.. 'l I . I .3 . ‘31.“! ‘1.I.. “1“" W‘ EI““W I 111.1 1.1513“‘ ““1“1““‘"“ I-"‘I‘I“'I 3515"“..1-1111‘5‘..5151..‘.‘I“ W 1““ ":5....5...1.5.1.5.I.III.I..I1I5I ‘1“‘11“““‘“‘!““““ 55‘5‘5‘51“ 15911111 .5553 m... . 5.1.II11‘1‘“.15I.““‘II‘ I“. » “I". "5555‘ 1.155555" I:I.'1.5. “5‘“ ""“.I'I. 5‘151.““I.I‘1‘.;;“51.. “(31.5 53. “ ““ . ‘I‘ 1. W ‘I ““‘151I5‘I1‘I1‘I‘I 35".11.11..I 151N111 ....I‘.‘.;.g.;1;11.115555.1I1{‘I... “‘31 1331.11 I311“). 3315...... ~ " ... ‘ ““““‘ 3“ “‘3‘“ ‘3"‘ I": “3““ ‘I‘"“I‘5“‘I"““““““I“I“ 1“"111‘1 1‘1“ 333‘ 1‘51“ 1‘1115 {“1. 13" 5:2: 51111;. A2913: 1. 8.3 ....1‘..1 1.1 .1.\IIII.I..~1 I.I.‘151..5i.1.1.1.2..I1.I.l.l.l....51..<.;.5..\l .‘M.“l5‘ Liz-iii“. _-,§1,A.;.3.R III ..1:-'1:11..1‘1‘11,1I11,.11111.‘5‘1‘1115‘... 1.. «v1. w II.‘.!1.1;.. .111: 11.. .1.‘1111 I1 ‘I'II..I‘1II..I.I.‘I‘ 11 I I5“. ‘1.“- ,.... ~...'.I.“§L}- finnfi... .. . 511.1% 55135515515158.1555. 1.51;1515....‘5.1‘. 5.5.53. ‘I.II.I1,..’1‘5 I.I-1.3171,. 3:113:11“. "a. 1:1? 15‘". IT“I‘I1.51.1I‘1II ““HI“‘1’“‘H‘1““&.:“‘““€‘7‘ . "I "‘ . "I . 1‘55... 5.....5....I...515115151511 5.115.554.5555..." III.I1.....'.551..11I{15......5.I.. .1.‘1‘“ . .51.. 3%‘3‘1‘3‘fifi ;1,...::}..;,1.:.;: H5151 I‘ IV‘.“II‘I‘111 . 1.1I11 11 .1 I. 1 .«:....:;‘_-_ 5... ~11“ .- ‘1 .111... .‘1.“!1:I:;‘;I§I1I‘I1: 15121111553151...I111;‘1;:“11.‘.1I‘1.5.1;1‘5‘1 . ' cafifixk‘ffiifi 1 .. 1.1"»‘I“I‘1‘11IIII5‘1"11"i""HI‘I‘I‘I‘I““1""I331‘ ‘1“I‘ “ ‘13“ {Ex}. _ I11“ 15. Hfifimwwmwmmmw“35513'1m1£$1I$15.531Ififigggfi- "fii 1‘.‘1,.‘.‘. “1“" I.“"15“I‘II5I.“1:I.I""‘I‘ "‘.‘-1‘I“II"‘I5‘:“‘51‘§' ‘ "A ‘ '3 ‘Y‘i“““:‘" ..1.1..I|.... 2 5.‘ .I‘.‘ ! .151.I.‘.I.l1 ”12“.. ..I“..1...1.5..1 1 , ‘5‘ ,_ {3%. S ‘11 “ -“‘ ‘ “"“‘5}‘.1".II“I1‘.:""“‘I‘“II1"' “ “‘I‘ ““ "1.1 ““5“ "”0"“ "‘31“ “ . "“I‘. I“ '1‘1‘H.'“'.1 ' ‘1; 1‘(‘ I‘ ‘Hbll‘fi‘ 1‘15!“ 1‘15"“ "‘3. L‘ R 1.1...15‘.1......I 4.5.5.5. .. .5... .;1.1. {‘55. ,.5........... I..1I ....1....I. 9.155. .1... . . .‘ “I555 5.55553151‘1‘I".1‘5“‘..5555..I I1II ‘5‘55‘5‘5 3:15": ““33““ l.‘; I“3351.‘5I“5“.:I5I1‘I‘15551I.‘5‘..5.5‘5:‘ ."“ .1.1I1I...11'1‘11'.. “I .‘.11II51’,‘1.1. 1'II15.11.11 ‘-.' .1». I.I"I‘11 I. 1111 .11.111.111.111111.11 .1. 111. 11111 IIIIIIIIIIIIII. 1.151. ‘1‘ III 11 1 I“I.I .1513 .I‘.II, I. ‘ I .. ‘I“‘ I“ “ ““‘I‘ “‘ “I‘ "““1 ‘1‘ 31“" "“"‘ "““‘I‘ “.‘5“‘.‘I“ 3““ “'5 ““3" “““““‘"‘o‘(' ““‘I‘1 " "“ {I'I‘ I-‘I'1I '11'1 11'1.I‘I.III ‘ ‘11‘ II ““1“ I III‘» I' I .1111 11‘wwwwywflII‘1N8138‘5‘55555. .1.1.1.1.I.I.51.I...I..‘. I;.1'I. _ .1'1I51:I.I.5.1‘1.11‘ .1. ‘11 .15!. 111...!“ 11.11.. . 1.1151 "1 ..‘1i...I.1‘I.I. .1. 11.1 5.11.I.I 1I.‘5I 1..1.1 1111. ...51I1‘11(1 1.1.1.1 II‘.‘I.I“1‘ ‘I51.111‘“I‘II 15.55‘65 .“I.1.1‘ 55.“: ! ' ‘i‘III‘1.1‘1‘1 1‘1, 1.“:- Id..11.1.I.I11“ I.III ;5..I.I‘.I.‘!. .I.I..l.‘.111.1. . ‘III5.15I.I.11‘1‘!.‘.!.51‘.15I.I.l.‘.I. ‘I.I ‘1‘-!.".w:1 I-I .. ...1.1‘;.I."5 ‘1“‘I U355 551...... “.5... ‘5‘1‘1352555‘51'3 “51.5131515155‘ “ 51.1.5813”;11151.1..55‘I‘15515Ig“‘3“1‘.