. The results suggested that the amount of media attention given an issue strongly influences its visibility to the public. E In investigating whether or not there was a strong relationship between the media and public agendas, McCombs and Shaw (1972) found strong agreement between the media and public agendas during a political campaign. The findings, based on responses from 100 non-decided voters showed that there was much greater agreement between their listing of important issues and the media agenda (daily newspapers in particular) than the candidate’s agendas. C. Williams and Semlak (1978) found no significant relationship between the media and the public agendas for Central Illinois residents during the New Hampshire Primary Campaign. The study sought to measure the effects of the 'media on the public agenda during a presidential primary campaign in New Hampshire. The researchers chose Central 26 Illinois to measure the media agenda’s effect in a campaign setting where the public would be exposed only to political news rather than political advertising. Although there was no general media agenda setting effect, the study did find that there was a significant media agenda setting effect for those members of the public who were involved politically. The contradictory results of the (22%; studies can be explained in part by differences in the orientation of each study. Funkhouser studied the effect of the agenda setting function over a longer period of time (one year) than the other two. McCombs and Shaw studied undecided voters in the midst of a political campaign where their need for orientation was great. Nilliams and Semlak, though conducting their study during a political campaign, none of the public in Central Illinois was being pressed to make a political decision in the near future. In short, then, the media agenda setting theory is not an axiom, but rather is what Merton (1957, p. 5) calls a theory of the middle range. That is it is situated between the day to day working hypothesis and the all embracing master laws of the physical sciences. The media’s agenda setting effect, as show in the above studies is limited by environmental and population conditions.r%/fl (:HBehr and Iyengar (1985) used real world indicators as a measure of the environment or reality to compare with the media agenda and the public agenda. For real world 27 indicators, Behr and Iyengar used the unemployment index to aeasurefiunemployment, the consumer price index to measure inflation, and oil supplies to measure energy. The media agenda was measured by the number of television news stories on the network news shows. The public agenda on the three issues was measured by public opinion polls. ,In the energy and inflation categories, the data pointed to a direct relationship between media and public agendas. Nith unemployment, the public seemed to be influenced more directly be economic conditions{:> Behr and Iyengar’s results show that the media agenda setting function is affected not only by the particular. population studied, but also the type of issue. // Protess et. a1. (1985) studied the agenda getting effects of a newspaper series on rape. They found a limited effect on the public with increased awareness of crime in general but not rape in particular. Policy makers, however, did show some greater awareness of the need for more effective law enforcement for rape crimes. No public policy, however, was initiated as a result of the series. After the series on rape, however, the newspapers in the city presented more rape and rape related stories than 28 before the series. The Protess study does reveal that the agenda setting function does effect different parts of the population differently. In this case, policy makers and the media itself were affected to a greater degree than the general population. Nhat the literature on the agenda setting function of the media points out is that the agenda setting function is affected by a number of conditions both in the media and in the audience. Other scholars have theorized on the reasons for the difference in agenda setting effects. Explanation of the Results of Agenda Setting Congitiogg Zucker states that the problem with agenda setting research is that it assumes that the media will have the same degree of influence over all sorts of issues (Zucker, p. 227). As the review above shows, this is clearly not the case. Zucker sees the assumption of universal media influence as a flaw in the agenda setting research. He goes on to specify the conditions in which media will have a high degree of influence and other conditions where media influence will be low. Zucker’s first condition (p. 227) is the issue’s obtrusiveness, that is how much direct experience people have with the issue in a given area. Where there is high obtrusiveness with the issue, the media will have little 29 influence. On the other hand, where there is low obtrusiveness, media will have high influence. Zucker’s second condition is duration. He defines duration as the amount of time an issue has been prominent in the news. When an issue has been in the news for a considerable period, people make up their minds about the issue and are less influenced by the media. Conversely, when an issue first reaches the news media, people will be highly influenced by the news media. //’ ((Adoni and Mane (1984), posit that the reason for the variability of agenda setting is the construction of social reality. That is, reality is socially defined and created \ \ through social interactions, either real or symbolic./ (:They continue to define three types of reality, objective, subjective and symbolic.Q)Objective reality is the world as experienced by the individual.C§Symbolic reality is the expression of objective reality through symbols like media content. :Subject reality fuses the objective reality and symbolic reality into the person’s consciousnessi> Their model also includes a distance factor for each type of reality. The distance factor measures how close a person is from the particular issue. Their model is presented below: Objective Reality Close ------------- Remote Symbolic Reality Subjective Reality Close ------------- Remote Close ------------- Remote (Adoni and Mane p. 327) 30 The distance factor (Close----Remote) indicates the closeness of experience the person has with the issue in each type of reality. For example, the two step flow model of communication is part of the symbolic reality but opinion leaders would have close experience with the media, while others would have remote experience with the media. In sum then, the media exercise different degrees of influence depending on the degree of involvement that people have with a particular issue. Zucker labels it obtrusiveness, while Adoni and Mane label it the distance factor. Zucker also used the concept of duration of an issue as an explanation of decreasing media influence. Adoni and Mane put the entire process in a social context realizing that media influence occurs within a socially defined reality.<{Although the effect of agenda setting function of the media varies with different conditions, there is little doubt that under a variety of circumstances the media do influence decisions. , Model of Motion Pictpre Decigion Mgking This study will propose that the agenda setting function of the media exercises a degree on control among a particular group of decision makers, motion picture producers. Most other studies except for Protess et. al. have concentrated on the agenda setting effects of the media 31 on the population at large (Funkhouser, Shaw and McCombs, Williams and Semlak). These general studies have then stratified the population according to various dimensions like undecidedness in a political campaign. Although other people are extremely important in the production of a motion picture, it is the producer who acts as gatekeeper and ultimately convinces others to produce the film. While the screenwriter is important, this role, has no power to decide whether or not the script is produced as a film. Secondly, the idea for a film may just as easily arise in the producer’s mind and he may have a script written from the idea (Monaco p. 27). It is proposed that motion picture producers, because of the uncertain conditions under which they must make decisions are influenced by the media agenda. When they are evaluating a film, they have no real idea of the audience tastes when the film is finally released. They do not have the luxury of a serial medium like broadcast television where characters can be developed over time and a program can find its audience through different programming strategies. For the motion picture producer, each film is a one shot effort that entails considerable risk. {/Under these conditions, motion picture producers rely on an imaginary image of their desired audience and choose both scripts and cast characteristics (Gans, p. 316). This audience image is composed of various ”publics" that the 32 producer hopes will pay to see the film. In order to attract this imagined audience, the producer tries to combine elements in his film that will satisfy various wants and needs of his imagined audience (Gans, p. 316).>’ (flt is proposed that both screenwriters and producers are effected by the iéaié agenda. The producer in particular, must rely on the imaginary image of his desired audience in making decisions all along the production .\ process.“ This study will examine the effect of the media agenda on the production of motion pictures by measuring the relationship between the frequency of periodical articles concerning black people and the production of films 35 featuring black actors. It is believed that there is a direct relationship between the priority given blacks in the media and the resulting number of motion pictures featuring black actors produced. This belief is based on three sets of facts. First of all, motion picture making is filled with risks. There is considerable uncertainty in the decision making process. Producers lack immediate feedback from their audience and have little idea of their competition as the film is being produced. Under these uncertain conditions, the producers are likely to rely on the media as the only available source of information in making their decisions. Secondiy,<éction picture firms or producers look for popular topics that will attract an audience.§>Jowett (p. 97) has documented that at least one motion picture firm surveys the public using newspaper articles to determine interesting topics that are likely to draw the public into the theater. No doubt, others at least informally use the media to get some feeling for popular trends. Finally, the media agenda research suggests that at least under certain conditions, the media act to create an awareness of different issues but also assign priorities to those topics. Although, one might expect that the agenda . setting function of the media would be operative only with high consumers of media, as Zucker (1978) points out, much of the research indicates that media affects both high and low consumers low consumers media set the leaders (high others. 36 of media. The mechanism for this effect among is the two step flow of communication. The topics of public discourse through opinion media consumers) who discuss issues with CHAPTER III METHODOLOGY The model developed in the last section is a representation of the process of motion picture production. It is however, only theoretical in nature. In order to test whether the posited relationships exists empirically, it is necessary to operationalize the media agenda and motion picture production. The media agenda is measured by the number of magazine articles on blacks and race issues. This is related to the number of films featuring black actors. Rationale for Tige Frgpe Selected (1950 to 1980) The time frame selected for this analysis is the period from 1950 to 1980. This period was selected for three reasons. 1950 was chosen as the beginning date because it was the beginning of a race consciousness, at least for whites in the U.S. Furthermore, Norld Nar II had ended and the motion picture industry had turned away from war pictures back to more conventional entertainment. Lastly, in 1948, President Truman issued Executive order 9981 directing "equality of treatment and opportunity in the U.S. Armed Services and created the Fair Practices Board of 37 38 the Civil Service Commission to rule on complaints of discrimination in government employment (Ploski and Williams). 