. »- W“ W ,W'EU‘KR «w... a” ”it“ 33. “.5.” 1%" 'N'.“-’§".. - ., HQ IO‘: 'fié‘l“ :I.’ :fifi wfiv‘zfi‘? (Rut)? I' i 9' ' d VS???“ . " TI‘PAI—t-W-II ‘1‘» ‘ . éfifl'.“ k‘t {iii} .2?“ h ‘irtt‘AwI ”3c :‘r 9).; 1‘\I;HI\‘N:I:I Mign‘h. H'- l.:\ If‘.’y I '31.“; .. "“ “"' "“ << SIP. ?&?u:§??¥ yg§r :yp'fih‘ \tI‘MI,|;‘ )\:'.'. -' I~....I‘ .ng §§*g’.92§{‘ - !!!! I ‘, ‘ ;.‘;' 3553" .- “I v 5I'Jt:~.~.—I ,.'M¢'m‘ ‘:“I::' ;:r W," "t...“ tux-14%! ‘ {-1. . I Jl\“"’>" . . 33' "I,” , 'I " ~ - "'1' 'III .2 I ‘. I I . .. I’I .I .. I.‘ . ‘I 'g .I f. -~I I. I ‘ “I|'(I \' 'II .4 ‘ -I-' I‘ll ‘ I .a 4 I. I I I' , IV 1 I '-I: l.‘\ -" \ f- u ‘f- " | LI}. ..'ll ()3 I' .. b‘)” . \J t. ‘I . ..' > (I - u. I. '1'I' h)... 3; II . I v . I. I/,I I"I .: If.” I ”é . I I II I“ ..,,, (I "\I‘H .II 5‘1- m... 1.. y." :LA‘I:I.)- >72?- 'I'; u 951:.- .:i;'I':‘.: I'; . . .13 Harms. '.IJ.'I':“;:II‘.'\.1‘I“I. wk % "“12“,!“ ' 3" h '34" .I‘Hjt‘s'fl ‘IIWL PI ' I I1'P|"'p|lt .' :9 ' . ., I Mi"- ' -?'. "" - {”3335 \,}.~-;I- ‘ 'MW'Q. 3. [-3. 16¢“. “‘. '-‘ I. 10“”; W. (N: ff“ ' I .‘ " ' Ifllfiflfifié’l " ‘ . ’Iv :‘ '~\ - I Adi. ‘5’: z:?'T Iv"{‘?b fifi‘ ‘fi. $§§£§II II Ikfip'u ‘ L. .D -.€‘A w ‘i F I v I .JIII'II:‘:I..I I- 'IV -.I ‘IH’E' f." .. . ‘9.“ ‘. I "f ‘ ”A WWI. 'd I l y I I I I I I -. -. \;)."l u‘c‘.‘ ”Irv-I "‘4, ANN “Vi “1-“ ,0 I}. “"A “.1534 I . “UH. ”:I‘i . I I _,'. 5"” "3 ‘ rIwaJ.J . , ’ . ‘.. ""..Ir.:. I‘. 'II . ‘. .‘. .,\,'\_ 1 . I [afhfichmvhhn , 3.; .. .II‘II‘IIi . U I . 'Ia.‘1)..'. I I 1:”? ‘IL ”3ij ' \ IHI‘M- I- 3.1.9. I ' “J ' ‘ "L ‘ -’ -Ul‘ok'lk‘q.' "II‘ . "' “' I'o \} I ”.1.” ‘Ik‘fl . ' H ' t "I "i. “J" I" 1}" ::-|':‘“"l> 3:; ;;1.‘|'/:r.r'~\‘l.,tl“kl . I .I' ‘. ‘ ...I 'I‘ ." _:.‘.. .I ‘ I 'I ‘5'. 1h;.1'?,:.q.v 1.: . .Ei. .','\I,,I‘."<'i)’.‘.":II .135. I‘ :q‘..'.. ”I "I 'I :hh" 38'; _, . .H, I.. IMIIQI ' II" I III I .4 .I.. I “.I VI'II I‘I.‘ ‘t'.’ W" 3“?“ " ‘ . T» ”I" 0.”, ,h‘ :WIIN‘.‘ “III-‘9)”- NW! ‘. ".I...".‘ J‘ImI' .. I‘I‘ ‘ .1. .rfi'o Inn-I, .I'l.'.".l;llt ‘ wnlocll "I‘ ‘3 .| “ 1. II‘,‘ 1.:l:'_" :‘" I'I I'll. ’IIIIIIX 'I '3“. .‘II_'( . ‘ ‘3" IJII\III-‘I:II>'I"I"" ”our-T: I 1‘! ..I.-'-.I:- . I. ‘ ‘ » ”va .IIIMIII' ‘-"I.‘ ..‘=I-::: 5‘ ..I I .. ~ ~ ' .I * . . .. h . . .-I,...~:$ l ' . If 93%.“ 2%} 1,”...- j: . (5| . . . .. L-- "i ' T3964 .7 I . I I I I ."I'. H .U — ,I‘ g I h II':.I' #‘ .‘l ' ‘L‘. 21.52%“: '.‘.‘ 1:11;:- . I ). - ‘ ‘ > n Ikl‘ 'r' ‘9'. 5,1. ' ; ' ‘ A3'HI thil .‘ \..'Il_'l. I‘. ‘2'?" (Hahn-u ‘ I .III \II‘ If -: . , ,I'a m "I‘.:;§:. 7f. ' IP‘ ;gll "m ‘I: ;':L|‘ WI. '1 I I. I . I'I"I| :‘j' ' I”. Ill" ' ' .‘g.’ 7’ ' ' V q I.‘ I I. ‘ , " #PL It :"H'I: any I';":-I'Ill1' “I?! :1" 'F “will“. .I ‘ \I ‘ .. ..‘ . . “‘ .I “I ‘ ' HI I”... I” II '-» 2’ ' ‘l I “I“ ('1 ‘ . , . " . ‘ I‘ I.‘ :1' I‘ > "' "-I ' I H I " "1(le- I|"‘ ' “'1'" I] ? r5” ',‘:'I"l|{?'?:.w .\. I I I‘- k. I ",‘IN‘Iy' '-"~| . 'I ‘ "III '1 " '. I?“ ‘:' l: .. "‘- l' “.5‘ l“ I |||| {II ”IJI‘ II .I: .l I I “I H ‘. I“ IL | I .‘)"" .(H ' ”ll" '|"I ' I): ' ’I |I. I " (I . II" P 'u' 'I‘ I‘” I I {III}. " .. J I '.. H ‘1 [II I .' ‘ o. ‘ ‘. I . “ III», . :‘m I | I: 'Iz‘l I I I I . I ' H . x W!" ','I M) ‘ I I ' 9“} I ' ! I II' III .III In”: In! LIBRARY ,. Michigan State rm‘m" . University This is to certify that the thesis entitled THE EFFECTS OF IMAGE SIZE ON JURORS' PERCEPTIONS OF A VIDEOTAPED DEPOSITION: A MEDIUM SHOT IN THE DARK presented by RONALD G . HALBE RT has been accepted towards fulfillment of the requirements for M.A. Conmunication degree in M am Major professor Date 8'3'78 0-7639 THE EFFECTS OF IMAGE SIZE ON JURORS' PERCEPTIONS OF A VIDEOTAPED DEPOSITION: A MEDIUM SHOT IN THE DARK BY Ronald G. Halbert A THESIS Submitted to Michigan State University in partial fulfillment of the requirements .for the degree of MASTER OF ARTS Department of Communication 1978 Accepted by the faculty of the Department of Communication, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Master of Arts @2 r 2 éd ; ge gag/Egg z, rector 0 es 8 degree. Guidance Committee: ABSTRACT THE EFFECTS OF IMAGE SIZE 0N JURORS' PERCEPTIONS OF A VIDEOTAPED DEPOSITION: A MEDIUM SHOT IN THE DARK By‘ Ronald G. Halbert This is an exploratory study of how various camera shots may affect a juror's opinion of a witness testifying in a civil suit via videotape. Prior research has provided inconclusive evidence that certain attributes of a video- taped subject may interact with the type of shot to differ- entially influence ratings by viewers, and perhaps bias a verdict. This study controlled for the attractiveness and sex of a witness by videotaping a male and female pretested as attractive, and a male and female pretested as unattractive. Each actor presented forty-five minutes of testimony in a close-up shot, medium shot, long shot, and extreme long shot, resulting in a 4 x 2 x 2 factorial design. A total of 215 Lansing area adults were each randomly assigned to view one of the sixteen videotaped experimental conditions and complete a questionnaire measuring the following variables: informa- tion retention, identification with the witness, source credibility and homophily, interpersonal attraction, and interest in the proceedings. Ronald G. Halbert Data analysis was done using three-way analyses of variance with a regression approach, in conjunction with the Tukey method for post hoc comparisons of means. The results indicated that the composure dimension of credibility ‘was significantly influenced by an interaction between camera shot and witness attractiveness. Physical attraction was affected by an interaction of shot and witness sex. Viewers' ratings of interest in the proceedings were also affected by the type of camera shot. Based upon these results and the findings of prac- titioners and past researchers, the medium shot was recommended as the most useful for videotaping a single, seated witness. ACKNOWLEDGMENTS Although these words are small tribute to the many people who made this research possible, I will attempt to express my gratitude for their time and effort. I wish to thank the members of committee, Dr. Gerald Miller, Dr. Katrina Simmons, and Dr. Thomas Muth for their patience and skill in guiding me past the pitfalls of scien- tific research. I owe a special debt to Dr. Miller and Dr. Norman Fontes for having the confidence in me to provide the resources and training necessary for this project, and for setting an example of research excellence. I am.also indebted to Ed Kaminski, Frank Easter, and the whole V.I.L.E. team.for all their labors and good advice, but especially to Joyce Bauchner for her invaluable assistance in the design and execution of the experiment. There are many other comrades and colleagues whose contributions are less obvious but still significant. Thank you Rob, Nancy, Eric, Laura, and Steve for putting up with me, and occasionally putting me up. Credit should also go to film and video researchers like Thomas McCain and Jack wakshlag for their advice and pioneering work, and to Sergei Eisenstein for starting it all. Appreciation goes to people like Thomas Gordon for showing me what satisfaction there is in discovery through research. ii I extend deepfelt gratitude to my parents, Leon and Sylvia, for supplying the right mixture of guidance, independence, and love. There is one other person whose love gave me the strength and determination I needed. Thank you, Lydia, for showing your love by listening and understanding, comforting and encouraging, typing and editing, but mainly by just being there and believing in me. iii CHAPTER II III IV TABLE OF CONTENTS RATIONALE AND REVIEW OF LITERATURE . Review of Literature . Practitioners . . Empirical Research Summary . Research Questions METHODS AND PROCEDURES . Procedures . Definitions . . . Statistical Design . RESULTS . Manipulation Checks Research Questions . Summary . DISCUSSION Experimental Manipulations . Research Questions . Limitations . . Implications . Recommendations . Summary . FOOTNOTES . APPENDIX A. Viewer Questionnaire H B. Three-way Analysis of Variance Tables for Non-significant Dependent Variables . C. Diagram of Studio Set- -up for Videotaping the Deposition . . BIBLIOGRAPHY . iv . 106 . 109 . llO TABLE 10. ll. 12. 13. LIST OF TABLES Rotated Factor Loadings for Identification Items . . . . . . . . . . . . . . . . Rotated Factor Loadings and Reliability for Mediated Source Credibility-Homophily Items (Principal Components Factor Analysis; Varimax Rotations) . . . . Rotated Factor Loadings and Reliabilities for Combined Attorney Credibility-Homophily Items . Rotated Factor Loadings and Reliability for Mediated Interpersonal Attraction Items . Means and Standard Deviations of the Two Dimensions of Physical Attraction for Attrac- tive and Unattractive Female and Male Witnesses . Summary Table for Analysis of Variance for Homophily Dimension of Attorney Credibility . Means of the Homophily Dimension of Attorney Credibility for Image Size . . Summary Table for Analysis of Variance for Information Retention . Means of Retained Information for Image Size and Witness Sex . . . . . . . Summary Table for Analysis of variance for Identification with the Witness . Means of Identification with the Witness . for Dmage Size and Witness Attractiveness Summary Table for Analysis of Variance of Composure Dimension of Witness Credibility . Means of Composure (Dimension of Witness Credibility) for Image Size and Witness Attractiveness . . . . . . V 39 41 44 51 53 53 55 55 56 56 5:8 58 LIST OF TABLES (Continued . . .) TABLE 14. 