“‘...1.1 ' v “ii t‘vE “S 51‘1‘.151I11‘I""1'2I,I.I‘I“ .I‘.‘.II1I‘11 1.1.181 .II‘II‘1II 1I111 ‘1”..5‘15' J “I‘ “‘5! 1“‘ "‘I‘.I'1.“5I5‘r "“5“‘ .‘5‘."" “MI“ ““I“1“11‘..I..5‘ “““ ‘.“.‘.‘551‘5 "51.15‘.I.\1.551 ‘H.f“! ‘i“““.“"1.‘.i“““‘ {“51 £55; :2. . IGI- MA I I‘ 1 .5... ‘1‘.‘1. “‘ “ ..I‘.. I‘ ' 1113.1 "““““““ 5‘ 1.1. ““‘5.. ““““.5‘ 1351.1“‘“‘1.“““‘I“‘15 "‘1 ‘1 "(31“‘155‘1‘151‘15‘33‘3‘533‘ .‘31‘1‘115511I1‘I‘I..1 .5 .1... 1.1.1.1 8.1.1.11. .I.1‘I‘. I;111.1;I.I.1}1‘11..I‘I1.1. 1.18.1.1“; "““II 1‘1 ‘1“1““1 ”.II‘51I1.1‘ .15 .5153.‘ I.1..I..I “5.55551 15.5.15 " .I.I.I.-'I1II 111331 13 ‘1'1.I., II.IV‘1.“‘ I‘“ “ ‘5‘1I1‘H 15“ I .. ‘111..I..‘11§ I .1 . 1.1. II .111 .1 5.51515 .1...15 51.51111. .55. 551. 8......5. 1.. .....I1.15151 5...... ...11“I1I1...I.5;15..“1....;.I.§.. 15.1.. ‘15.... 1 ‘11... I '1111111‘I1 ..11 1.11 ‘1‘“1 I I.1 “M“ I“ 55I‘I‘I“‘I““5I‘I“‘I“ 1.I“.““““‘5“ 55“" 31“"‘5‘5‘ 35‘3““ “““‘5$ “““‘“ “‘51.. ..H. 111‘ .. 1 I‘1.".‘1‘1.‘I 1..“‘51‘1.":“5“I" “11‘ 51‘1“ ‘15 ‘1 I5““" I“.‘1I ".‘5 I5II 51.1.1 1.11 “1 “‘“ ““l“"‘ ‘.‘.‘5.‘I ““5““ .1... .1...(II. 5355 1.1511311 “"5I“I I .I5I.I. I 1. ”1.5111. I5 .‘11 1.1; ““1! 1 ‘I“.H‘5... .1;.1 55“ ‘55'5‘55 :‘55‘15H‘I‘HIH‘ 5 .1" 5511555‘. I...“ 1 "'I .1 . :5..1.....“I.511.1151.1.5555.“.1.‘...I“.5.11‘1I15.1..1....3515... .1.555‘.‘15“.5555555I5. . 51.1.1155: .1111 .II1....1‘.1.“.159‘..1..‘.‘.I..5.. ...5. 5.1.5551. “.15. (1.1“. I5..5.5 1.5.1 I‘I.“5‘ 11..I...551‘5 5.‘ .“5..5I {51.55“ 55‘ ..5IH ‘“‘| ”MN“ 511‘: .5‘. 1“5 ‘5‘“ 1 I. 1.1 ‘I .1 “1‘1 "511‘ ‘1‘ 1““ ‘1II.1.I“5 ..1.‘..5‘5‘“.I55“II1I5I "‘1‘1‘355‘.““‘I‘ “““ 5111. .55‘8 ““‘1I“I“1‘II““51. ““5II““ 5 II‘ “““““‘1“55I“““5“‘5“' 1““ 5‘1 5 1.1.15 ‘1" "“31“ II""II!. ““ ““‘I 13““ ““5“ 5‘; ".51‘3““ .I““ ‘.....5 .I‘I.l. . 11‘; "55‘“. I 5 I I 1..1 I I ‘ I I -' “33m 885WWW‘MMIWMMWfiWWH‘Ww‘WWWW I‘ .1‘1“‘-I.I‘. ‘I“‘l‘!“"““ 5."."“““‘ ‘1‘1 1‘1 3““ “‘ "‘H ““I .1‘1 111 I ‘I 11.I. ‘I. 11 ‘ .. ‘..(5 ‘1I ‘ ““1“ $3.151“ . .11 ‘13 I “1‘1 ‘ II .‘ » l 13 IMHIW‘fiWWMU5‘W‘w‘ .M‘WI‘I‘ .I .... ..3151...‘....155555‘.55.55511.15.. .5.1:i..‘1‘1‘1,.”1.155.5}.‘.555. ' “ I‘II|‘1"1II‘III5 ‘II ‘1.;‘II I‘I; I1.111‘I‘I11‘1‘I‘II ‘I “51‘. .I 11‘». ‘55:“‘5.55I. 155155“ :5“ .5 55‘5‘55 31‘9595‘15‘5‘...5.. .155" 5“‘..' .5... ‘51E {.5555 41““ “' 31“” 51“" “‘ I‘ ‘1 . ‘111‘5155‘5‘ ‘5‘151 1‘1“..5I1I1I.““‘ ‘8 ‘I ‘5‘5 . 31‘1 ‘1“ 15“"“55 ‘55 ‘ 1‘. . . ‘_ H I I; I ~ . 1.. 5...... 1.1.1.51. 1.......1I‘1.1.1.1 ‘1‘5‘1115 15555.1“:.‘15.1.I.;‘5I1.1...1....‘.5. ‘55.. 1- 1.35513 ‘.. I . ‘ I“‘“““ ‘1‘11 I “““‘I"‘ 5“" “I‘-I“‘ 5“"‘1'5‘ .““1“‘1I“‘ t ». ‘1 I 1...”. . 51.111” ‘111‘5‘“‘1‘I11I11IIII‘III“1“II ‘ ‘ II. 1 ‘1 ""II 11‘1It‘1‘1‘1‘133‘1‘1- ““II‘ I"“I““‘I-“‘,““ '3‘93‘3 “ I1 ‘1 ‘I;I‘I“1‘I.“1‘"‘_‘I. .11 1 ‘ "5‘5“" ““I‘I‘III‘I‘1I. ““1511I5I‘II‘ ‘ ‘ ‘ 1.159411 ‘1‘“‘31 51‘..II' I I“ ““I‘ ‘ ‘ ‘1'»‘I‘I‘115I.'.5I.'.1‘ . .. ‘- H55... ‘I1‘5y5..‘| 11551.5 ‘11“'15‘.‘..;.“I“.“‘5 .055555‘5§. 5151‘1. 