1980 was chosen as the cutoff year because after 1980 other distribution means (cable and home video players) may drastically affect the motion picture industry. Although this period represents three decades of tremendous social change in the United States, it was felt that the period was sufficiently long and diverse enough to avoid short term spurious relationships between variables. Additionally, this period the marks passage of major civil rights legislation and court decisions. During this period although race relations has been an important issue at times, it was not the only major issue. The Vietnam war, campus unrest, and pollution were all among the issues rated as the most important problem facing America (Funkhouser, 1973). This variation in importance on the national agenda clearly marks the period as suitable for study. The Variables The model proposed in the previous section leads to the relationships that will be examined in this study. Concepts must first be defined and their relationships specified. In its broadest scope, this study proposes that there is a relationship between the domestic news media agenda and the production of domestic motion pictures. In short this 39 study posits that these two variables are related as: Domestic News Media Agenda ====> Domestic Motion Picture Production For purposes of this study we sought to determine the relationship between the domestic production of films featuring black actors and the number of articles about blacks in the news media. This relationship between the two variables is expressed as: Number of news =====> Domestic Production of articles on blacks Motion pictures featuring black actors This further refinement of the two variables avoids a multitude of categorization and measurement problems while maintaining the fundamental relationship between the media agenda and the production of motion pictures. The relationship, then, should be generalizable to other refinements of the general relationship. Opergtionaliggtion of Variables The two variables were operationalized as follows. The number of news articles on blacks was found using the Rggger’s Guide to Periodical Literature for the years of 1946 to 1980. The number of films featuring black actors was found using Bogle (1973), Ploski and Williams (1983), Cripps (1975), and Murray (1973). Where there was some disagreement about the release year of a film, the 40 Anniversary Issues of Variety were used as the ultimate arbitrator. The number of domestic films released each year was found using Steinberg (1979). The table below presents the data sources for each variable. Table 1 Data Sources for Variables Variable Source Number of news Reader’s Guide articles on blacks and Race Relations (1946 to 1980) Number of films Bogle featuring black Cripps (1975) actors ’ Murray (1950 to 1980 ) Ploski and Nilliams Variety Number Domestic Steinberg films (1950 to 1978) The next section will review the data sources and discuss possible intervening variables. The Datg Sourcep_gnd Interveningllarigbleg No data source is entirely reliable. For this reason, the data used in this study were carefully checked against other sources where possible. None of the data in the sources had been collected with this particular study in mind. Each source, then, must be evaluated by its rules of data collection and categorization and the possibility of 41 intervening variables. Newg Articles about Black Americans: The source for these data is the Egggerlplgpige to Periodicgl Literatgyg. Unfortunately, over the period of the study, 1) the number of magazines indexed, and 2) the number of black oriented black magazines indexed changed. The total number of magazines indexed, could conceivably bias the number of articles indexed since a greater number of indexed magazines implies a greater number of articles indexed. Over the period, the Reader’s Guide has indexed from 84 (1957 to 1959) to 188 magazines (1978 to 1979). Although this seems to be a large variation, it seems to represent changes in the magazine industry. As a major index, the Reader’s Epigg only indexes established magazines that are in general national circulation. Additionally, each magazine is indexed over its life span. The Reader’g Guide, then, indexes only the magazines and periodicals that are established in the national consciousness and impact that consciousness. The number of black oriented magazines indexed, has ranged from 1 to 5 over the period from 1946 to 1980. Once again, there is a possibility that a greater number of black oriented magazines implies a greater number of articles indexed on blacks. These magazines, also must be in general circulation and have been established as major national magazines. Although the number of black oriented magazines 42 has increased, each has established a niche in the national market. The nupber of Filps Featpring_81ack Actors The number of films featuring black actors is taken from a compilation from an assortment of sources. Bogle (1973), Ploski and Williams (1985), and Variety provided the list of films featuring black actors. This list was then cross-checked in Murray (1975) and Cripps (1975). Only in rare circumstances was there disagreement among the sources. The major area of disagreement was on the release year of the film. In these cases, Variety was used as the ultimate arbitrator. Although each source had slightly different ways of defining the films, the resulting listings of films agreed. Bogle(p. vii) listed his films through his own research efforts and by "going through albums of pictures and clippings at the Schomburg Library, through personal scrapbooks of black personalities at their homes, through reams of old yellowed newspaper movie reviews." Ploski and Williams (p. 1189) defined films featuring black actors as films "in which blacks have either starred, played feature roles (e.g. supporting characters), or otherwise made significant contributions". Dgpestic F11! Prodggtion The number of films released domestically each year was found in Steinberg (1979). The number of films was found for the years 1950 to 1978. 43 Domestic film production includes not only the major motion picture production firms but also independent productions. During the time frame analyzed the motion picture production industry has changed. Steinberg (1979, p. 42) speaking of the change in the industry over the period, states that although there had been a decline in the number of motion pictures produced by the majors in the in the 50s and 603, there has been a slight surge in domestic production in the 70s. This surge has been the result of an increase in the "market for porno flicks and to the increase in small, independent production (Steinberg, 1979, p. 42). The number of films featuring black actors is a subset of the number of domestic films produced. Films featuring black actors fall into all three film production categories; major studios, independent production, and pornographic films. Coding and Classification of Data It was felt that there were additional relationships between motion picture production and the news media agenda. The data were consequently coded as to the type of motion picture and the type of news article. Classification of Newa Media Articles The number of articles indexed under ”blacks or Negroes in the United States" was counted in the Reader’s Guide. This category, however, is also subdivided into a number of 44 sub-categories. The categories used in this study and consistent throughout the period were: 0 Blacks or Negroes in the United States: This category includes all index articles pertaining to blacks. The following topics are sub-categories of this category. 0 Economic Conditions: This category includes articles on the general economic conditions of black Americans. It includes articles on the average income of blacks and black consumer behavior. 0 Education: This category includes articles on educational achievement of blacks as well as articles on desegregation. 0 Employment: This category includes articles on the general employment or unemployment levels of blacks and the employment of blacks in particular industries. 0 Equal Rights/Civil Rights: This category includes articles on civil rights protests, legislative actions, and public policy debates. 0 Housing: This category includes articles on general housing conditions of blacks and legislative actions on housing opportunities. 0 Occupations: This category is generally either profiles of individual blacks in unusual occupations or articles on career opportunities aimed at blacks. 0 Politics and Suffrage: This category includes both protest over voting rights and political activities of blacks. 0 Social Conditions: This category includes articles on the general living conditions blacks. Race Relationa: Since much of the news about black Americans in this period has revolved around Race Relations, data from this category were also collected for this study. 45 Race Relations is a separate category in the index. The Race Relations category includes both articles on racial theories and the cooperation between races. Over the course of the study, the Reader’s Guide used both the categories of Race Relations and Interracial Cooperation. At no time, however were both titles included in the index. The change in name reflects only a change in title rather than a fundamental change in the type of article indexed. Qlaaaification of Filaa Featuring Black Actors As reported above, the number of films featuring black actors was taken compiled from a number of sources and verified by Variety. Each film was then categorized into one of ten categories. The Categorization was performed using Ploski and Williams film descriptions. The categories used for this analysis are: 0 action/Adventure These films include detective stories, war, horror, westerns, and other action oriented films. 0 Biography These are films that portray a person’s life and are at least loosely based on fact. 0 Comedies These films are those whose primary focus is humor. This type includes situation comedies and films of comedic acts. 0 Documentaries These films are reality films that concern actual historical events without any added dramatic or filmic elements added. 0 Drama These films focus primarily on the dramatic elements of the plot in the theatrical sense. 46 o Historigal Themes These films were ones set in the civil war, or slavery. That is films whose historical period was focused on the slave system or its aftermath, the civil war. 0 Musicals These films are adaptations of broadway musical plays or films in which the plot is carried through musical sketches. 0 Race Relations In these films the primary theme is the relationship between whites and blacks. In particular, this category focuses on the interactions of whites and blacks in society. 0 Other This category is used for those films whose primary focus is not covered in any of the other categories. Each film was counted only once and categorized by its predominate theme. The categories are both exhaustive and exclusive. It must be remembered, however, that although each film may contain elements that would justify placing it in another category, films were categorized by their predominate characteristic according to the film synopsis in Ploski and Williams. The Hypotheses and Statistical Teating The model of motion picture production and its relationship with news media agenda setting leads to a two hypotheses that are testable using the operationalization of the variables. The two hypotheses test the relationship between the production of motion pictures featuring black actors and the news media agenda. Multiple linear regression will be used as the tool of statistical analysis for the two hypothesis of this study. 