15. 16. 17. 18. 19. B1 BZ BB B4 BS B6 Summary Table for Analysis of Variance of Physical Attraction Dimension of Inter- personal Attraction . . . . . . . Means of Physical Attraction (Dimension of Interpersonal Attraction) for Image Size and Witness Sex . . . . . . . . . . Summary Table for Analysis of Variance for Physical Unattraction Dimension of Inter- personal Attraction . . . . . . . . . Means of Physical Unattraction (Dimension of Interpersonal Attraction) for Image Size and Witness Sex . . . . . Summary Table for Analysis of Variance for Interest in Proceedings . Means of Interest in Proceedings for Image Size . . . . . . . . . . . Summary Table for Analysis of Variance of Sociability Dimension of Witness Credibility . . . Summary Table for Analysis of Variance of Competence Dimension of Witness Credibility . Summary Table for Analysis of Variance of Character Dimension of Witness Credibility Summary Table for Analysis of Variance of Perceived Homophily . . . . Summary Table for Analysis of Variance of Social Attraction Dimension of Interpersonal Attraction . . . . . . . . . . . . . Summary Table for Analysis of Variance of Task Attraction Dimension of Interpersonal Attraction . . . . . . . . vi 60 62 62 63 63 . 106 . 106 . 107 . 107 . 108 . 108 FIGURE ‘1. Cl. LIST OF FIGURES Geometric Dispay of a 4 x 2 X 2 Factorial Des ign O O O O O O O O O I O O O O O O 0 Diagram of Studio Set-up for Videotaping the Deposition . . . . . . . . . . . . . vii 109 CHAPTER I RATIONALE AND REVIEW OF LITERATURE The advent of video technology into the courts has focused attention upon a relatively unexplored area of media research. Before videotape recording of depositions, and entire trials, can gain widespread acceptance, legal experts and social scientists must assess the potential effects upon a legal system that has been reluctant to change for hundreds of years. Rather than contending here with the controversy of advantages and disadvantages of videotape versus an all live trial (see Kaminski, 1977), we will examine an aspect of videotape usage--techniques used to record testimony. As part of the controversy, much discussion centers on what production techniques should be specified for testi- mony recording. As videotaped depositions replace live testi- mony, the camera and microphone intervene, or mediate, in jurors' acquisition of information needed to reach a verdict. If mediation by these electromechanical devices introduces excessive bias due to recording methods, then the jurors may be basing their decisions upon information significantly different from that available in a live presentation. Although a microphone will record non-selectively the same sounds that would be normally heard in a court, the video 1 2 camera has certain inherent limiting properties. Mbreover, the recording of visual events is complicated by the wide number of decisions made by the camera operator (or the person directing him.or her). Discounting any special effects, the most basic camera selections include: camera location, angle of view, focal length, aperture size, compo- sition, and framing. Gerbner (1964) stresses the necessity of being more conscious of these production variables: We are not always aware of the existence or nature of these more subtle elements . Manipulation of these elements can therefore lead to changes of perception (meaning) with relatively little awareness of manipulation. (p. 253) The question confronting researchers is whether or not certain production "manipulations" present a danger to the concept of a fair trial. Bermant, McGuire, and Chappell (1975) warn us that "the techniques of film and television art will soon become applied to videotaped depositions and testimony" (p. 8). Even the novice television directing student knows of simple ways to transform a lifeless actor into a dramatic hero. Kaminski (1977) points out that current rules govern- ing production equipment and methods are minimal. "It is conceivable that if the costs of losing the case were high enough, and the defendent or plaintiff has the money, some depositions could become M.G.M. productions" (p. 6). Optimally, guidelines should originate from research findings. To date, enough research has not been conducted to allow court systems to generate specific suggestions for 3 permissible camera techniques. Due to the nature of the material being recorded, it becomes increasingly important to determine the degree of objectivity of videotaping. Given the vast array of possible production tech- niques, it was necessary to limit the scope to selections made most frequently. Whenever a deposition is recorded, the camera operator has to decide how to present the witness within the limited size of the television frame. The sub- ject's image size relative to the television screen is re- ferred to as a "camera shot, or also the "image size" (throughout this study, the two phrases will be used inter- changeably, but always with reference to the same concept). Camera shot is a function of the focal length of the camera lens, and the camera-subject distance. There are currently four camera shots useful for videotaping testimony:1 (I) Close-up Shot -- tight focus on the head and shoulders of one focal in- dividual with nobody else in the frame. (2) Medium Shot -- view of not more than two focal individuals from.the waist up. (3) Long Shot -- view of entire bodies of multiple focal individuals. (4) Extreme Long Shot panoramic view of multiple focal individuals and sett- ing, with any given person occupying only a fraction of the frame. The decision of which type of shot (or combination of shots) to use is an important one because of the influence a particular shot may have upon the amount and type of 4 information that is recorded. Both social scientists and legal experts disagree on the type of shot that should be used when recording testimony. Doret (1974) typifies the ambivalence that commonly occurs when choosing a camera shot. He suggests that the panoramic view (extreme long shot) provides information that "deviates least, in terms of the visual field offered the jury, from the traditional trial and offers the jury the widest possible universe of sensory data to formulate his impressions upon." However, the panoramic view cannot "capture the nuances of the de- ‘meanor of the witness" (p. 228). Conversely, a close-up view of the witness could supply greater detail and definition of the witness's face than could be seen by a jury sitting in a courtroom, but it would severely restrict the normal field of vision. Since only the witness is visible in a close-up, the gain in facial information might not compensate for the inability to observe the behavior of other trial participants. Mention has been made of only some of the consider- ations that require careful examination before the courts can set guidelines for videotaping a deposition. The Speci- fic purpose of this thesis is to determine what effects, if any, different camera shots have upon jurors' perceptions of a videotaped deposition and the testifying witness. Review of Literature The remainder of this chapter will be devoted to an examination of literature relevant to the use and effects of camera shots. Also included in this review will be literature discussing other variables that mediate the effects of camera shot. Some of the sources cited will be referring to motion picture production rather than video. This analogi- cal approach was felt justified since any discussion of video or television production techniques will necessarily find its roots in the cinema. Although some terminology is exclusive to one medium.or the other, the essential elements of theory and technique are shared. According to Roy Medsen (1973), "Electronic photography for videotape requires the same principles of composition and perception as does motion picture cinematography" (p. 125). In reviewing the literature regarding the effects of different image sizes upon television and film audiences, one soon notices an obvious dichotomy of authors. One group con- sists of media practitioner-theorists who base their state- ments upon observation, personal experience and intuition. These writers suggest when certain shots should be used and 6 what effect may be expected, although they rarely offer reasons why a given effect occurs. The other group is com- posed of researchers who base their conclusions on the results of data collection and analysis. Both sources will be cited, making it clear to the reader into which camp a particular author falls. Practitioners D;W. Griffith is considered one of the earliest fihm- makers to make use of camera shots other than the long shot, standard in the first years of cinema. Although he supposedly first used the closeup in an effort to save money (Montague, 1964, p.100), he soon realized its great cinematic potential. In a 1908 film, Griffith used a close-up of the heroine's face to prepare the viewers for the succeeding shot of the object of her thoughts. The close-up served to bring the audience in closer and then closer still, into her imagina- tion and emotions (Kracauer, 1960, p.47). Ever since Griffith's revolutionary