55.?55 555555515: ”“5555; “551 55.351; . “‘ "““' 515153 ‘3‘ 1.1.; “.'.III1.I51‘I‘I11 5.". 15151355515 5... .55‘8‘3‘ I31 “5...”. ““‘....‘I“:,‘ I5 ‘1‘5‘. . . - 1 ...,.I.5551.I.".1..I1.1.11.1.l ‘..‘ 55511.. ‘ ill‘ .ef 1‘. :3." {13%; "‘3‘“ ““““““‘“‘1“““““"“ “I"IIIIIq‘I‘I ‘IIII‘I‘ . .11. .11... .‘I‘ 1‘5‘5‘3 “'3 ‘55....15515M 11‘15‘55...5“.5.515....‘.I“..1. 5.. I 111 .1 .111 11 I»I 511“"; “I" “III“I‘ h.1‘ 1‘35‘1III ‘1 “‘ “1‘1‘51 I‘.‘ “ “"" “‘3‘" .i .I‘III.‘,‘II.‘ II.I.I! ‘4 ‘1“ $.5555 .;5‘.I.. I.111. 1') 1.55! .5.‘..5. 1.11151 11311.15 ‘! .Mfl §m%w "‘15. ‘1‘5‘ uh .IIII .‘H I‘ 3.1:. .1 5“ 1‘5‘ ““45“ ‘5‘ l"' .[ Idm‘ 5 ‘ 5.1.I5‘. ;- :35: .b £49: “£2. 3 xfi- I; 853.23 .5. 1‘3"): ‘ “:a‘f' *- A: A "5....1.‘ A _' 4:53;? ‘I“I "I II‘I‘I.“ ‘n‘. -.—.x. 0.. 3...: pm'fy‘zwm'r-‘H 3 4—1": 2&5; ”0' 2:. “W":— 5 . 5.54.“; ‘ ‘1‘".1 m.“ “ 5. ““F‘ . ‘5‘"... a“ ‘ “II “ 1. 13 "I‘., I‘m“. ~- u‘k I‘ " “ ‘51“‘1111 m ”55" 5 Michigan State Univcxsity This is to certify that the thesis entitled SOME COIVIMUNICATION STRATEGIES IN THE COURTROOM: A SIMULATED JURY STUDY ‘ presented by Jose Ruben Jara Elias has been accepted towards fulfillment of the requirements for _Ph._D.._ degree in Emmi-0n Major professor Date M £14”? y . I 0-7639 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIAL§5 Place in book return to remove charge from circulation records SOME COMMUNICATION STRATEGIES IN THE COURTROOM: A SIMULATED JURY STUDY BY Jose Ruben Jara Elias A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1978 6//’23c23/ Accepted by the faculty of the Department of Communication, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Doctor of \ \\ - Philos0phy degree. _/ )7, ’ / , /. /- / / / o ,/¢H;;/ Acting ‘j 3.r‘«‘ , Chairman Guidance Committee: ABSTRACT SOME COMMUNICATION STRATEGIES IN THE COURTROOM: A SIMULATED JURY STUDY BY Jose Ruben Jara Elias This work is concerned with the effects of some com- munication strategies used by an alleged offender standing trial on the decisions rendered by a simulated jury regard- ing the offender's degree of responsibility. The theoret- ical underpinnings for the study come from attribution and equity theories. Equity theory provides a conceptual equa- tion for arriving at decisions regarding the fate of a defen— dant based on a variety of perceived rewards and costs experienced by the offender and his victim in the course of their relationship. Attribution theory provides some guid- ance in identifying the relevant inputs and outcomes to be entered in the equity equation. One of such inputs is the juror's liking for the defendant. Such liking can be great- ly influenced by communication variables such as the content of the information regarding the defendant's behavior and by the source of such information. The interaction of these two variables also affects the jurors' stereotyping of the defendant, their feelings of sympathy