47 Multiple regression is one method for studying the effects and magnitudes of the effects of more than one independent variable on an dependent variable (Kerlinger, p. 603 Chatterjee and Price p. 3) and is part of the general linear model (along with ANOVA). The relationship between the dependent (Y) variable and the independent (X) variables is expressed in a regression equation in the form: Y = b0 + blxl + bzxz ..... +ann In this equation b0, b1, b2,...bn, are called regression coefficients and are determined from the data. Each b or regression coefficient gives the magnitude of the relationship between an individual independent (X) variable and the dependent (Y) variable. The regression coefficients (b) represent how much Y will change for a unit change in the particular X variable. For example, a regression coefficient with a value of .03 indicates that for each change of one in the X variable, the Y variable will change .03 units. The effect of each independent variable (X), then can be expressed by its regression coefficient. The overall magnitude of the effect for all independent variables on the dependent variable is measured by the Multiple R value (R). The value of the Multiple R ranges from negative one to positive one. The Multiple R value, when squared (R2) expresses the magnitude of the effect of all independent variables on the dependent variable. An R2 48 value of 0 indicates that the independent variables had no effect and explain none of the variation in the dependent variable. A value of 1 indicates that the independent variables explain all of the variation in the dependent variable. An R2 value of 1 also indicates that the regression equation provides perfect prediction from the independent variables to the dependent variable (Huntsberger, Croft, and Billingsley, p. 450 ). Multiple regression, then will be used in the statistical test of the hypothesis of this study. The R2 value will be used to express magnitude of the relationship between films featuring black actors and various media agenda variables. The regression coefficients will be used as measures of the individual independent variables’ (media agenda) effect on the dependent variable (films featuring black actors). Hypothesis 1: The Media_AgendafPrecedea the Prodaation of Moviea‘Feagaring,Black Actor: This hypothesis is the first to test the relation between the media agenda and the production of motion pictures. The number of films featuring black actors will be regressed (using SPSS multiple regression) on the number of articles on blacks. The articles on blacks will be lagged for a ten year period. ,The lag period that has the highest significant R2 value will be entered first in the equation and used to determine not only whether or not there is a relationship between the 49 media agenda at a particular time but also the magnitude of that relationship. Additionally, other lagged media agenda periods will be added to the regression. The addition of these variables will increase the R2 value and therefore the explanatory power of the relationship. Since it was felt that the number of films featuring black actors was related to total domestic production of films, the percentage of the total number of films featuring blacks will also be regressed against the lagged number of articles indexed on blacks. In short, the general equations being tested are: Table 2 Regression of Films on Media Agenda Y Variable X Variable Number of f(Articles on blacks at time t to films featuring time t-lO) black actors at time t f(Articles on blacks at time t to Percent of films featuring time t-lO) black actors at time t where t = the current year Forward selection will be specified in the test of the regressions. Forward selection enters the independent variable (number of articles) with the greatest relationship with the dependent variable (number of films) first. At each subsequent step, the variable with the most significant 50 partial correlation with the dependent (Y) variable (number of films) is entered until the probability limit of .05 is reached. This procedure will not only yield the total amount of variation in motion picture production explained by the news media agenda, but also the magnitude of that relationship and the combination of variables that best predict the number of films featuring black actors. Since forward entry procedure is being used, the lag period with the greatest association with motion picture production will be entered first and so on until the limits of the probability of error (.05) is reached. Hypothesis 2: The Media Story Type is Related to the Type of Motion Pictare Prodaaed: After the lag period for the overall relationship is established, each film type will regressed against all the media story types and the total number of articles on blacks. In short the relationship to be tested is: Table 3 Regression of Fila type against News story type Film type = f(News story type at time t to at time t time t-10 and Articles on blacks at time t to time t-10) where t = current year Each of the film types will be regressed against each of the news story types and the total number of articles on 51 blacks over a ten year period. The news story types and the total number of articles on blacks will be lagged for the ten previous years. The regressions will be run separately for each type of film and the R2 values will be compared for the combination of variables that explain the greatest variation in the number of each film type produced. Like hypothesis 1, forward entry of the variables will be used and entry into the regression equation will be limited by a probability of error of .05. The results of these tests will indicate not only which news story type is most significant for each type of film produced but also the combination of news story type and total number of articles that best explain the production of each film type. CHAPTER IV RESULTS Before reporting the test results of each specific hypothesis, the data will first be described and then evaluated to establish the magnitude and variation of the major variables. Simple descriptive statistics will be presented for both the number of films featuring black actors and the number of articles indexed on black Americans. Descriptive Analysis of the Naaber of Filga Featuring _lackgactors: Three hundred and thirty three films featuring black actors have been released over the period of 1950 to 1980. The table below presents the frequency distribution of the number of films released each year. (See Appendix A for a table of the number of films released by year.) 52 53 Table 4 Frequency Diapribation of Films Featuriag Black Actora Number of Frequency Percentage Cumulative Films Percentage 0-5 9 29.03 29.03 6-10 11 35.48 64.51 11-15 5 16.13 80.64 16-20 2 6.45 87.09 21-25 1 3.23 90.32 26-30 2 6.45 96.77 31-35 1 3.23 100.00 Total 31 100.00 As Table 4 shows for the majority of years, 10 or fewer films (64%) have been released featuring black actors. In 3 years, however, more than 25 films have been released. To further understand the variation of the number of films released each year, other descriptive statistics were calculated. Table 5 presents those statistics: Table 5 Mean, Standard Deviation.aand Range of Filaa Mean Number Standard Deviation Range of films 10.8 8.5 2 to 32 Table 5 shows that the average number of films produced each year featuring black actors is about 11. The standard 54 deviation, however, at 8.5 points out that there is a great deal of variation in the production of films featuring black actors. The range from 2 to 32 films per year reinforces this variability of the distribution of the number of films released. Over the period, then, the number of films featuring black actors varied from a low of two in four of the years to a high of 32 in one year. The two tables taken together show that although the average number films featuring black actors released each year is about 11, in the majority (64%) of the years fewer, than 11 films have been released. The difference is caused by the great increase in number of films released in years where more than 11 films were released. Descriptive Analysis of the Naaber of4Articlea indexed on Black Americans: Over the period of this study 4,931 articles had been indexed in the Reader’s Guide under the category of blacks or Negroes in the United States. The table below presents the frequency distribution of the number of articles indexed each year and the percentage of the total number of films released over the period. (See Appendix A for a table of the number of articles indexed each year.) 55 Table 6 Fregaency Distribution of Filas Featuring Black Actors Number of Frequency Percentage Cumulative Articles Percentage 50—75 1 3.23 3.23 76-100 8 25.80 29.03 101-125 7 22.57 51.60 126-150 2 6.45 58.05 151-175 1 3.23 61.28 176-200 2 6.45 67.73 201-225 3 9.68 77.41 226-250 2 6.45 83.86 251-275 3 9.68 93.54 276-300 1 3.23 96.77 301-325 0 0.00 96.77 326-350 1 3.23 100.00 Total 31 100.003 To further understand the variation of the number of articles indexed on black Americans each year, other descriptive statistics were calculated. Table 7 presents those statistics: 56 Table 7 Mesa; Standard Deviation, and Range of Articles Mean Number Standard Deviation Range of Articles per year 159.1 72.9 64 to 328 The number of articles indexed on black American has averaged about 159 per year as table 7 shows. The standard deviation of the number of articles is about 73. This indicates that there is a considerable degree of variability in the distribution of articles over the period. The range from 64 to 328 articles per year reinforces the variability of the distribution. As table 6 shows more than half (58*) of the years had a number of articles below the mean value for the whole period. gpaparing:the Diatributions of the Variables Both distributions show a degree of variability. One statistic that can be used to compare the variability of a distribution is the coefficient of variation (Huntsberger, Croft, and Billingsley, 1980, p. 67-68). The coefficient of variation is the standard deviation divided by the mean of the sample. The resulting ratio expresses the standard deviation as a percent of the mean. These coefficients can then be compared to determine the relative variation of two or more samples. The coefficient of variation for both the number of 57 films released featuring black actors and the number of articles indexed on blacks/Negroes in the United States is shown in Table 8: Table 8 Coefficient of Variation Variable Standard Mean Coefficient of Deviation Variation Number 8.5 10.8 78.73 of Films Number 159.1 72.9 45.8% of Articles Table 8 shows that the number of films featuring black actors is more variable over the period (1950 to 1980) than the number of articles indexed on blacks/Negroes in the United States. In general, than the number of films featuring black actors has varied more relative to its average value than the number of articles indexed on blacks/Negroes in the United States. Comparing the Distributiona,over Tiaa Both the number of films featuring black actors and the number of articles indexed on blacks/Negroes in the United States have varied over the period. Since this study is concerned not only with the general variability but also with the variability of each variable over time, the yearly distribution of both the number of films and number of articles indexed has been 58 graphed against time. Graph 1 (Films featuring black actors) is shows of the number of motion pictures featuring black actors graphed against the year. Graph 1 shows that the peak number of films featuring black actors (32) was released in 1973. The graph also demonstrates that the release of films featuring black actors has not been progressed steadily over time. The number of films released in 1979 (4) almost matches the number of films featuring black actors in the 1950. Between these two periods, there has been a considerable degree of difference in the number of films released as evidenced by the relatively high coefficient of variation (78.7%). Although the coefficient of variation for the number of articles indexed on blacks/Negroes in the United States is much smaller (45.8%), it too has varied over time. Graph 2 (Articles indexed on black Americans) shows the yearly distribution of number of articles. This distribution over time shows that the two possible intervening variables, the number of magazines indexed and the number of black oriented magazines, do not seem to have had much effect. The number of magazines indexed has increased from a low of 84 (1957 to 1959) to a high of 188 (1978 to 1979) the number of articles indexed at both time periods is almost the same (79 in 1959 and 125 in 1979). The number of magazines indexed has little effect on the number of articles indexed under blacks/Negroes in the United States. 