breakthroughs filmmakers have been conscious of selecting the best image size to bring about the desired audience response. Eisenstein (1965) encourages the purposeful use of different image sizes within a "montage" (i.e. a rapid succession of informative vieWpoints where the whole is greater than the sum of the parts). Other than simply making things larger, he indicates that the close-up has a distinctly different effect than the long shot. By excluding details that compete for the atten- tion of the viewer, the filmmaker guides the audience to the 7 important subject and gives it increased connotative mean- ing. To better understand how long shots, medium shots, and close-up shots differ, it would be helpful to discuss them from a "content transfer" point of view as suggested . by Rod Whitaker (1970). Considering film as a language, he defines two classes of content: facts and attitudes. Facts may be conceptualized as "those aspects of the film message that can be given a noun label, even outside the context of the film, and without reference to the subjective reading of the film viewer." Facts may be photographically simple such as a "man, bottle, window and tree, or complex, such as "marriage, urbanization and economic depression" (p. 42). Attitudes are the subjective aspects that cannot be photographed directly and depend upon the emotional state of the viewer. Through a cinematic arrangement of facts, the audience is led to perceive attitudes such as "good and bad, love, and emotional depression" (p. 42). Whitaker suggests that due to the much larger field area, the longer shot is a much better carrier of fact than the close shot. This increase in facts reduces the attitude potential, allowing the audience to view the facts from a more detached point of view. Conversely, the close shot excludes distracting facts and emphasizes affect displays. Its limitation as a carrier of fact increases its potential as a carrier of attitudes. Also, distance acts as "emotional 8 insulation," both limiting viewer involvement in the long shot, and increasing it in the close-up. The ability of selected shots to transfer factual or attitudinal content offers an explanation why film and television theorists discuss the close-up shot as heighten- ing emotion, and the long shot as primarily expositional. Over the years, a "grammar," or rules of production, has evolved which incorporates these ideas. Madsen (1973, p.87) refers to the basic editing approach as: long shot--medium shot--close-up shot--long shot. The long shot supplies the full view of the actors and the setting. The medium.shot is used to focus attention on a character performing some activity relevant to the plot. The close-up provides a more detailed and intimate look at the person revealing a mood or feeling. Then the camera returns to the long shot to show a new character or dramatic action. In all shots, each new view should yield some information not readily available in previous views. Each shot seems to serve a particular function, corresponding with fact-attitude limitations. The long shot is most effective when providing an opportunity for the audi- ence to familiarize themselves with spatial relations within a specific setting. But, as Roberts and Sharples (1971) warn, the long shot should not be needlessly maintained or used without reason since it decreases control of the audience's attention. This would imply that Doret's suggestion of a panoramic view would suffice to establish locale, but may sacrifice interest. On the other hand, Millerson objects to overlong use of the close-up, claiming that a close-up fragment held too long may seem to be detached from.the rest of the scene, and "this can cause us to lose our orientation, sense of location, or forget the relationship of the portion to the whole" (1972, p. 224). In a later writing, Millerson (1976) advises that screen filling views are dramatic for the presentation of people, but skin imperfections and dental problems will become more noticeable (p. 32). Although it cannot be specified which shots are most appropriate for viewing testimony, certain uses of the close-up, as well as longer shots, will likely have a detrimental effect on the production as an objective record (i.e. a representation that does not introduce significantly more bias than the live version). So far, discussion has centered on some of the more generally accepted theories that apply to both cinema and television. When we consider the differences between viewing a filmed deposition and a videotaped deposition, one major dissimilarity is apparent. Unless a court has access to a modern video projector, the screen size of the television will be considerably smaller than the average sized 16mm movie screen. Due to this smaller reproduction size, Zettl (1968, p. 420) claims that close-up and medium shots are more beneficial because they enlarge objects relative to the screen. Millerson (1976) suggests that for television use, 10 longer shots function best to reveal a location, estab- lish a mood, or follow action. . . . In practice, the type of shot that predominates varies with the kind of produc- tion" (p. 30). Although it would depend upon the nature of a particular deposition, it would seem.that the closer shots would be better for the average single witness testi- mony. This review of relevant statements by practitioners indicates that different image sizes may play an important role in recording a deposition. Their conjectures suggest that close-up shots and medium shots of a witness would pro- vide increased picture clarity and more attitudinal informa- tion. However, the long shot supplies a larger field of view thus providing a truer approximation of the total scene. In the widest view, the extreme long shot, the juror would have the greatest choice for focusing attention upon different trial participants, although the amount of perceivable detail would be minimal. Depending upon the shots selected, the jury could receive an overabundance of factual informa- tion and not enough insight into the attitudinal data that would contribute to a verdict, or vice versa. Empirical Research The discussion now turns to a review of several studies that are the product of systematic data collection and statistical analyses. The first pair of studies focus on the effects of image size on perceptions of source ll credibility and interpersonal attraction. Two other studies try to isolate specific variables that might interact with image size: (1) body type and sex, and (2) acting style. A more general discussion proposes that speaker physical attractiveness might also be of heuristic value for examining image size's effects. This section closes by considering two dependent variables, identification and interest, that might be influenced differently by different camera shots. Image size. A series of studies conducted at Illinois State University contributes the best evidence of the influ- ence of image size on viewers' perceptions of a mediated pre- sentation. McCain and Repensky (1972) manipulated camera shot to determine its effects on perceived interpersonal attraction of two comedy performers. Three cameras simultan- eously recorded a close-up, medium, and long shot of two pro- fessional comedians doing their routine. Interpersonal attraction was operationally defined as a three-dimensional construct consisting of physical attraction, social attrac- tion, and task attraction. Although there were no significant differences in social attraction between different shot conditions, physical and task attraction were both influenced by particular combi- nations of shot and comedian. One explanation for these interactions is that the impact of image size may be affected by certain physical characteristics of the videotaped subjects. For instance, the taller of the two comedians was judged more physically attractive in shots emphasizing the height 12 discrepancy (the medium and long shots). The authors also suggest that the differences observed in task attraction may be a function of the funny man-straight man relation- ship of the two comedians. If this is true, then variation in the physical appearance of two witnesses may produce differential perceptions by the jury, due to shot emphasis. Some of the problems encountered in this initial attempt were eliminated in a study by wakshlag (1973) who examined the effects of camera angle and image size on inter- personal attraction and source credibility. Since few tele- vision shows ever use a single static shot of an actor, wakshlag decided "that each treatment should have a pre- ponderance of the desired camera angle and image size which would be complimented by switching back and forth to a refer- ent angle and image size" (p. 17). Dmage size treatments consisted of a preponderant medium shot with a close-up refer- ent, and a preponderant close-up with a medium shot referent (referent shots were shown 25% of the time). Stimulus sub- jects were two male broadcasting students delivering a three minute newscast. wakshlag administered semantic differential credibil- ity scales designed by McCroskey (1971). They factored into four dimensions: dynamism, composure, character, and socia- bility. Comparisons of image size produced significant differences only on