for and similarity with the defendant, and their perceptions of his honesty. In order to test the relationship between these vari- ables and attribution of responsibility, 120 undergraduate Jose Ruben Jara Elias students enrolled in various social science courses were asked to imagine themselves as jurors, and render a judgment regarding the responsibility of a defendant accused of mur— dering his father while attempting to defend his sister from the father's attacks. In a 2 x 3 design, subjects heard either the defendant (self—disclosure condition) or a witness (other—disclosure condition) disclose actions per- formed by the defendant which were either negative (negative avowal condition), positive (positive avowal condition) or neutral (neutral avowal condition). Results indicate: a) The lack of a significant relationship between the experimental manipulations and attribution of responsibility. b) A significant relationship (canonical r = .68) be— tween liking and attribution of responsibility indicators. c) Significant relationships (canonical r = .60) be- tween liking and three hypothetical dimensions: perceived honesty, perceived similarity, and sympathy. In the discussion, several operational flaws that ex— plain the failure of the experimental manipulations to take effect are pointed out. It is suggested that future research should be aimed at identifying the communication behaviors that serve as antecedent conditions of these psychological processes, and to increase the mathematical rigor of equity theory formulations by casting them in the form of informa— tion integration models. It is also stressed the need to Jose Ruben Jara Elias carry out similar studies within the context of social situ- ations with different outcome structures, in order to study the effects of such structures on human behavior. To my Mother, who has hoped for this To my Wife, who always work more than anyone in the world knew when to push ii ACKNOWLEDGMENT S I would like to express my deepest appreciation to all the people who helped me carry out this work: Professor Edward L. Fink was a continuous source of encouragement and help as my dissertation director; Dr. Randall Harrison, chairman of my dissertation committee was most helpful under very difficult circumstances; Drs. Donald Cushman, Joseph Woelfel and Eugene Jacobson, members of my dissertation committee, were all crucial influences in my professional development. Special thanks are also due to Dr. Gordon Thomas for his help during the final stage of this work, to my colleague, Ms. Nancy Richardson, for help throughout this research, and Mrs. Ruth Langenbacher for typing the manu- script and handling the many details required to finish up a Ph.D. program. Finally, I would like to express my appre— ciation to the Universidad Iberoamericana where I found the working and financial