58a FILMS FEATURING BLACK ACTORS GRAPH 1 32 "‘ 30--I 23A 26- 24- 22-4 20-: iaJ 16-1 14— 12- 10- 6— 4.4 1946 1951 1956 1961 1966 1971 1976 ARTICLES INDEXED ON BLACK AMERICANS GRAPH 2 1946 1951 1956 1961 1966 1971 1976 59 Additionally, the variation of articles by time seems to neither increase nor decrease as the result of the total number of magazines indexed. The second variable thought to possibly intervene was the number of black oriented magazines indexed. Once again, the number of black oriented magazines indexed seems to have little effect on the number of articles indexed. In fact the number of articles indexed on blacks in the U.S. actually reached it’s high point (1963 with 328) while three black oriented magazines were indexed and its low point (64 in 1977) while 5 black oriented magazines were indexed. Graph 2 clearly shows that the variation in the number of articles indexed on blacks has fluctuated greatly over time. Although one might expect that the distribution of articles would be rather constant over time, Graph 2 shows that this is not the case and the variation has not been effected by either the total number of magazines or the number of black oriented magazines indexed. The last possible intervening variable is the relationship between the total number of films indexed on blacks yearly and that number as a percentage of the total number of articles indexed per year. Estimates were made of the total number articles indexed each year by estimating the number of articles per page in gaaderafGaide and the percentage of those articles indexed on blacks was calculated. The percentage of black articles indexed was 60 regressed on the total number of articles indexed on blacks to determine the magnitude of the relationship. As might be expected the resulting R2 value (.9339) was extremely high indicating that percentage and total number of articles indexed are equivalent measures. Since the measures are for all useful purposes equivalent, the total number of articles indexed on blacks will be used to preserve computer memory. Any conclusion made for one will be valid for the other. Statistigaa Taapa of the Hypothesea Having established that there is some degree of variation in both the number of films featuring black actors and the number of articles, the study will now proceed test the relationship between the two. Hypothaais l: The aedia_agenda precedaa,theiproduction of aggieaafeaturing black actors. For this hypothesis multiple linear regression was chosen as the test. The media agenda (the number of articles indexed on blacks/Negroes in the United States) was lagged for a ten year period. Lagging a variable means that the previous year’s number of articles was regressed against that year’s number of films featuring black actors. For example, a no year lag would regress 1960 films against 1960 articles. A one year lag, on the other hand, would regress 1960 films against 1959 articles. Forward entry was chosen as the method of entry of variables for the regression. Forward entry (SPSS Inc., p. 61 666) enters the variables according to the greatest amount of variation in the dependent variable (films) explained by the independent (articles). The decisions on entry are based on the correlation of the independent with the dependent variable. The general form of the regression equation is: Y 3 b0 + blxl +b2X2....ann Of interest for this hypothesis is the regression coefficients b as the estimate of the effect of the independent variable (articles) on the dependent variable (films). Additionally, the R2 value for the entire regression equation tells the amount of variance in the dependent variable (films) that is explained by the independent variable (articles). The media agenda at lagged times from 0 (current) to ten years was regressed against the number of films released featuring black actors and the percentage of domestic films featuring black actors. fiagresaion of Total Naaber of Filaa Featuring Black Actors against Nuaber of Articles at Lagged Tiae Perioaa: Three variables were entered into the equation before the probability limits were reached. The Analysis of Variance of the regression was significant at less than the .00001 level. The Analysis of Variance table for this test is presented below. 62 Table 9 Analysis of Variance of Regression of Total Number of films featuring Black Actapa df SS Mean Square F P Regression 3 1326.39 422.13 32.27 >.0001 (explained) Residual 21 287.71 13.70 (unexplained) The variables entered the regression in the following manner, the number of articles 6 years earlier (t-6) was first being most associated with the number of films at time t. The number of articles nine years earlier (t-9) was next, and finally the number of articles four years earlier (t-4). The unstandardized regression equation is presented in Table 10 below: Table 10 Regression Equation for Nuaber of Filaa Number of = - 9.78 + .052(t-6) + .035(t-9) + .039(t-4) films at time t (where t-6, t-9, and t-4 are the number articles at each of the lagged time periods.) R2 = .73457 The R2 value of the regression was .73457. This indicates that the three lagged number of articles variables explain 73x of the variation of the in number of films released. This means that the three media agenda variables are strongly related to the number of films released per 63 year. aggression of Percentage of Filaa Featuring Black Actors against Naabgr of_§rticles at Legged Tiae Periods: The same procedure was followed as above for the percentage of domestic films featuring black actors. The percentage of films featuring black actors is the number of films featuring black actors released each year divided by the number of domestic films released for the same year. As with the regression of number of films, three lagged time periods combined to explain a significant among of variation in percent of films featuring black actors. The Analysis of Variance of the regression table is presented below: Table 11 Analyais of Variance of Regression of Percent of filaa featuring Black Actors df 88 Mean Square F P Regression 3 238.77 79.59 29.81 >.0001 (explained) Residual 19 50.74 2.67 (unexplained) The lagged variables entered the equation in the following manner. The number of articles five years perviously (t-5), followed by the number of articles three years previously (t-3), and finally the number of articles nine years previously (t-9). The unstandardized regression equation is presented below: 64 Table 12 Regression Eguation for Percent of Nuaber of Filaa Percent of = - 3.99 + .026(t-5) + .020(t-3) + .011(t-9) films at time t (where t-5, t-3, and t-9 are the number articles at each of the lagged time periods.) R2 = .82475 The three entered variables yield a R2 value of .82. This value indicates that approximately 823 of the variation in the percentage of films is explained by the three entered variables. Comparison of the two Regression Eguations: The process used in obtaining the two equations differs only in the dependent variable. In the first case, the dependent variable is the total number of films. In the second case, the percent of films featuring black actors is used. Over the period, the number of domestic films is negatively correlated (r = -.052) with the number of films featuring black actors. In other words, as the number of domestic films decreases, the number of films featuring black actors increases. This tends to increase the variation in percentage of films featuring black actors. Interestingly enough, this increase in variation is also explained by the entered media agenda variable. If the both regression equations are compared, the there is little difference between the two. 65 - 9.78 + .052(t-6) + .035(t-9) + .039(t-4) Number of films at time t R2 = .73457 Percent of films at time t - 3.99 + .026(t—5) + .020(t-3) + .011(t-9) 32 = .82475 In interpreting the results of the two equations, each set of independent variables (media agenda variables) must be interpreted as a group. The removal of any of the variables changes the R2 value and the coefficients of the other entered variables. As a result, there is little real difference between the two. Both sets of independent variables explain a considerable amount of the variation in the dependent variables (number or percent of films). The difficulty with interpreting the results of the regression of percent of films featuring black actors is that two negatively correlated variables are estimated (films featuring black actors and domestic production of motion pictures). Since there is little difference between the two in terms of their explanatory power and the lagged time periods entered are similar subsequent tests will be performed only on the total number of films rather than the percent of domestic films. 66 Hypothesis 2: The Media:Story Type is Related to the Type of Motion Pictape Produced. Like the previous hypothesis, multiple linear regression was used as the test. In this case, however, the type of article was regressed against the type of film produced. Each type of film and the total number of films were also regressed against the total number of articles indexed on blacks. The table below presents the all variables, both film type and article type that were used. Table 13 Variables used in Regression of F11! Type against Article Type Film Type Variables Media Story Type Variables Total films featuring Total articles indexed on Black Actors blacks in the U.S. Action/Adventure Economic Conditions Biography Education Comedies Employment Documentaries Equal Rights/Civil Rights Drama Housing Historical Themes Occupations Musicals Politics and Suffrage Race Relations Social Conditions Other Race Relations Before testing this hypothesis, however, a descriptive 67 analysis of both the film type and media story type will be conducted. The following table presents film type, total, percent of total, and mean number per year for each type of film released over 1950 to 1980 period. Table 14 Descriptive Analysis of Fila Typaa Film Type Variable Total Percent Mean Number of Total per year Total films featuring 333 100 10.8 Black Actors Action/Adventure 111 33.3 3.6 Biography 13 3.9 .4 Comedies 25 7.5 .8 Documentaries 16 4.8 .5 Drama 31 9.3 1.0 Historical Themes 20 6.0 .6 Musicals 9 2.7 .9 Race Relations 85 25.5 2 7 Other 20 6.0 .6 The predominate film type featuring black actors over the period is the Action/Adventure category. About one in three films featuring black actors fall into this category. As might be expected, Race Relation films are the next most predominate film type featuring black actors. Over the period, over one fourth (25.5X) of the films featuring black 68 actors were race relations films. Similarly, the number of articles indexed on blacks has fluctuated over the years. Table 15 is a descriptive analysis of media story types for the articles indexed on blacks in the U.S. and Race Relations. Table 15 Descriptive Analysis of News Story Typea Media Story Type Variables Total Percent Mean Number of Total per year Total articles indexed on blacks in the U.S. 4,931 100.0% 159.1 Economic Conditions 183 3.7 5.9 Education 556 11.3 17.9 Employment 322 6.5 10.4 Equal Rights/ 605 12.3 19.5 Civil Rights Housing 118 2.4 3.81 Occupations 27 0.5 .3 Politics 558 11.3 18.0 and Suffrage Social Conditions 227 4.6 7.3 Race Relations* 351 100.02 11.3 *Race Relations is a separate category As the above table shows, over the period searched (1946 to 1980) the predominate news story type was about 69 Equal Rights/Civil rights. Approximately 20 articles per year were indexed under this subhead. The second most frequent news story type was Politics and Suffrage. The Race Relations Category, although indexed separately, did average about 11 articles per year. Regression Analysis of Fila type and Mediaistory Type: This hypothesis states that Film type is some function of media story type. In order to test this, each film type was regressed against the Media story types to determine which media story type was most related to the production of the film type. Each of the media story type variables was lagged for a ten year period to determine what lag period and what combination of variables and time lags best explained the relationship between film type and media story type. Results of Ragression of Total Nuaber of Filaa Featuring Black Actors: Although the previous hypothesis showed a strong relationship between the number of films released and the number of articles, it was felt that some media story types would also explain the variation of the number of films released. As a result, the total number of films was regressed against the total number of articles, and the news story type. The resulting equation was: Total Filmst = -7.29 + .060(Artic1est_6) + .231(Politicst_8) + .031(Articlest_4) R2=.8494l p < .0001 (t = current time period) 70 In other words, the three variables, the media agenda as measured by the total number of articles at 6 and 4 years earlier, and the number of political articles at 8 years earlier account for over 84% of the variation in the total number of films. The probability of sampling error is less than .0001. Results of Regreaaion of ActionjAdventare FilaiType: Action/Adventure films have been the predominate type that feature black actors. The regression of this type resulted in the following equation: Action/Adventuret = .04 +.286(Race Relationst_5) -.l75(Race Relationst_8) +.044(Education -3) +.133(Race Relationst_6) R2=.88063 p < .0001 The release of Action/Adventure films featuring black actors is a function of Race Relations at three time periods (5, 8, and 6 years pervious) and the number of articles indexed on education three years previously. The results are significant at the .0001 level. Over 88% of the variation in Action/Adventure films can be