sociability, where the medium shot received higher ratings than the close-up. wakshlag contended that the stoic expressions of the newscasters were uncomplimentary 13 in the emphasis of the close-up (p. 41). The same interpersonal attraction scale was used as in the McCain and Repensky study. Physical and social attraction showed higher ratings in the medium.shot than in the preponderant close-up. wakshlag was unable to account for what particular characteristic(s) of the newscasters produced the difference. He concludes that three factors interact with image size: "1. The image size(s) with which it is juxtaposed; 2. How long the viewer is exposed to each; (and) 3. The object(s) and/or person(s) in the sequence" (p. 44). These two studies indicate that the image size of a mediated source affects a viewer's interpersonal judgments of the stimulus subject. It may be inferred that some physic- ally observable attribute(s) of a witness might also be differentially perceived in various shots, but there is in- sufficient information to hypothesize what it (they) may be. The next study to be discussed determines if somatotype (body type) is possibly one such attribute. Somatotype (body type). A third study from.Illinois State University tested for the effect of body type and camera shot on interpersonal attraction and credibility. To control for sex differences a male and female of each of the three body types were viewed: ectomorph (thin), mesomorph (muscular), and endomorph (fat). These six speakers pre- sented a neutral message and were videotaped in three views: a close-up shot, a medium shot, and a long shot. This l4 resulted in 18 experimental conditions, or a 3 x 2 x 3 factorial design. Although there were no main effects of shot (i.e. all speakers combined), camera shot and body type inter— acted to produce significant differences in scores for physical attraction, and the sociability dimension of cred- ibility. These differences were attributed to availability of body type information due to variations in camera-subject distances. The authors suggest that the medium shot pro- vides the most detailed VIBW’Of body type, since the close- up is restricted to the face only, and the distance of the long shot makes detail harder to see. They feel that the ‘medium.shot is the most likely to receive ratings corres- ponding to Viewers' predispositions toward particular body types. The validity of these assertions is questionable. This study fails to control for or measure other potentially mediating variables such as speaking style and facial attrac- tiveness. Since speakers were chosen by experimenters, rather than randomly selected, the results may not be general- izable to any other set of speakers. This same limitation applies to the prior two image size studies as well. Speaker sex. McCain and Divers (1973) found in post hoc tests that speaker sex interacted significantly with image size to produce unhypothesized effects on the depend- ent variables. Sex also had influence as a main effect, in interaction with body type, and in three-way interactions 15 with body type and image size. The impact observed in this study caused the researchers to advise that the sex of the stimulus be controlled for in future research. Acting style (expressiveness). Another attempt at isolating mediating characteristics was made by wurtzel and Dominick (1972), who varied the degree of expressiveness of their speakers. The researchers compared television acting style with stage acting style to determine what effect differing camera shots would produce. In two of the experi- mental treatments three male actors used a normal television acting style. In the other two videotaped treatments the actors used the more exaggerated gestures and facial expres- sions associated with stage acting. For each acting style one videotape consisted of close-ups and medium shots, and the other tape only medium shots. An interaction of camera shot and acting style supported the hypothesis that close-ups in conjunction with a low key acting style will be evaluated more positively than a close view of stage acting, This would suggest that a wit- ness who expresses him/herself with too much intensity would suffer in a close-up view, whereas a more introverted speaker would be rated lower in a medium shot. Physical attractiveness. All the studies cited thus far seem.to suggest that physical appearances and behavior of a speaker interact with image size of a videotaped pre- sentation. Surprisingly, no studies have controlled the physical attractiveness of the actors. Past research in areas of persuasion (Mills and Aronson, 1965), reactions to 16 evaluations (Sigall and Aronson, 1969), and person percep- tion (Dion, Berscheid and Walster, 1972; Sigall and Landy, 1973) seem to point to a strong influence of the source's physical attractiveness on communicative interactions with others. In fact, Dion et a1. maintain: The results suggest that a physical attrac: tiveness stereotype exists and that its content is perfectly compatible with the "What is beautiful is good" thesis. Not only are physic- ally attractive persons assumed to possess more socially desirable personalities than those of lesser attractiveness, but it is presumed that their lives will be happier and more successful. Landy and Aronson (1969) tested the sentencing of defendants described as attractive, neutral, or unattrac- tive, and found that the unattractive one received a signi- ficantly longer sentence. However, according to Friend and.Vinson (1973), jurors committed to being unbiased may overcompensate, resulting in more leniency for the unattrac- tive defendants. In all the studies reviewed, the degree of perceived attractiveness exerted an influence on inter- personal evaluations. It may be that a particular camera shot that makes a witness appear more, or less, attractive could add a bias to jurors' decisions. Identification. One of the main goals of any film or television director is to present the protagonist so that the viewer can identify with the characterization. Madsen (1973, p. 25) maintains that one of the most crucial elements of a film.ar television production is its "sub- jectivity,‘ or the potential for the audience to metaphoric- ally "see" events through the eyes of the character. The l7 viewer is expected to be less aware of self and identify more with the lives portrayed on the screen. The importance of this concept to mass media research is emphasized by Weiss (1968) when noting the great attractiveness of heroes. He states: Basically, "identification" refers to a per- son's involvement in the depicted events through a psychological relationship with one or another of the participants. . . . As a result, he participates vicariously in the events, feelings, and behavior that relate to the object of identification, and experiences the communication more personally and deeply. (p. 98) This "psychological relationship" is demonstrated in research by Tannenbaum and Gear (1965) who found that subjects who identified with the endangered protagonist exhibited more stress arousal to stressful film endings than subjects who rates themselves less similar to the protagonist. Identification with a source is related to the source's physical attractiveness. Landy and Aronson (1969) propose that jurors may find it easier to identify with an attractive or neutral defendant than an unattractive one be- cause it is ". . . easier to imagine themselves involved in a similar situation . . . because they had potentially more in common with the defendant" (p. 151). Other attributes of a character have been suggeSted as contributing to identification potential, such as sex similarity of character and subject (Maccoby and Wilson, 1957; Wall and Simonson, 1951), rewarded behavior (Kagan, 1958; Zajonc, 1954), and Viewers' aspired social class l8 (Maccoby and Wilson, 1957). In addition to these variables, the actual audio-visual presentation could influence percep- tions such that identification may be increased or decreased. Although there is a paucity of research linking image size and identification, many practitioners claim that the choice of shot plays an essential role in bringing the audience emotionally closer to the screen reality, Russ and Silverstein (1968) state that "close shots, which peer at objects, create intensity; long shots, by their distance, imply detaChment" (p. 116). Kracauer (1960) uses similar phrasing when he comments on one of D.W. Griffith's emotion- laden close-ups as "intensifying our participation in the total situation" (p. 47). This can be partially explained by referring back to Whitaker's (1970) theory that the close-up shows attitudes better than the longer shot, thus allowing the viewer to better observe and vicariously experience por- trayed feelings. Identification theory and research suggest that, depending upon how a witness looks and behaves in various camera shots, a jury might identify to different degrees. Any identification with a witness may have two effects: (1) the juror may feel psychologically closer to the witness and react with more positive evaluations and therefore find the testimony more credible; or (2) the juror may feel as though he/she has a better understanding of the witness, and therefore feel more confident