facilities to finish my dissertation. Chapter I II III TABLE OF CONTENTS THEORY . . . Introduction. . . . . . . Equity Theory . . . Attribution Theory . . Liking and Attribution of Responsibility. . . Communication Strategies and Attribution of Responsibility Intervening Variables and Attribution of Responsibility . METHOD . . . . . . . . . Design Subjects. . . . Procedures . . . . . The Trial . . . . . The Pretest . . . The Questionnaire . . . RESULTS . . . . . . . Equivalency of Subjects in the Experimental Groups . . Influence of Subjects' Background Characteristics on Attribution of Responsibility. . Subjects' Perceptions of the Experimental Variables. Perceived Positiveness of Content Perceived Negativeness of Content Perceived Ingratiation . . Perceived Openness . . . . iv Page GDU'IH 10 13 21 Chapter Page Chapter III (cont'd.) Subjects' Perceptions of the Intervening Variables . . . . 58 Perceived Honesty . . . . . 60 Perceived Similarity . . 60 Positive and Negative Labeling. . 61 Sympathy . . . . 65 Canonical Correlation Analysis. . . 68 Hypothesis 1 . . . . . . . 7O Hypothesis 2 . . . . . . . 73 Hypothesis 3 . . . . . . . 81 Summary . . . . . . . . 90 IV DISCUSSION . . . . . . . . . 92 Summary . . . . . . . . 92 Discussion. . . . . . 95 The Theoretical Framework . . . 96 Regarding the Victim-Related Variables. . 96 Regarding the Defendant— Related Variables. . . . . 97 Liking and its Dimensions . . 98 Regarding the Theoretical Rigor of the Equity Model . . . 99 The Independent and Intervening Variables . . . . 101 Perceived Honesty . . . 102 Positive and Negative Labeling. . 106 Similarity. . . . . . . 108 Sympathy . . . . 110 Generalizability of Findings . . . 112 Suggestions for Future Research . . 115 Regarding the Victim-Related Variables. . . 116 Regarding the Defendant- Related Variables. . . 117 Regarding the Mathematical Rigor of the Relationships . . 119 Regarding Other Related Fields of the Study . . . . 120 Conclusion. . . . . . . . 122 v Chapter APPENDICES Appendix A: REFERENCES The Experimental Question— naire . . . . I: Introduction . . . II: Juror Identification Form. III: Transcripts from Trial IV: Juror's Decisions. . . Transcript of the Tape- recorded Statements . Experimental Manipulations of the Independent Variables Means and Standard Deviations of the Variables in the Study Additional Intercorrelations Among Variables . . . . Canonical Weights and Canonical Correlations Between Indicators of Intervening Variables and Indicators of Liking . Vi Page 125 125 126 130 131 138 145 153 165 167 170 171 LIST OF TABLES The Overall Design of the Study The Design Under Analysis. . . . Summary of Measured Variables. Correlations Between Subjects' Background Variables and the Independent Variables. Correlations Between Subjects' Background Variables and Indicators of Attribution of Responsibility . . . Means