explained by these three variables. Results of Regression of Biography: No significant relationship could be found for the production of the biography film type. It should be remembered, however, that only 13 were produced over the period. Factors other than the media agenda explain the variation in biography films 71 produced. Results of Ragpaaaion of Caaediaa: Comedies represent 7.5x of the total number of films featuring black actors. The resulting regression equation was: Comediest = -.21 +.128(Economicst_7) +.100(Economics _ ) -.058(Socia1 Cond1tionst_5) +.028(Politicst_4) R2=.80126 p < .0001 The three variables entered in the equation account for about 803 of the variation of Comedies. The probability of this being the result of sampling error is less than .0001. The negative relationship of Social Conditions indicates that as the number of articles indexed on social conditions increases, the number of comedies will decrease. The variable that has the greatest relationship with the release of comedies featuring black actors is the number of articles indexed under Economic Conditions 7 years earlier. Results of Regression on Docaaentariea: Although over the period, only 16 documentary films (including "live” recordings of events) a significant relationship was discovered. The resulting equation is presented below: Documentaryt = -l.05 +.47l(0ccupationt_6) +.008(Articlest_4) R2 = .66793 p < .0001 Two variables explain almost 67% of the variation in the production documentaries featuring black actors. The 72 number of articles 6 years previously, and the total number of articles four year previously, together explain the majority of variation in the number of documentaries. Both are positively related to the number of documentaries indicating that an increase in either results in an increase in the number of documentaries released that feature black actors. Results of Regression of Draaa: Of the nine film types, Dramas represent 9.3x of the total number of films released that feature black actors. The resulting regression equation was: Dramat = .46 +.05(Race Relationst_8) R2 = .34724 p = .0019 The Number of Articles indexed under Race Relations for the period 8 years previously, then, accounts for about 35% of the variation in the number of Dramas released featuring black actors. Since the sign of Race Relations (eight years previously) is positive indicating that as the number of articles indexed on race relations increases the number of Dramas featuring black actors will also increase. Regression of Historical Themes: Six percent of the films over the period had historical themes, being either about slavery or the civil war. The resulting regression equation was: 73 Historyt = -.69 +.502(Occupations -6) +.006(Articlest_4) 32 = ..51681 p = .0002 The number of articles indexed under Occupations six years previously and the total number of articles four years previously accounts for over 51% of the variation in History films that feature black actors. The probability of this happening by chance is only .0002. Ragresaion of Maaiaal Filaa: Only 2.7: of the number of films released over the period featuring black actors are musicals. No news story type or lagged number of news article resulted contributed significantly to the variation of musical films featuring black actors. Regression of Race Relation Filaa: Over 253 of the films featuring black actors were race relation films. The resulting regression equation was: Race Relation Filmst = -1.1 +.065(Equa1 Rightst_4) +.017(Articlest-4) R2 = .68415 p < .0001 Over 68% of the variation in the number of race relation films is accounted for bye the number of articles indexed under Equal Rights four years previously, and the total number of articles indexed four years previously. Since both are positively related, an increase in either of them results in an increase in the number of race relations films. 74 Overall Results of Hypotheais a: The test of the second hypothesis (Motion picture type is related to news story type), is confirmed for six of the eight film types. The Hypothesis also tested whether or not there was a relationship between the total number of films featuring black actors and the news story type. The variation in six of the eight film types featuring black actors and the total number of films featuring black actors can be explained in part by either particular news story types and the total number of articles indexed on blacks in the U.S. or Race Relations. Results of both Hypotheaaa: To a large degree, both alternative hypotheses have been confirmed. The number of films featuring black actors is related to the number of articles indexed under blacks/Negroes in the United States. Finally, six of the eight categories of film types are related to various news story types at various times (Hypothesis 2). The overall production of films featuring black actors is also related to news story type. CHAPTER V SUMMARY AND CONCLUSIONS Review of Results: The two research hypotheses were both confirmed. The results of this study clearly indicate that there is a strong relationship between the number of articles indexed under blacks/Negroes in the United States and the number of films released featuring black actors. The two hypotheses tested the relationship between the number of motion pictures featuring black actors released and the number of news articles indexed on blacks/Negroes in the United States. The first hypothesis tested the relationship between the total number of films featuring black actors and the total number of articles indexed under blacks/Negroes in the United States. This hypothesis showed that the number of articles lagged over three periods accounted for 73% of the variation in number of films released. When the percentage of films featuring black actors was used, over 828 of the variation of percent of films was explained by number of articles. The last hypothesis tested whether there was a relationship between motion picture type and news story type. The total number of films featuring black actors was 75 76 also regressed to determine if any particular news story type was significantly related the total number of films. The variation in total number of films and for six of the eight film types was strongly related to the number of articles indexed under blacks/Negroes in the United States and or Race Relations. Interpretation of Reaalpa: Both hypotheses require some interpretation since they test the relationship between the number of films released and number of articles indexed on blacks/Negroes in the United States or Race Relations. The results of the statistical tests require the creation of some additional concepts to aid in interpretation. Trigger and Contextual Variables: In order to fully explain the results of the hypotheses it is necessary to propose an explanation for why the variables entered the equations as they did. The method used was forward entry of the variables in a multiple regression equation. Forward entry of the regression enters the independent variable most related to the dependent variable first, the second most related variable next (while controlling for the first), and so on until the probability limit of .05 is reached. The resulting R2 value gives the percent of variation in the dependent variable explained by all entered independent variables. Since there was no strong theoretical basis for how the variables should have entered the equation, forward entry 77 was chosen to ensure that all of the most related variables would be entered into the equation in the order of their statistical significance. The variables in the resulting equations, then, are ordered according to their relationship to the dependent variable rather than any particular time sequence. In other words, the entry sequence of the independent variables has no relationship to their chronological sequence. The interpretation of the independent variables must, therefore, be taken as a one group of variables rather than as individual variables. In order to facilitate this interpretation, two different concepts will be proposed to classify the variables. The first concept is that of the ”trigger" variables. This variable is that enters the regression equation with the smallest lag period. It is felt that no matter where the variable enters the regression equation, its effect is to "trigger" the motion picture production process. The other variables (those with higher lags) will be considered contextual variables. Contextual variables, though possibly more related to the dependent variable (number of films) are considered more as being part of the producer’s history and memory of past events to form the context in which his audience image is composed. The . contextual variables may also signal a gestation period for the film. The film may have been conceived long before the final release given the complexities of the motion picture 78 industry. Interpretation of Hypothesis 1: Thaaaadia agenda precedeaathe production ofvaation pictapaayfaaturinggblack actors. The testing of this hypothesis resulted in two regression equations. Regression analysis was performed using both the total number of films featuring black actors and the percent of domestic films featuring black actors. As discussed previously, since films featuring black actors and domestic production of films were negatively correlated, resulting in increased variation due to the percentage itself, the total number of films was used for interpretation. The resulting equation was: Total Number of = -9.78 +.052(Articlest_6) Films Featuring +.035(Artic1est_9) Black Actors +.039(Articlest_4)t at time t R2 = .73457 *trigger variable Together, these three variables account for more than 73% of the variation in the total number of films featuring black actors. Since all of the coefficients are positive, each increase in the number of articles results in an increase in the number of films featuring black actors produced. The trigger variable in this equation is the number of articles less than four years (but greater than three) previous to the release of the film. It is proposed here that the number of articles four years previous to the release year of the film "triggers" 79 not only the production of the films but also memories as represented by the articles at 6 and 9 years previous to the release of the film. The motion picture producer as the key player in the production of a motion picture is not only affected by what is happening at the present time (four years before release of the film) but also the memories of the past few years. The contextual variables may also represent the gestation period for the films. Additionally a number of films may require a longer gestation period (from conception to release) than others. Given that a film may require from 18 months to 15 years to move from conception to release, the trigger of four years previous seems reasonable. It should also be remembered that not only the motion picture producer, but also various investors evaluate the potential of the film.to return their investment in the film. What the related variables at the various lagged periods (six and nine years) say is that evaluation is based not only on the current media agenda, but also on the memories, and history of the period. Films could reasonably be conceived nine years before release, but only triggered to release by the media agenda at that point (four years previous to release). The strong relationship between the number of films released featuring black actors and the number of articles on blacks/Negroes in the United States is clear evidence of 80 the strong effect that one media can exert on another. Interpretation of Hypothesis 2: The aedia story type is related to the type of aption picture produced. In this hypothesis, all of the film types and the total number of films were hypothesized to be related to either the total number of articles in a particular news story type or the number of total articles. The total number of films was also regressed to determine if there was any relationship between particular story types and the total number of films featuring black actors. Interpretation of Resulta of Total Naaaer of Filaa Featarinnglack Actors: The total number of films featuring black actors was also regressed against not only the overall number of articles indexed on black/Negroes in the United States, but also on the specific media story type. The results of the regression was: Total Number of = -7.29 +.060(Articlest_6) Films Featuring +.231(Politicst_8) Black Actors +.031(Articlest_4)* at time t R2 = .84941 *trigger variable Since all of the coefficients of all variables are positive an increase in any of them results in an increase in the