in any evaluations, either positive or negative. Both these considerations lend 19 importance to determining what effect image size has upon juror identification with a videotaped witness. Interest. Judges and attorneys often complain of difficulty in maintaining juror interest throughout lengthy and frequently tedious trial proceedings. Videotape may either alleviate or aggravate this serious hindrance. Williams (1968) varied film shots to assess the effect on viewer interest level. Showing a filmed lecture to students, he found, contrary to expectations, that a close-up did not significantly raise the interest level of a film, although a long shot did decrease it. When shots were congruent with judged interest levels of sentences (i.e. close-up for high interest, medium shot for moderate interest, and longshot for low interest), only the moderate sentences increased in interest. When image size conflicted with inherent sentence interest, measured interest decreased. Although these results are far from conclusive, they suggest that the shots used in a videotaped deposition are likely to affect the interest level of the jurors. Summary. The empirical research reviewed in this section suggests that image size does have a significant influence on such communication variables as perceived source credibility and interpersonal attraction. However, it appears that certain characteristics of the source interact with, or mediate the effect of image size. Some mediating variables specifically investigated were: speaker body type, sex, and acting style. Physical attractiveness was discussed 20 as possibly interacting with image size. Research indi- cates that image size may also have an impact on identifi- cation and interest, which may in turn influence jurors decisions. One of the common problems of image size research is that stimulus speakers are not randomly selected from a population of speakers, but are instead chosen by the researchers. Also, most of the studies do not directly per- tain to jurors viewing a deposition. However, the research cited does provide an empirical starting point and general procedures for implementing further research. 21 Summary This chapter has pointed out the need to assess the effects of the current practice of videotaping depositions on the guarantee of a fair and speedy trial. Justification was given for examining necessary, but potentially biasing, production decisions such as the choice of camera shot. If jurors are differentially influenced by various camera shots, then production guidelines may be necessary. The literature of media practitioners found intui- tive and experiential speculation that a scene or character may be perceived differently depending upon the camera shot used. For instance, it was suggested that closer shots tend to increase audience involvement and may be more effective for presentation of emotions. The longer shots may carry more factual information in a more detached atmosphere. Most practitioners merely give advantages and disadvantages of each shot without providing empirical support or reasons. Research by social scientists also indicates differ- ences in Viewers' perceptions when image size is varied. Image size was shown to affect source credibility, inter- personal attraction, and interest, but researchers concluded that certain source variables may have had a mediating effect. Investigation in this area is still exploratory and 22 many potentially influential characteristics of the speaker have not been controlled, and many relevant dependent variables have not been measured. More extensive research is needed before any empirical generalizations can be made. Source credibility, interpersonal attraction and interest are three variables that may play an important role in judgments of witness veracity made by jurors. If these variables are influenced by varying camera shot, then jurors' decisions may reflect the bias. Two other variables that may be related to jurors' decisions, and may be affected by shot, are retained information (the amount of testimony remembered by a juror) and identification with the witness. The goal of this study was to determine if varying the camera shot used to videotape a deposition would have an impact on these decision related variables. Rather than just compare camera shots of some set of testifying witnesses, control variables were employed to help specify particular reasons for any effects observed. Based upon information found in the review of literature, the physical attractiveness of the witness was believed to be a potentially mediating variable and of heuristic value. This study examined the possibility that different camera shots might influence jurors' perceptions depending upon the attractiveness of the witness. The literature review suggested witness sex as a second control variable. Since sex has been shown as medi- ating effects of image size (McCain and Divers, 1973), and 23 physical attractiveness might vary with witness sex, it was decided to add explanatory power to the study by also con- trolling the sex of the witness. Research Questions Due to the speculative nature of statements by practitioners, and the paucity of empirical research, it would be difficult to make predictions of specific potential effects. Rather than stating particular hypotheses of image size's influence on decision making variables, this thesis took a question-centered approach, asking whether or not image size makes a significant difference. Specifically, the following questions will be examined: 1. Do subjects exposed to differences in image size demonstrate differences in retention of deposition related information? 2. Do subjects exposed to differences in image size demonstrate differences in identifica- tion with a videotaped witness? 3. Do subjects exposed to differences in image size demonstrate differences in perceived source credibility of a videotaped witness? 4. Do subjects exposed to differences in image size demonstrate differences in perceived interpersonal attraction of a videotaped witness? 5. Do subjects exposed to differences in image size demonstrate differences in the degree of interest in a videotaped deposition? CHAPTER II METHODS AND PROCEDURES In the previous chapter justification was given for further research on videotape in a legal environment with specific reference to the types of camera shots used when videotaping a deposition. writings of practitioners and social scientists have provided no general conclusions, but do suggest possible effects which led to the five research questions posed at the end of Chapter I. The purpose of this chapter is to describe the methods and procedures used to obtain the answers to the ques- tions, and to further clarify the variables used in the research. The chapter opens with a presentation of the ex- perimental design followed by detailed descriptions of the creation of the stimulus videotape and the selection of the sample. The definitions section provides conceptual and operational definitions for the following dependent vari- ables: (l) retained information; (2) identification; (3) mediated source credibility; (4) perceived homophily; (5) mediated interpersonal attraction; and (6) interest in the proceedings. A discussion of the statistical design used concludes the chapter. 24 25 Procedures Experimental desigp. The following three independ- ent variables were manipulated in this experiment: image size (i.e., types of shots used); the physical attractive- ness of the testifying witness; and the sex of the witness. Figure 1 presents a geometric display of the result- ant 4 X 2 X 2 factorial design. The four levels of image size were close-up shot, medium shot, long shot, and extreme long shot (a1 to a4 in Figure 1). Physical attractiveness ‘ was divided into two levels - attractive and unattractive (b1 and b2 in Figure 1). Two levels of witness sex, male and female, are shown in Figure l as c1 and c2. This is a cross- classified design, meaning that each level of each variable enters into combination with each level of every other vari- able, thus forming a total of 16 unique experimental condi- tions. The execution of this design will be discussed next. ./’///i 1////’ _ 44/ Aht‘éi‘éfiivebl / ...‘ . a Unattractive ’/,/”/;2 Ng%ness Witness b F'male 2 c1 Witness 8'1 a2 a3 34 . Close-up Medium. Long Extreme Long Shot Shot Sho Shot Figure l. Geometric Display of a 4 X 2 X 2 Factorial Design. 26 Selecting the Stimulus. To enhance the external validity of findings, it was decided to select an actual deposition transcript rather than attempting to script a plausible simulation. The following criteria were used in selecting the transcript: 1. It had to be part of a civil case. 2. The content of the testimony had to be of the nature that either a male or female witness could present it. 3. The deposition could not exceed one hour in length. 