and Analyses of Variance for Per- ceived Positiveness of Content, by Dis- closure Source and by Avowal Type . Means and Analyses of Variance for Per— ceived Negativeness of Content, by Dis- closure Source and by Avowal Type Means and Analyses of Variance for Per— ceived Ingratiation, by Disclosure Source and by Avowal Type . . . . . . Means amd Analyses of Variance for Per- ceived Openness, by Disclosure Source and by Avowal Type . . . . . . Means and Analyses of Variance for Per- ceived Honesty, by Disclosure Source and by Avowal Type . . . . Means and Analyses of Variance for Per— ceived Similarity, by Disclosure Source and by Avowal Type . . . . . . Means and Analyses of Variance for Posi— tive Evaluation, by Disclosure Source and by Avowal Type . . . . . . vii Page 28 30 41 46 48 51 53 55 57 59 62 63 Table Page Means and Analyses of Variance for Neg— ative Evaluation, by Disclosure Source and by Avowal Type. . . . . . . 64 Means and Analyses of Variance for Sym- pathy, by Disclosure Source and by Avowal Type . . . . . 67 Correlations Between Indicators of Liking and Indicators of Attribution of Respon— sibility for the Total Sample . . . . 71 Canonical Weights and Canonical Correla— tions Between Indicators of Liking and Indicators of Attribution of Responsi- bility for Four Cell and Sex-Cell Designs . . . . 74 Means and Analyses of Variance for Respon- sibility, by Disclosure Source and by Avowal Type . . . . . . . . 76 Means and Analyses of Variance for Guilt, by Disclosure Source and by Avowal Type . 78 Means and Analyses of Variance for Second Degree Murder Verdict, by Disclosure Source and by Avowal Type . . . . . 79 Means and Analyses of Variance for Not Guilty Verdict, by Disclosure Source and by Avowal Type. . . . . . . 80 Means and Analyses of Variance for Sentence, by Disclosure Source and by Avowal Type . . . . . . . . 82 Canonical Weights and Canonical Correla— tions Between Experimental Independent Variables and Indicators of Attribution of Responsibility . . . . . . . 83 Correlations Between Intervening Variables and Attribution of Responsibility Indicators. . . . . . . . 85 Canonical Weights and Canonical Correla— tions Between Indicators of Intervening Variables and Indicators of Attribution of Responsibility for Four Cell and Six- Cell Designs . . . . . . . . 88 viii Means and Standard Deviations of the Variables in the Study Correlations Between Indicators and Indicators of Attribution sibility (N=72). . Correlations Between Indicators and Indicators of Attribution sibility (N=108) . . . of Liking of Respon— of Liking of Respon— Intercorrelations of Intervening Variables for Four Cell and Six—Cell Designs Canonical Weights and Canonical Correlations Between Indicators of Intervening Variables and Indicators of Liking ix Page 165 167 168 170 Figure 1.1 LIST OF FIGURES Summary of Hypothesis Relating