number of films featuring black actors. The trigger variable in this equation is again (as in Hypothesis 1) the total number of articles four years previous to the release 81 of the film. The contextual variables are blacks in politics eight years previous to the release of the film and the total number of articles six years previous to the release of the film. The entry of the Politics variable indicates that this news story type is particularly related to production of motion pictures featuring black actors. The other contextual variable to enter the equation is the total number of articles six years previous to the release of the film. These two variables are the most important in helping to determine the audience image of the motion picture producer at the conception stage of the motion picture. Politics eight years previous to the release of the film because it is the kind of news story that would be remembered and used in creating the audience image of the motion picture producer. Also some films could have been originally conceived during those periods (six and eight years before release) and because of circumstances not triggered to release for a 2 or 4 year period. Interpretation Results of Action/AdvenpaaaiFila Type: The regression of the Action/Adventure film type on total number of articles and news story type resulted in the following equation: 82 Action/Adventuret = .04 +.286(Race Relationst-5) -.l75(Race Relationst_8) +.044(Education _3)* +.123(Race Relationst_6) R2 = .88063 *trigger variable The trigger variable in this equation is the number of articles on blacks in education three years previous to the release of the film. Action/Adventure films tend to be targeted at younger audiences. The trigger variable of blacks and education is evidence of this. Since the producer is aiming the film at a young audience, it would seem entirely appropriate that his attention would be focused on news stories about young people. The Education category is the one news story type that is predominately about young people. The contextual variables in this case are all articles about race relations at various times. The race relations (or interracial cooperation) articles indicate that the casting decisions on featured black actors is related to how popular a topic interracial cooperation is as a background for the creating or casting the role. Action/Adventure films rely heavily on the cooperation of characters to overcome the dangers of the adventure. The negative relationship between Race Relations eight years previous to the release of the film is mitigated by 1 its positive correlation (r = .426) with Race Relations five years previous. In other words as Race Relations eight 83 years earlier increases, so does Race Relations five years earlier. The overall effect on the equation is to set limits on the effect of the negative relationship that Race Relations eight years previous has in the equation. In summary, the Race Relations variables as contextual variables tend to imply that interracial cooperation as a topic is positively related to the production of Action/Adventure films featuring black actors. Even the negative relationship with Race Relations eight years previous is limited by its positive correlation with Race Relations eight years previous. As above, these contextual variables may represent different gestation periods from the conception of the film to actual release. Interpretation of Resulaa of Biquaphy Fila Type: No significant relationship could be found for the release of the Biography film type. This film type, however, is probably much more affected by the individual subject of the film, his or her life and accomplishments. The number of articles on black Americans would have very little effect on the number of biography films released. The individual subject, and his life story are rather the critical factors in producing the producer’s audience image which in turn affects casting decisions. Interpretation of Results of Qaaadies: The regression of the Comedies film type on total number of articles and news story type resulted in the following equation: 84 Comediest = .10 +.128(Economicst_7) +.100(Economics _ ) -.058(Social Condltionst_5) +.028(Politicst_4)* R2 = .80126 *trigger variable The trigger variable for the release of comedies featuring black actors is the number of articles indexed on politics and suffrage of black Americans. The politics variable is the variable that appears to be one news story type that is sets the conditions for production of comedies featuring black actors. The contextual variables (Economics, and Social Conditions) would be factors that producers of comedy motion pictures would be sensitive to since they are serious subjects that form the basis of most comedic situations. These variables would form the background context in which the trigger of politics would enter to spark production of the film. Additionally, these contextual variables could represent different gestation periods for films. Interpretation of Results of Doaaaentary Fila Type: The regression of the Documentary film type on total number of articles and news story type resulted in the following equation: Documentaryt = -l.05 +.47l(Occupationt_ ) +.008(Articlest_4)§ R2 = .66793 *trigger variable The trigger variable in the release of documentary is the total number of articles on blacks in America four years 85 previously. The effect of the total number of articles is to help create the audience image of the documentary film producer. The contextual variable is Occupations. This news story type is generally about individual blacks or career choices. It would seem that this kind of news sets a tone that helps encourage the production of documentary films featuring black actors. This relationship may be explained since documentaries are about actual events and people and publication of articles on occupations may signal a trend toward an audience interest in reality rather than fantasy. The total number of articles four years previous to the release of the film would be the immediate impetus to the conception or production of the film. Interpretation of Results of Drama FilagType: The regression of the Drama film type on total number of articles and news story type resulted in the following equation: Dramat = .46 + .046(Race Relationst_8) R2 = .34724 Only race relations eight years prior to the release of dramas featuring black actors was significantly related to the release of dramas. This variable accounts for approximately 353 of the variation in the number of dramas produced. Given that the lag time is so long, it would seem likely that factors other than the media agenda are also of 86 prime importance in the release of dramas featuring black actors. One possible explanation for this of relationship is that dramas are generally much more risky films and producers of dramas base their audience images more directly on the scripts and the actual or potential performance of the dramas in the legitimate theater. The audience image, in this case would be the theater going audience. Race Relations and interracial cooperation’s relationship may indicate a longer gestation period for dramas and like the Action/Adventure film type, dramas featuring black actors rely on interracial cooperation as a key film element. Interpratation of Resalga of Hiatorica} Theaaa: Historical themes are films that are either set during slavery or in the civil war. The resulting equation of the regression of historical themes was: Historyt = -.69 +.502(Occupations -6) +.006(Articlest_4)* 32 = .51681 *trigger variable The trigger variable in this equation is the total number of articles on blacks/Negroes in the United States four years before the film is released. Once again, this would seem to indicate that the motion picture producer is strongly affected by the number of articles on black Americans. .One would expect Historical (set in slavery or the Civil War) films with this theme to be particularly 87 affected by the number of articles since the setting almost requires the casting of black actors. Occupations six years previous to the release of the film as a contextual variable may indicate that producers of films with History (slavery and civil war periods) are affected in creating their audience image by their attention to feature or soft news stories about blacks. Additionally, this variable may represent a gestation period for the conception of the film to its release. Interpretation of Resulaagof Masical Fila_Typaa: No significant relationship was found between the total number of articles or news story type and the release of musicals featuring black actors. Like Dramas, the audience image of the producers of these films may rest with past performance or the potential performance of the film with traditional theater—going audiences. Interpretation of Results of Race Relatioaa Fila Types: The resulting equation of the regression of Race Relations film types was: Race Relationst = —l.1 +.065(Equal Rights _4)* +.017(Articlest_4)£ R2 = .68415 *trigger variable Interestingly enough, the two variables that entered the equation for the regression of Race Relations can both be considered trigger variables. The producers of these 88 films no doubt intend to appeal to both races. Their audience image, is influenced by the equal rights/civil rights articles and the total number of articles on blacks in the United States. In this case, however, the nature of the subject matter necessitates casting black actors in roles. The results indicate that Race Relation films are released four years after there has been variation in the number of equal right/civil rights and total number of articles on blacks. The results clearly point out that at the time of conception, the equal rights, and total number of articles create an atmosphere of what is popular to the producer. The producer constructs his audience image based on these news story types. General Interpreaation of Reaylta Overall, the general relationship of the media agenda as measured and the number of films produced featuring black actors is strongly supported. The trigger variable concept as an analytic tool helps to clarify the relationship between the two. Table 16 presents the both the film types and the trigger news story types: 89 Table 16 Fila_Types and Trigger News StorygTypes Film type Trigger Variable Total films Articlest_4 featuring black Actorst Action/Adventuret Educationt_3 Biographyt None identified Comediest Politicst-4 Documentaryt Articlest_4 Bra-at Race Relationst_8 Historyt Articlest_4 Musicalst None identified Race Relationst Equal Rightst_4 Articlest_4 Table 16 clearly shows that the total number of articles is the trigger variable not only in total films but also in three of the eight film types. In two cases no trigger variable could be found. Of those film types that could be explained by the variables in the study, three of six had trigger variables of the total number of articles on blacks four years previous. What the results of this study indicate is that there is a strong relationship between the media agenda as measured by the number of articles published in magazines 90 and the casting of motion pictures. The mechanism for this relationship is the audience image of the motion picture producer. This audience image of the motion picture producer is impacted by the media agenda. That is the motion picture producer whether consciously or unconsciously relies on the media agenda to make decisions about the audience his film will attract. The trigger variable for not only the Total number of films, but also for three of six film types was the total number of articles four years previous. For one additional type (Comedies) the trigger variable was politics four years previous to release. Since the lag time in the trigger variables (four years) matches the average time from conception to release, the evidence for the impact of the media agenda on the motion picture producer’s audience image is compelling. From the producer’s audience image, all major decisions about the film are made. These decisions also include the casting of the film. Since the major hypothesis of the study can be accepted (that is the media agenda affects motion picture production), this model of the decision making process in the industry is supported. There had been sketchy evidence that the media agenda does affect the motion picture production process since one motion picture production company sends people across the country to question people on newspaper articles (Jowett, 91 1980, p. 97). There was no study, however, that attempted to quantify this relationship. This study is also one of the few to empirically test the relationship between two