4. The deposed witness had to be a person whose credibility was not already established due to their occupation or status (e.g. a physician or police officer). The rationale for the above criteria is as follows. The first stipulation contributes to the generalizability of the study since the majority of videotaped depositions are currently generated by civil suits. The second criterion was necessary to manipulate the sex of the witness as an independent variable required by the research design. The third requirement emanated from pragmatic consider- ations. To use available adult subjects, the experiment-- including introduction, presentation of the stimulus, and questionnaire completion--had to be less than two hours. Moreover, the reels of tape used to record a deposition are one hour in length. Attempting to work with a longer deposition would thus become quite unwieldy and expensive. Although depositions vary widely in duration, several advisers agreed that one hour was a fair representation. 27 The fourth criterion was necessary to insure an un- biased evaluation of credibility by the subjects based only on behavior observed during the deposition. It is possible that a witness introduced as a medical expert would be view- ed by subjects as having a certain level of credibility that would not fluctuate appreciably due to experimental treat- ments. After a substantial search, with the assistance of legal advisers, a deposition was selected, that met all these criteria. Like many other civil case depositions, testimony was taken from an individual who had witnessed a serious motor vehicle collision. The witness was called upon to present evidence involving precise judgments of time and speed of the vehicles, which could not be corroborated with any other evi- dence. Jurors had to evaluate the veracity of the testimony based upon perceptions of the witness during the deposition. The questioning by the defense and plaintiff's attorneys lasted forty-five minutes. There were no opening or closing statements. Selectinggthe aetors. Eight actors and eight actesses from the Lansing, Michigan, vicinity responded to audition announcements distributed to local acting groups. To pretest for physical attractiveness, each were videotaped for sixty seconds using a medium.shot. This tape was then shown to a sample of volunteer undergraduate student subjects. The videotape was st0pped after each person appeared and the sub- jects were instructed to rate the physical attractiveness of 28 the individual just viewed on the screen. Perceived physical attractiveness was operationalized as scores on the following eleven-point semantic-differential scale: attractive __:__:__:_:___:__:_:___:___:__: un- attractive. Based upon these ratings, the most attractive male and female, and the most unattractive male and female were chosen to play the part of the witness in the enactment of the deposition transcripts. All looked to be in their early twenties, cerresponding with age of the actual witness. Rather than using actors to play the parts of the attorneys, these parts were typecast by employing law students, with previous trial experience, who could add to the realism of the simulation. They were both in their early thirties and had had extensive courtroom.exposure. Taping the stimulus: After all the actors had memor- ized and rehearsed their parts, the deposition was reenacted and videotaped. The production took place in Michigan State University's Telecommunication Department's television studio, using Panasonic color cameras and 1/2 inch Sony color videotape recorders. A light blue cyclorama (curtain) provided the back- drop, and the only props used were a six foot wood grain table and three chairs placed on a large shag carpet. The two attorneys were seated at one end of the table with the witness positioned at the other end so the attorneys would not be in the same shot as the witness. (See studio diagram in Appendix C) Only four cameras and videotape recorders were avail- able to produce the stimulus tape. After an opening establishing 29 shot showing all three participants, two cameras were used to record two different shots of the witness, a third camera to record both attorneys, and the fourth served as a back-up. Each of the four witnesses presented the testi- mony twice, resulting in the following four shots taken at 'witness eye level: 1) Close-up Shot showing head and shoulders of seated witness 2) Medium Shot showing seated witness from top of head down to hands on table 3) Long Shot -- showing seated witness framed from.head to feet 4) Extreme Long Shot -- panoramic view in which seated witness occupies only a small part of the total view. For each of the deposition reenactments, a separate recording was made of the questioning attorney. In order to relieve the monotony of a single fixed view of only the wit- ness, shots of the attorneys were insert-edited at appropriate times during the questioning using the following criteria: (1) at a new line of questioning, (2) during an objection, or (3) during the handling of exhibits. The shots were inserted at the same location in each treatment tape thereby keeping the attorney interaction constant across conditions.2 Subjects. Due to the unavailability of a courtroom setting and actual jurors, potential jurors were used in this experiment to role-play the part. This procedure has been criticized (Freeman, 1969) because it may lack the realism of the actual situation. Jurors probably have a greater sense 30 of responsibility since their decisions will have more serious consequences than subjects in an experiment. An assumption of this study is that the judgment process and res- ponses of adult subjects will not be significantly different from that of jurors viewing a deposition in a trial context. In an effort to approximate the demographic compo- sition of a typical jury, adults eligible for jury duty were recruited from the Lansing, Michigan area. Members from a variety of service organizations, veterans groups, parent- teacher associations, and adult social groups volunteered.- A total of 215 adults were randomly assigned to the sixteen experimental conditions. When each group of subjects met, they were informed that this experiment was part of a research grant sponsored by the National Science Foundation, and that the researchers were interested in finding out their impressions toward videotaped depositions. It was explained that they were about to view a forty-five minute deposition of an actual wit- ness who was unable to appear in court. The subjects were asked to role-play the part of jurors and pay careful atten- tion to the videotape because they would be questioned about it when it was over. In each condition, the stimulus videotape was viewed on two l9-inch Sony Trinitron color television monitors by not more than fifteen subjects at a time, to permit a comfort- able viewing distance without crowding. Room.lights were 31 left on to avoid eye fatigue since the experiment was con- ducted on weekday evenings. After viewing the entire deposition without interruption, subjects completed the questionnaire (Appendix A), were fully debriefed, and cautioned not to disclose the exact nature of the study to friends until experimentation was completed. All subjects were asked if they had recognized any of the actors or had guessed the manipulations, with none responding affirmativ- ely. For each participating subject, the organization with which s/he was associated received a $5.00 donation. Definitions This section will provide the conceptual and opera- tional definitions for each of the dependent variables generated by the research questions. When appropriate, factor analysis of some multidimensional variables is pre- sented. The factor structure will be discussed as part of the operational definition. Reliability coefficients will be provided for each dependent variable scale. Retained information. This variable was defined in this study as the amount of the witness's testimony which is remembered by a subject at the conclusion of the deposi- tion. The construct was operationalized as responses to twenty-seven multiple choice items (one correct choice with three foils), and twelve completion items requiring one or two word answers. The questions pertained to relevant 32 material evenly distributed throughout the stimulus testi- mony. One question, regarding the birthdate of the wit- ness was misprinted on the questionnaire and was eliminated from.the analyses. Reliability computations for the thirty-eight item retained information scale produced a Cronbach's alpha coefficient of .82. (Unless otherwise in- dicated, all calculations are done by computer using the Statistical Package for the Social Sciences, version 6.0). Identification. This was defined as the degree to. which a subject is vicariously involved in the proceedings through a psychological relationship with the witness. The identifying subject would experience an emotional state simi- lar to that which s/he believes the witness is experiencing. To operationalize this construct, a seven item, comparative judgment, ratio scale was developed. According to Torgerson's (1958) "Law of Comparative Judgment," if two stimuli are presented, an observer can "judge which is higher on the psychological continuum.