different media and the influence of one on the other. Although, the operationalization of the variables was blacks in motion pictures and blacks/Negroes in the U.S., the results are so strong that they should be generalizable to other relations between different media. The lag periods of three to four years between trigger variable and release of the film also has implications for a greater understanding of at least some social events. The lag period indicates that although an issue may be newsworthy in the sense that it appears frequently in news media, its apparent social significance may be distorted by its appearance in motion pictures. An event or issue will appear in motion pictures four years after it has been high on the news media agenda. As a result, an interested social observer may be deceived into thinking that an issue continues to rank high on the nation’s agenda of important issues while, in fact the news media agenda as moved to other issues. I The observer’s selective exposure to information would tend to bias his overall view of the importance of the issue over time, since he or she would be likely be to exposed to the issue first in the news media (when it ranked high) and four years latter in the entertainment media. In the 92 intervening time period, the issue may have dropped from the news media agenda. Liaitations of the Study: Although this study found a strong relationship between the media agenda and the production of motion pictures, it does face some limitations. As with much media agenda research, the measurement of the media agenda is limited by having to refer to indexes and having to depend on the codification of the index makers (for example see Funkhouser, 1973). Secondly, since the lists of films featuring black actors was again compiled from other sources, it was necessary to rely on other’s work. This limitation was partly controlled for by compiling the tested list of films from a variety of sources. Aside from these data collection limitations, there is an additional limitation in that neither list indexes significance. The News Media Agenda setting measure measured only the number of articles indexed under blacks/Negroes in the United States. It does not measure the impact of particular articles or even the placement or number of words. A longer article in a news magazine may have a greater impact on motion picture producers than a shorter article in a general interest magazine. Given the number of articles in the study, it would seem that these matters of significance or impact would be randomly distributed. 93 The list of films does not measure the number of black actors in a film but only measures whether or not a black actor was featured in a role. It counts a film the same whether there are blacks in more than one featured role. Given the strong support for the hypotheses, the number of blacks with featured roles in a film may not be as critical to the relationship as a single black actor being featured. Another limitation of this study is that it did not consider the kinds of roles that blacks portrayed in the film. This is really only a perceived limitation in that analysis is beyond the scope of this study. The possible effect of social agitation by civil rights groups for a greater number of roles although a possible limitation was not one for the period (1950 to 1980) studied. Until 1982 (Heron, 1982, Cripps, 1977) the focus of the NAACP was on type of the role that featured black actors rather than on the number of roles. The final limitation of this study, is that the intervals used were so large (one year). This interval’s length, does not measure whether or not intense publicity for a few months during the year, has the same effect as less intense publicity over a longer part of the year. The results do not distinguish the duration of an issue within the period. Directions forgfiarther Raaaarch: The results of this study clearly establish that there is a strong relationship 94 between the news media agenda and the film industry. Other entertainment media, like network and local television, H _,__—-._1 cable, and book publishing can be studied to determine if the same relationship exists in those media. For example, broadcast network television can be analyzed as the dependent variable using types of roles for women regressed on the number of articles on feminist issues. If in other media, this relationship does not exist, then, what factors are different about entertainment production in those media. Another direction is to test whether or not a shorter interval of measure makes a difference in the results. Shaw and McCombs (1977) have posited that time is one variable that affects the agenda setting function. They feel that it takes a certain amount of time for the media agenda to exert its effect. Since the media agenda is constantly changing, a shorter interval of measure on the media agenda could help determine whether or not the duration or intensity of an issue has the greatest impact on motion picture production. In other words does a publishing great number of articles in a few months exert a greater effect on the motion picture industry than publishing the same number of articles over a longer period. Although the operationalization of the variables was chosen partly for the availability of the data (e.g. lists of articles and films were readily obtainable), the study could be replicated with other operationalizations. For 95 example, articles on the women’s movement may have an effect on the type of role (nontraditional) that actresses were cast. It is believed that these results are generalizable over a wide range of news media agenda/motion picture relationships. Although the results strongly supported the tested hypotheses, there have been no similar studies done to test the limits of the news media agenda setting function. Initially, there seem to be some issues that although high on the media agenda do not result in a direct relationship with motion picture production. For example, the Vietnam war was one of the leading media issues of the 60s and 70s (Funkhouser, 1973), but not until the 803 were there a number of films released that were openly about the war. This suggests that some issues though covered extensively by the news media are too controversial for the entertainment media. There would therefore seem to be some limit to the effect of the agenda setting function based on the controversial nature of the subject. Finally, this study can be replicated over time as the motion picture industry changes to determine if changes in the industry cause changes in the relationships found. For example, home video is rapidly becoming the major profit source for motion picture production companies. The expansion of the home video market and its characteristics 96 may influence how producers create their audience images. Theatrical release, will no longer be the prime profit generating market for motion pictures. As a result, the producers may become more sensitive to a home audience rather than the theater audience and base decisions on what people would rent or buy to watch at home, rather than what they would go to a theater to see. This and other changes in the motion picture industry may effect not only how the films are distributed but also what subjects are filmed. The effect of news media agenda on the motion picture industry, and the lag time of four years may change as a result as the industry adopts to the change in distribution channels. APPEND IX A Table 17 Number of Filas Featurianglack Actors and Articles indexed on Blacka in the United States by Year Films Articles Featuring Indexed on Year Black Actors Blacks in the U.S. 1946 5 105 1947 4 110 1948 3 100 1949 6 82 1950 2 90 1951 ll 86 1952 3 101 1953 2 100 1954 2 131 1955 2 116 1956 4 234 1957 6 206 1958 4 100 1959 8 79 1960 6 193 1961 5 184 1962 6 221 1963 7 328 1964 9 269 1965 9 274 1966 14 213 1967 13 264 1968 19 286 1969 28 227 1970 15 172 1971 20 109 1972 27 79 1973 32 100 1974 25 112 1975 13 97 1976 6 103 1977 9 64 1978 6 125 1979 4 143 1980 10 125 97 APPENDIX B Table 18 Titlea_of Filaa FeaturigggBlack Actors 1950-1980 Title Year No Way Out 1950 Stars in My Crown 1950 Bright Victory 1951 Lydia Bailey 1951 Member of the Wedding 1951 Native Son 1951 Show Boat 1951 The Breaking Point 1951 The Harlem Globetrotters 1951 The Medium 1951 The Nell 1951 To LIve Together 1951 The Steel Helmet 1951 Cry the Beloved Country 1952 Red Ball Express 1952 Tarzan’s Perils 1952 Bright Road 1953 The Joe Louis Story 1953 Go Man Go 1954 New Faces 1954 The Glenn Miller Story 1955 Trail 1955 Blackboard Jungle 1956 Goodybye, My Lady 1956 Safari 1956 That Certain Feeling 1956 Edge of the City 1957 Island in the Sun 1957 Meet Me in Las Vegas 1957 Something of Value 1957 The Benny Goodman Story 1957 The Defiant Ones 1957 Anna Lucasta 1958 St. Louis Blues 1958 The Decks Ran Red 1958 The March of the Hawk 1958 98 99 Table 18 (cont.) Titles of Films Feaparing Black Actors 1950-1980 Title Black Orpheus Night of the Quarter Moon Odds Against Tomorrow Porgy and Bess Sapphire Tamano The Sound and the Fury The World, the Flesh and the Devil All the Young Men Sergeant Rutledge Shadows Take a Giant Step The Adventures of Huckleberry Finn The Crowning Experience Biography of a Rookie Guns of the Trees The Intruder The Young One A Raisin in the Sun A Taste of Honey Carmen Jones Oceans 11 Paris Blues Pressure Point The Connection An Affair of the Skin Convicts 4 Gone are the Days Sergeants 3 The Cool World The Greenwich Village Story To Kill a Mockingbird Black Like Me Free, White and 21 Johnny Cool Lilies of the Field Living Between Two Worlds Nothing but a Man One Potato, Two Potato The Long Ships Year 1959 1959 1959 1959 1959 1959 1959 1959 1960 1960 1960 1960 1960 1960 1961 1961 1961 1961 1961 1962 1962 1962 1962 1962 1962 1963 1963 1963 1963 1963 1963 1963 1964 1964 1964 1964 1964 1964 1964 1964 100 Table 18 (cont.) Titles of Films Featuriag_Black Actora 1950-1980 Title The Streets of Greenwood Cat Ballou Major Dundee None but the Brave Synanon The Carpet Baggers The Greatest Story Ever Told The Hill The New Interns The Pawnbroker A Man Called Adam Blues for Lovers Booker T. Washington Duel at Diablo Hurry Sundown Our Man Flint Patch of Blue Rio Conchos The Appaloosa The Battle of Algers The Bedford Incident The Cincinnati Kid The Girl Nobody Knew The Slender Thread A Time for Burning Doctor Doolittle Dutchman In the Heat of the Night Now is the Time Protrait of Jason Some of My Best Friends are White The Dirty Dozen The Night of the Living Dead The President’s Analyst The Story of a Three Day Pass The Weapons of Gordon Parks To Sir, With Love Up the Down Staircase Dark of the Sun Finian’s Rainbow For Love of Ivy Year 1964 1965 1965 1965 1965 1965 1965 1965 1965 1965 1966 1966 1966 1966 1966 1966 1966 1966 1966 1966 1966 1966 1966 1966 1967 1967 1967 1967 1967 1967 1967 1967 1967 1967 1967 1967 1967 1967 1968 1968 1968 101 Table 18 (cont.) Titles of Films FeaturingaBlack Actors 1950-1980 Title Guess Who’s Coming to Dinner Ice Station Zebra If he Hollers, Let Him Go Joanna Mingus Negro Kingdoms P. J. Robby Salt and Pepper Slaves Split Decision The Biggest Bundle of Them All The Heart Is a Lonely Hunter The Scalphunters The Split The Story of a Three Day Pass Uptight 100 Rifles Ace High Burn Bye, Bye Braverman Castle Keep Change of Mind Death of a Gunfighter First World Festival of Negro Arts Flame in the Streets Float Like a Butterfly, Sting Like a Bee Hello Dolly I Have a Dream Joanna Martin Luther King: The Man and the March No Vietnamese Ever Called Me Nigger Putney Swope Sweet Charity Terry Whitmore for Example The Comedians The Informer The Learning Tree The Lost Man The Rievers The Riot Three in the Attic Year 1968 1968 1968 1968 1968 1968 1968 1968 1968 1968 1968 1968 1968 1968 1968 1968 1968 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 102 Table 18 (cont.) Titles of Filaa FeataaiagABlack Actors 1950-1980 Title Topaz Two Gentlemen Sharing Wild in the Streets Cotton Comes to Harlem End of the Road Halls of Anger Its Your Thing Last of the Mobile Hot-Shots My Sweet Charlie Patton The Angle Levine The Great White Hope The Landlord The Liberation of L. B. Jones The McMaster The Watermelon Man They Call Me Mr. Tibbs Tick...Tick...Tick A.K.A. Cassius Clay Black Chariot Black Jesus Brian’s Song Brother John High Yellow Honkey Man and Boy Right On Shaft Skin Game Soul Soldier Soul to Soul Sweet Sweetback’s Badasssss Song The Ali-Frazier Fight The Bus Is Coming THE Man from C.O.T.T.O.N. The Organization Uncle Tom’s Cabin Across 110th Street Black Girl Black Gunn Black Rodeo Blacula Buck and the Preacher Year 1969 1969 1969 1970 1970 1970 1970 1970 1970 1970 1970 1970 1970 1970 1970 1970 1970 1970 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1971 1972 1972 1972 1972 1972 1972 103 Table 18 (cont.) Titlealof Films FeaturingaBlack Actors 1950-1980 Title Come Back Charleston Blue Cool Breeze Farewell Uncle Tom Georgia, Georgia Hammer Hickey and Boggs Hit Man Lady Sings the Blues Malcolm X Man and Boy Melinda Shaft’s Big Score Soul Soldier Sounder Super Fly The Final Comedown The Legend of Nigger Charley The Limit The Man Top of the Heap Trouble Man A Warm December Black Bart Black Caesar Black Mama, White Mama Book of Numbers Charley One Eye Cleopatra Jones Cobby Five on the Black Hand Side Ganja and Hess Gordon’s War Honey Baby, Honey Baby Jimi Plays Berkeley Lost in the Stars Maurie Savage Save the Children Scream, Blacula Scream Shaft in Africa Slaughter’s Big Ripoff Superfly T.N.T. Sweet Jesus, Preacher Man The Harder They Fall The Hit Year 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 1973 104 Table 18 (cont.) Titlea of Films FeataringyBlack Actors 1950-1980 Title The House on Skull Mountain The Mack The Slams The Soul of Nigger Charley The Spook Who Sat by the Door Three the Hardway Trick Baby Wattstax Abby Amazing Grace Black Belt Jones Blazing Saddles Bone Catch My Soul Claudine Conrack Foxy Brown Hell Up in Harlem Lost in the Stars Sugar Hill That Man Bolt The Black Six The Education of Sonny Carson The Klansmen The Super Cops The Take Thomasine and Bushrod Three The Hard Way Three Tough Guys Together Brothers Truck Turner Uptown Saturday Night Willie Dynamite Aaron Loves Angela Boss Nigger Cleopatra and the Casino of Gold Cooley High Cornbread, Earl and Me Dolomite Friday Foster Let’s Do it Again Mahogony Mandingo Report to the Comissioner Year 1973 1973 1973 1973 1973 1973 1973 1973 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1974 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 1975 105 Table 18 (cont.) Titles of FilaaAFeaturing Black Actors 1950-1980 Title Sheba Baby TNT Jackson Bingo Long and the Traveling All-Stars and Motor Kings Car Wash Countdown at Kusini Drum Mother, Jugs, and Speed Norman Is That You? Rocky Silver Streak A Hero Ain't Nothing but a Sandwich A Piece of the Action Brothers Greased Lighting Short Eyes The Cassandra Crossing The Deep The Greatest Which Way is Up? Blue Collar California-Suite FM Scott Joplin The Boys in Company C. The Wiz Apocolypse Now Richard Pryor in Concert Rocky II The Fish That Saved Pittsburgh Airplane All that Jazz Brubaker Fame Fort Apache The Bronx Stir Crazy The Blues Brothers The Empire Strikes Back The Hunter Wholly Moses Year 1975 1975 1976 1976 1976 1976 1976 1976 1976 1976 1977 1977 1977 1977 1977 1977 1977 1977 1977 1978 1978 1978 1978 1978 1978 1979 1979 1979 1979 1980 1980 1980 1980 1980 1980 1980 1980 1980 1980 REFERENCES REFERENCES p/zAdoni, H. and Mane, 8., "Media and the Social Construction ‘ of Reality; toward an Integration of Theory and Research" Communication Reaearch, Vol. 11, no. 3 July 1984 pp. 323-340. Ashton, C. R.,, The Changing Imaga of Blacks in Aaerican Film, 1944-1973, University Microfilm, Ann Arbor, 1981. Becher, S. L., Pepper, R., Wenner, L. A., and Kim, K., "Viewership of the Debate" in Krause, Sidney (ed.) Tag Great Debates;, Caater_ya. Ford, 1976, Indiana University Press, 1978. Behr, R. L., and Iyengar, 8., "Television News, Real World Cues, and Changes in the Public Agenda" Public Opinion Quarterly, Vol. 49, no. 1, Spring, 1985 pp. 38- 57. Beniger, J., "Media Content as Social Indicators: The Greenfield Index of Agenda Setting" Communication Research, Vol. 5, #4 October 1978. Berelson, 8., Content Analysis in Communication Research Free Press, New York, N.Y., 1952. Blalock, H.,Socialastatistiaa McGraw-Hill. New York 1972. Bogle, D., Toms, Coons, Mulattoes, Mammies, and Bucks: An Interpretive Historypof Blacka in Aaerican Filaa. Banton Books, New York, 1973. Chadwick, D., Bahr, H., and Albrecht, S. gpcial Science Research Methods Prentice—Hall, Englewood Cliffs, N.J., 1984. Chaffe, S., and Hochliemer, J., "The beginnings of Political Communications Research in the United States: The Limited Effects Model" pp. 75-93 in Gurevitch, Michael, and Levy, Mark R., (eds.) Mass Communications Yearbook, vol. 5, Sage Publications, Beverly Hills, 1985. Chatterjee, S. and Bertram, P., Regression Analysis by Example, John Wiley and Sons, Inc., New York, 1977. 106 107 Cohen, B.,The Press, the public,,and foreign pplicy Princeton University Press, Princeton, N.J., 1963. Cohen, Jacob and Cohen, P., Applied Multiple RegressionjCorrelation Analysis for the Behavioral Sciences Lawrence Erlbaum, Hillsdale, N.J., 1975. Cohen, Jacob Statistical Power Analyais for the Behavioral Sciences Academic Press, New York, N.Y. 1969. Cohen, John, Behavior in Uncertainty and its Social Implicationa,Basic Books Inc., New York, 1964. Comstock, G., "Social and Cultural Impact of Mass Media" in Abel, Elie, (ed.) What’s News, Institute for Contemporary Studies, San Francisco, 1981 pp. 233-235. Coppola, P., "On the Director" pp. 50-70, in Baker, F. and Firestone,R. (eds.) Movie People The Douglas Book Corporation, New York, 1972. Cripps, T., Black Film as Genre.Indiana Press, Blooming Indiana, 1975. Cripps, T., Slow Fade to Black, The Negro in Aaerican Fila; 1900 to 194;. Oxford University Press, New York, New York, 1977. Eyal 0., Winter, J., and DeGeorge, W., "The Concept of Time Frame in Agenda Setting". in G. Wilhoit & H DeBoch, (eds.,) Mass Communication review yearbook, Vol. II, Sage Publications, Beverly Hills, 0a., 1981. Funkhouser, G., "The Issues of the Sixties: An Exploratory Study in the Dynamics of Public Opinions" Public Opinion Quarterly, Vol. 37, No. l, 1973 (spring). Candy, 0., Beyond Agenda Setting; Inforaation Subsidies and flablic Policies. Ablex Publishing Co., Norwood, New Jersey, 1982. \/’Gans, H., "The Creator-Audience Relationship in the Mass Media: An analysis of Movie Making." in Rosenberg, B. and White, Da. (ed.), Mass Cultare: The Popular Artaain America, The Free Press, New York, 1957. Gerbner, G., Holsti, 0., Krippendorf, K., Paisley, W., and Stone, P. The analysis of Communication Content John Wiley, New York, N.Y. 1969. Gormley, W., "Newspaper agendas and political elites". Journalism Quarterly, 52(2), 304-308. 108 Gurevitch, M., and Levy, M., (ed.) Mass Communications YearBook, Vol. 5, Sage Publications, Beverly Hills, 1985. Gubak, T., "Theatrical Film" in Compaine, B. M., (ed.) Who Owna the Media. Knowledge Industry Publications, White Plains, New York. 1979. Heron, K., ”Is Movieland Segregated?". Detroit Free Press, February 7, 1982 p. l. Holsti, 0., Content Analysis for the Social Sciences and Humanitiea. Addison-Wesley, Reading, MA., 1969. Huntsberger, D., Croft, J., and Billingsley, P., Statistical Inference for Management and Economics. 2nd. Allyn and Bacon, Inc., Boston, Ma., 1980. H. W. Wilson Company, The Readega:Gaide to Periodical Literature. Vols. 15-40 (May, 1945 to Feb. 1981), The H. W. Wilson Company, New York. r\a/lJowett, G., and Linton, J., Movies as Mass Communication. Sage Publications, Beverly Hills, 1980. Kerlinger, F., Foundationa of Behavioral Research Holt, Rinehart and Winston, New York, N.Y. 1973. Kindem, G., (ed.), The American Movie Indaatry: The Business of Motion Pictures. Southern University Press, Carbondale, 111., 1982. Klapper, J., The effects of Mass Communication, The Free Press, New York, 1960. Krippendorf, K., Content Analysis: An introaaction to its aethodology. Sage publications, Beverly Hills, Ca., 1980. Lang, K., and Lang, G., "The Mass Media and Voting" in Berelson and Janowtz, M., (eds.) Readings in Pablic Opinion and Communication Free Press, N.Y. 1966. Larmatt, J., Elias, S., and Schwortz, F., Analysis and Conclaaioaa of the Washington Task Force on Motion Pictures. U.S. Government Printing Office, Washington, D. C. 1978. Leab, D., From Sambo to Saperspade: The black experience in Motion Pictures. Houughton Mifflin, Boston, 1975. 109 Lees, D., and Berkwitz, S., The Movie Business Vintage Books, Random House, New York 1981. Lerner, D., and Lyle, N., Communication Research-A half Centary Appraisal. University Press of Hawaii, 1977. Lewis, R., "On the Producer", in Baker, Fred and Firestone,Ross (eds.) Movie People The Douglas Book Corporation, New York, 1972. Lumet, 8., ”On the Director" pp. 35-50 in Baker, Fred and Firestone,Ross (eds.) Movie People. The Douglas Book Corporation, New York, 1972. McCombs, M., and Shaw, D., ”The Agenda Setting Function of Mass Media", Paplic Opinion Quarterl , Vol. 36, Summer, 1972 pp. 176-187. McQuail, D., Communication, Longman Group Limited, London, 1975. Merton, R., pacial Thaory and Social Stracture, Collier-MacMillan, Canada Ltd., Toronto, 1957. Molotch, H., and Lester, M., "News as Purposive Behavior: On the Strategic use of Routine events, Accidents, and Scandals", Aperican Sociolggical Review, Vol. 39, No. 1, 101-112. Monaco, J., Ameriaan Film Now, The New American Library, New York, 1984. Murray, J., To find an Image: Black filaa_from Uncle Toafito Superfly, Hobbs-Merrill Inc., New York, 1973. Nafziger, R. and White, D. (eds.) Introduction to Mass Communication Research. Louisiana State University Press, 1958. Patterson, T. and Mc Clure, R., The Ugaeeing Eye. G.P. Putman, New York, N. Y. 1976. Picker, D., "On the Distributor" pp. 23-34, in Baker, F. and Firestone,R. (eds.) Movie People The Douglas Book Corporation, New York, 1972. Protess, D., Leff, D., Brooks, 8., and Gordon, M., ”Uncovering Rape: The Watchdog Press and the Limits of Agenda Setting", prlic Opinion Quarterl , Vol. 49, No. l,Spring, 1985 pp. 19-37. 110 Roberts, D. and Bachen, C., "Mass Communications Effects" in Rosenzweig, M., and Porter, L, (eds.)Annual Review of PsychologyL Annual Reviews, Palo Alto, Ca., 1981 pp. 307-356. Roscoe, J. T., Fundamental research statistics for the behavioral aaiencea. Holt, Rinehart and Winston, New York, 1975. Rosse, J., "Mass Media: The Economic Setting" in Abel, Elie,(ed.) What’s News Institute for Contemporary Studies, San Francisco, 1981 pp. 233-235. Scott, W., "Reliability of Content Analysis: The Case of Nominal Scale Coding. Pablic Opinion Qaarterly, 17, 1955, 321-325. Schram, W. "The Challenge to Communication Research," in R. O. Nafziger, and D. White (eds.) Introdaction to Mass Communication Research, Louisiana State University Press, Baton Rouge, Louisiana, 1958. Servin, W., and Tankard, J., Communicationa Theorieap Origin , Methodaa Uses. Hastings House, New York, 1979. Shaw, D., and McCombs, M., The Emergence oprEericaa Political Issues: The agenda Setting Fanction of the Press. West Publishing Co., St. Paul Min., 1977. SPSS Inc., User’a Gaide: SPSSx, 2nd. edition, SPSS INC., Chicago, Illinois, 1986. Steinberg, C., Film Facts, Facts on File, Inc. New York, N.Y. 1980. Stuart, F., Theaaffecta of Televiaion on the Motion Picture and Radio Indaatriag. Arno Press Cinema Program, Carleton University, 1960. Tan, A., Mass Communication Tagories and Reseapch. John Wiley and Sons, New York, 1981. Thorndike, R. M., Correlation Procedures for Research. Gardner Press, New York, N.Y. 1978. Tipton, L., Haney, R., and Baseheart, J., "Media Agenda Setting in City and State Election Campaigns" Journalism Quarterly, Vol. 52, No. l, 1975, pp. 15-22. Variety, Annivegaary Issues, January 1950 to January 1980. lll Walizer, M. H., and Wienir, P. L., Reseaach methods and Analysis, Harper and Row, New York, N.Y. 1978. Weaver, D., "Political Issues and Voter Need for Orientation" in M. McCombs and D. Shaw (Eds.) The Emergence of Aaerican political issues, West Publishing, St. Paul, MN., 1977. Westley, B., and MacLean R., "A Conceptual Model for Communications Research" Journalism Quarterly, Vol. 34, Winter 1957 pp. 31-38. Williams, F., Reasoning with Statistics, Holt, Rinehart and Winston, New York, N.Y. 1979. Williams, W., and Semlak, W. "Campaign ’76: Agenda Setting During the New Hampshire Primary", Journal of Broadcasting, Vol. 22, #4, 1978 pp. 531-541. Williams, W., Agenda Setting Research in Dominick, Joseph R., and Flitcher, James F., (ed.) Broadcast Research Methods Allyn and Bacon Inc., Boston, 1985. Wilson, C., and Gutierrez, F., Minorities and Media: Diveraity and the End of Maaa Communication. Sage Publications, Beverly Hills, 1985. Winkler, R. and Hays, W. Statistica: Probability, Inference, and Decision. Holt, Rinehart and Winston, New York, N.Y. 1975. Woodrum, E., "Mainstreaming Content Analysis in Social Science: Methodological Advantages, Obstacles, and Solutions." Social Science Research, Vol. 13, no. 1 March 1984 pp. 1-19. Wright, G., (ed.) Behavioral Deciaion Making, Plenium Press, New York, 1985. Zucker, R., "The Variable Nature of News Media Influence" in B. D. Ruben, Ed. Communication Yearbopa, 2, New Brunswick, N.J. Transaction Books, 1978. “IIIIIIIIIIIII'IIIIII