(e.g. which is louder, heavier, or more beautiful)" (p. 160). The observer can al- so make judgments as to the amount that one stimulus is higher than the other, beginning at some arbitrary zero point and using some numerical rating to express the degree of difference. In this study, the viewer-subjects were asked to con- sider the amount of identification they felt they would experience when viewing a live deposition. This amount of identification was arbitrarily valued at a constant rating 33 of 50. For each of the seven identification items, the subjects compared the identification they felt during the videotaped deposition to their impressions of a live pre- sentation, and responded with an appropriate rating of either more or less than, or equal to the standard of 50. For example: How much did you feel the same emotions the witness felt? If the viewer believed s/he felt the same emotions to a greater degree during the deposition than they would during a live testimony, s/he chose some number larger than 50 (e.g. 100 to represent twice as much identification). If the viewer felt s/he would experience less emotional in- volvement watching the videotape than if the witness were viewed in person, the subject reaponded with some number smaller than 50 to represent the difference in intensities. If the response was "50", it meant that the subject believed s/he would identify with the witness to the same degree, live or videotaped. (See Appendix A for questionnaire) After data collection, examination of the ratings for each identification item demonstrated that subjects had developed a response set. Positive correlations among all items indicated that the subjects responded to every ques- tion as measuring positive identification with the witness. However, item.#3 (see Table 1) was intended to measure detachment from the witness (i.e. negative identification), and should have been negatively correlated with the other responses. Further support of this conclusion was supplied 34 when the scale items for identification were submitted to principal components factor analysis with varimax (ortho- gonal) rotation. This resulted in a single factor with all items loading +.60 or better, explaining 57.8% of the variance in the set of measures (see Table 1). Item #3 should have had a negative loading reflecting its negative association with the factor. Due to the suspected invalid- ity of the item it was dropped from the scale and not used in subsequent analyses. Reliability computations for the six item.identification scale produced an alpha coefficient of .86. Table 1 Rotated Factor Loadings for Identification Items Item Loading 1. How much did you identify with the witness? .78 2. How much did you feel the same emotions that the witness felt? .78 3. How detached from the witness did you feel?8 .72 4. How well did you feel you knew the witness? .85 5. How close did you feel toward the witness? .81 6. How restless did you feel while view- ing the witness? .64 7. How much did you feel you were a part of the action of the case being tried? .65 Cumulative Variance (after rotation) .58 8Item dropped from subsequent analyses. 35 By measuring identification with a videotaped wit- ness, relative to feelings toward live testimony, we can gain some insight into juror's preferences. 'With the same scale we can also compare varying degrees of identifica- tion across treatment conditions to assess the effects of the independent variables. Mediated source credibility. This variable was con- ceptualized as a viewer-subject's evaluation of a videotaped" witness as a communicator of case related information. Credibility is not determined by any inherent qualities of a source, but rather by characteristics receivers perceive in a source based upon five hypothesized dimensions: (1) socia- bility, (2) extroversion, (3) competence, (4) composure, and (5) character. The items used for the operational definition similar to Wakshlag's (1973) study of image size, are a combination of scales used by McCroskey, Hamilton, and Wiener (1974) and ‘by McCroskey, Jensen, and Todd (1972). The resultant mediated source credibility scale consisted of 17 nine-point semantic differential items. Each hypothesized dimension was measured by three items, except for competence and character (most relevant for evaluating a witness) which were measured by four items each. An example of the scales follows: Good ___:_:__:_:__:_:_:___:_ Bad The bipolar adjectives composing the sociability dimension were: good natured-irritable; cheerful-gloomy; 36 sociable-unsociable. The bipolar adjectives used for the extroversion dimension were: aggressive-meek; verbal-quiet; and interesting-boring. The bipolar adjectives making up the competence dimension were: competent—incompetent; expert-inexpert; intelligent-unintelligent; and qualified- unqualified. The bipolar adjectives used for the composure dimension were: poised-nervous; relaxed-tense; and calms anxious. Bipolar adjectives composing the character dimen- sion were: honest-dishonest; believable-unbelievable; reliable-unreliable; and good-bad. To avoid response set, eight of the bipolar adjectives were reversed on the ques- tionnaire (see Appendix A). Perceived homophily. ‘McCroskey, et a1. (1974, page 43) contend that homophily is a construct closely related to credibility and is of heuristic value for predicting communi- cation interaction. Rogers and Shoemaker (1971) define homo- phily as "the degree to which pairs of individuals who in- teract are similar in certain attributes, such as beliefs, values, education, social status, and the like" (p. 14). They maintain that individuals who are homophilous will come ‘municate more effectively with each other than those who are not. This principle, combined with research on characteris- tics of "opinion leaders" (Katz and Lazarsfield, 1955) suggests the concept of "optflmal heterophily." "The most influential source is highly homophilous with the receiver in all other aspects, but is perceived as somewhat more com- petent on the topic in question" (McCroskey, et al., 1974, 37 p. 43). In their study of interaction behavior, McCroskey, et a1. (1974) measured homophily in conjunction with source credibility, using three semantic-differential items to guage attitudinal similarity. Since a deposed witness deals mainly with observations of events, a juror would tend to judge homophily based on how the witness looks and be- haves. For this study, perceived homophily was defined as the degree to which a subject perceives the observed be- havior of a trial participant as being similar to his or her own behavior. This variable was operationalized as ratings on three nine-point semantic-differential scales similar to the credibility scales above. The bipolar descriptors used were: speaks like me-doesn't speak like me; like me-unlike me; and acts like me-doesn't act like me. Based upon measurement of homophily in past research, these items were inserted among the 17 credibility items on the questionnaire. This resulted in a 20 item Credibility-Homophily Scale resembling the one used by McCroskey, et a1. (1974). To verify the existence of the six dimensions hypo- thesized for mediated source credibility-perceived homophily, the scales were submitted to principal components factor analysis with varimax rotation. The criteria employed for interpretation of the factors were as follows: 1. Each orthogonal factor will be interpreted as a unique dimension. 2. For an item to be judged as measuring a parti- cular dimension it must have a primary loading on that 38 factor of .55 or higher with no secondary loading of .35 or higher. 3. In order to use a dimension in later analyses it must contain at least two items that meet criterion #2. The results of the factor analysis are reported in Table 2. Seven items did not meet the criteria for inter- pretation. All other items loaded as predicted on factors designated by sociability, competence, composure, character, and homophily. The three items associated with the extro- version dimension did not load well. It was decided to drop the dimension from.subsequent analyses. Four other items did not load properly and were not used in later computa- tions. They were: expert-inexpert and qualified-unqualified (competence dimension); sociable-unsociable (sociability); good-bad (character dimension). The five factors used in analyses accounted for 59.4% of the variance. Reliabilities were computed independently for the five sets of items. The alpha coefficients were adequately high to conclude that the items making up each dimension were internally consistent (see Table 2). Each orthogonal dimension was treated as an independent dimension in later calculations. Although the behavior of the two questioning attor- neys was held constant for each condition, a check was made for any confounding effect due to their presence during the deposition. Subjects were asked to rate both attorneys 39 Table 2 Rotated Factor Loadings and Reliability for Mediated Source Credibility-Homophily Items (Principal Components Factor Analysis; Varimax Rotation) Scales Canpo- Char— Socia- HQID- Carpe- Exorc- sure acter bility phily tense version PoisedrNervous .87* .06 .03 .02 .18 -.01 Tense-Relaxed. -.75* -.03 -.03 .Ol -.09 -.28 Calm-Amcious .70* .13 .10 .08 .10 .05 Ikbelhambhe- Believable -.02 -.77* .00 .08 -.13 .02 Dishonest-honest -.O6 -.69*’ .02 -.03 -.17 -.OO Reliable-Unreliable -.00 .59* .31 -.02 .13 -.06 GbodsBad. .15 .44? .38 .08 .28 -.17 Goxinamned- Irritable .07 .15 .82*’ -.03 .09 -.07 Cheerful-Cloany . 18 .06 .77* - .01 .10 .10 Uhsociable-Sociable -.06 -.42 -.36 -.00 -.10 -.39 timeksiumeimr- .10 -.L1 -.Q3 .68* .12 .04 Duanfltzaxmkzlfluame Acusliketmr Doesn't act like me .04 -.02 .03 .61* -.06 -.07 Unlike me-like me .02 -.ll .00 -.55* -.20 -.21 Dmxnpeuamh competent -.18 -.25 -.16 -.Ol -.74* -.16 Uhuuxilhyamk immerugent -.L3 -u15 -.G2 «~12 -.WW* 'nl3 Qmflifflak» Uhqualified .19 .24 .17 .15 .46 -.32 Expert-Inexpert .24 .23 .19 .10 .35 -.24 Boring-Interesting -.12 -.09 -.15 -.20 -.36 -.55 Feekwhggresfina -u22 .13 .08 -.04 .04 -.51 Verbal-Quiet .28 -.05 .17 .06 .07 .02 Infinmsllhfliaaflury of Sets of Adequate Loading Items .84 .71 .72 .64 .71 *Factor loadings and corresponding scales which met criteria. 40 using the mediated source credibility-perceived homophily scales used in rating the witness. Items in each set of measures were combined to form one 18 item scale to assess the credibility-homophily of both attorneys. This scale was factor analyzed using principal components with varimax rotation resulting in four usable factors (see Table 3). Two of the previously hypothesized dimensions, sociability and homophily, each formed a distinct factor. However, items composing extroversion, competence, and charac- ter were divided among two orthogonal factors. Comparison of the two factors with the questionnaire (Appendix A) revealed a pattern existing in the reaponses. Those items loading high on Factor I had been expressed on the questionnaire as adjectives with a negative connotation on the right hand side, or "9" end of the nine-point bipolar scale. Those items comprising Factor II were expressed as positive adjec- tives on the right side of the scale. This suggests the possible occurrence of a "left hand effect." Respondents seem more prone to rate a stimulus closer to the right, or upper end of the scale rather than the left, or lower end when either is positive. It is not known why this phenomenon did not occur in measures of the witness. Reliabilities for Factor I and Factor II items were .91 and .86 respectively. These two orthogonal factors were treated as two separate indices in later computations. Composure itams did not meet the criteria for accept- able factor loadings and were dropped from subsequent analyses (see Table 3). 41 TABLE 3 Rotated Factor Loadings and Reliabilities for Combined Attorney Credibility-Homophily Items Seuss lkmior Pecan: Hmnr- quar Pecan: ~ #1 #2 lflfily Ifility #5 QmflifHXP Ummuflifflai .89* .J9 .01 .06 .13 Expert-Inexpert .79* .23 .01 .08 .16 Reliable—Unreliable .79* .27 -.03 .27 - 03 Ififlsedikrwnm. .73* .19 -I£i .20 .40 verbal-Quiet .69* .18 -.11 .04 -.05 GbodrBad .67*' .32 -.Ol .30 .06 'UunmeUigan> Intelligent -.29 -.76* -.02 .06 -.26 Unbelievable- Believable -.27 -.75* .02 -.21 -.04 Dishonest-Honest -.34» -.71* -.03 -.28 .03 Imxnpeuzms Competent -.30 -.70* .06 .07 -.18 Boringrlnteresting -.Ol -.58* -.14 -.19 -.14 lkmkdagraxuve -u25 -n46 In. .12 .07 Speaks like me- Doesn't speak like me -.01 .03 .90* .ll -.06 Actsliketmr Doesn't act like me -.12 .02 .87* .05 -.03 Unlike me-Like me -.01 -.04 -.71* -.04 -.04 Cheerful-Gloomy' .28 .06 .10 .80* .14 Gbod.naturedrlrritable .33 .01 .10 .71* .19 Uhsociable-Sociable .08 -.46 -.09 -.52 -.16 TensesRelaxed -.02 -.39 -.05 .23 -.72 Calm-Andous .39 .03 .01 -.25 .56 Innamallmdiafiflity offkmsanAdepmme loading Items .92 .86 .91 .84 * Pecanrloafingsamd pow aslpom ownvsvm mo EmmmmHo . Ho omswwm o Amhosaoapdv Wsmwo ' .am ma mmonpfls gosm maofiwo 51” among? .Ho «93. . 6:“ NN whaaopfl. .am mm Jaw am.“ mozm ‘ :Iomoao . unwafioflpamml o ‘ 4 ponm «.350 .. ‘ 5:.ch wmmm ‘ .pm H u.:a « poam msoq «hon oaohfinm mpcwmfiowphwm op mahoaao 89C. moosmpma BIBLIOGRAPHY BIBLIOGRAPHY Bermant, G., McGuire, M. and Chappell, D. Videotaped Trial Presentation: Advanta es and Disadvanta es, BatteIIe MEESrIaI InstItute, 1975 ZStencIIS. Berscheid, E. and walster, E.H. Interpersonal Attraction. Reading, Mass.: Addison-Wesley, I969. Dion, K.K., Berscheid, E. and Walster, E. What is beauti- ful is good. JOurnal‘of‘Personality and Social Psychology, 1972, £5, 2854290. Doret, D. Trial by Videotape - can justice be seen to be done?, Temple Law Quarterly, 1974,‘11, 228-268. Eisenstein, S. Film.FOrm and the Film Sense, translated by Leyda, CIeveIand, UHIo: WorId PuEIIshing Co., 1965. . Freedman, J.L. Role-playing: Psychology in consensus. Journal of Personality and Social Psychology, 1969, I3, 107-114. Friend, R. and Vinson, M. Leaning over backwards: Jurors responses to defendants' attractiveness. Journal of Communication, 1974,‘24:3, 124-129. Hays, W. and Winkler, R. Statistics: Probabilit Inference, ‘and‘Decision (Vol. 2), New York: HoIt, RInEHart and Winston, Inc., 1970. . Huss, R. and Silverstein, N. The Film.Experience, New York: Delta, 1968. Kagan, J. The concept of identification. Psychological ReviewI 1958, 62, 296-305. Kaminski, E. The effect of differing editing techniques on juror responses, Unpublished Masters thesis, Michigan State University, 1977. Katz, E. and Larsfield, P.F. Personal Influence, Glencoe, 111.: Free Press, 1955. Kim, J. and Kohout, F. Analysis of variance and covariance: Subprograms ANOVA and ONEWAY. In N. Nie, C. Hull, J. Jenkins, K. Steinbrenner, and D. Brent, Statistical 'Package for the Secial Sciences (2nd ed.). New York: McGraw-Hill, 1975. 110 111 Kracauer, S. From Caligari to Hitler (2nd ed.), New York: Noonday Press, 196D. Landy, D., and Aronson, E. The influence of the charac- ter of the criminal and his victim on decisions of simulated jurors. Journal of Experimental Social Psychology, 1969, §,_1£1-152. Lewis, C. The TV Director/Interpreter. New York: HastingsHouse,1970. Lindquist, E. Design and Analysis of Efiperiments in Psychology and Education. Boston: oug ton- lin, 1 53. Maccoby, E. and Wilson, W. Identification and observation- al learning from.fi1ms. gggxngl of Abnormal and Social Psychology, 1957, __, 76-87. MadsigéaR. The Impact of Film. New York: Mammillan, McCain, T. and Divers, L., et al. The effect of body type and camera shot on interpersonal attraction and source credibility. Paper presented at Speech Com- munication Association Annual Meeting, New York, 1973. McCain, T. and Repensky, G. The effects of camera shot on interpersonal attraction of comedy performers. Paper presented at the Speech Communication Associ- ation Annual Meeting, Chicago, 1972. McCroskey, J. Scales for the measurement of peer and 'media credibility. Unpublished manuscript, Illinois State University, Communication Research Center, 1971. McCroskey, J., Hamilton, P. and weiner, A. The effect of interaction behavior on source credibility, homo- phily, and interpersonal attraction.‘ Human Cemmunica- tion Research, 1974, 1:1, 42-52. McCroskey, J., Jensen, T., and Todd, C. The generaliz- ability of source credibility scales for public figures. Paper presented at the Speech Communication Association Convention, Chicago, 1972. _ Millerson, G.‘ The Technique of Television Production (2nd ed.), New YorE}HastingsHouse, 1972. ‘Millerson, G. Effective TV Production. New York: Hastings House, 1976. 112 Mills, J. and Aronson, E. Opinion change as a function of the communicator's attractiveness and desire to influence. JOurnal'of Personality and SOCial Psychology, 1965, I: 173-177. Montegu, I. Film.Wbr1d: A Guide to Cinema. Baltimore: Penguin Books, Nie, N., Hull, C. Jenkins, J., Steinbrenner, K. Brent, D. Statistical Packa e for the Social Sciences (2nd ed.), ew or c raw- Rogers, E. and Shoemaker, F. Communication of Innovations. New York: Free Press, 1971. Ruch, F. and Zimbardo, P. Ps cholo and Life (8th ed.), Glenview, 111.: Scotts, Foresman an o., 971. Siga11,H. and Aronson, E. Liking of an evaluator as a function of her physical attractiveness and the nature of the evaluations. Journal of Experbmental Social Psychology, 1969,‘§, 299- 306. Sigall, H. and Landy, D. Radiating beauty: The effects of having a physically attractive partner on person perception. Journal of Personality and Socia PsychologyL I973, gg, 218- 224. Tannenbaum, P. and Fosdick, J. The effects of lighting angle on the judgment of photographed subjects. Audio Visual Communication Review, 1960, 8, 253- 262. Tannenbaum, P. and Jaer, E. Mood chan e as a function of stress of protagonist and degree 0 identification in a filmrviewing situation. Journal of Personality and 'Social'PsychOIOgy, 1965, 2:4, 612-616. Torgerson, W. Theory and Methods of Scaling. New York: John Wiley & Sons, 1958. wall, W. and Simonson, W. The responses of adolescent groups to certain films. 'British JOurnal of Educational'SOciology, 1951,621, 81-88. Wakshlag, J. The effects of camera and hmage size on mediated source credibility and interpersonal attrac- tion. Unpublished Masters thesis, Illinois State University, 1973. . Weiss, W. Effects of the mass media of communication. nThe Handbook of Social Psychology, J. Lindzey and E. Aronson (2nd ed. ,) Vol. 3, Raising, Mass. : Addison- Wesley, 1969, 77-195. 113 Whitaker, R. The Lang%age of Film. Englewood Cliffs, N.J.: Prent ce- a , . Williams, R. The value of varying film.shot on interest level. Speech Mbnographs, 1968, gg, 166-169. Wurtzel, A. and Dominick, J. Evaluation of television drama: Interaction of acting and shot selection. Journal of Broadcasting, 1971-1972, 16:1, 103-110. Zajonc, R. Some effects of the "space" serials. Public Opinion Quarterly, 1954,‘1§, 367-374. Zettl. H. Television Production Handbook (2nd ed.), Belmont, a.: a swort , .