Commun— ication Dimensions and Attribution of Responsibility. A Model of the Relationship Between Message Strategies and Attribution of Responsibility. . . . . . Summary Version of the General Causal Model of the Study and Pair-wise Com- parisons Between Sets of Variables. . Causal Model of the Relationship Between Intervening Variables and AR Variables. Correlations Between Independent Vari— ables, and Error Terms of Indicators of AR not Included. . . . . . Interaction Between Perceived Honesty, Avowal Type and Interpersonal Evaluation. . . . . . . A Coorientation Model of the Juror- Victim—Defendant Relationships. Page 20 25 70 89 105 117 CHAPTER I THEORY Introduction In this thesis, an attempt is made to bring together some theoretical issues of importance for the discipline of communication, and some pragmatic concerns of relevance for the legal system in the United States. Thus, the research reported in it is based on the premise that, in many ways a trial is a specialized communication event amenable for scientific study following the guidelines prevalent in the discipline of human communication. This study focuses par- ticularly on one aspect of the information—processing func— tion of the jury system. Its purpose is the investigation of some of the communication variables relevant to the process whereby a juror attributes responsibility to a defendent and renders a verdict regarding his culpability or innocence. In the process of arriving at a decision a juror is exposed to several sources of information input. First, there is a formal communication occurring inside the courtroom where the juror attends to witnesses' testimony, lawyers' arguments, and judge's instructions. Second, there are the procedural characteristics of a trial such as the form of presentation of l 2 the information (e.g., adversary rather than inquisitorial), the order in which the information is presented including arguments and witnesses, and the number and variety of alter- natives given to the jurors in order to render their verdicts. Third, there is the informal communication inputs going on in the courtroom such as audience's reactions, counselors' cred- ibility, and a host of nonverbal cues such as the defendant's sex, physical attractiveness, race, clothing, and demeanor. Finally, there is the informal communication from the total culture impinging upon the jurors through literature, public opinion, interpersonal influences, and, more generally, his- tory and tradition. These general inputs of the culture are manifested in the personality traits of the jurors, as well as in their attitudes toward the issues discussed in a par— ticular trial. While legal practitioners have been aware of the importance of these variables for a long time, as Mitchell and Byrne (1973:123) point out, "only recently have systematic investigations of evaluative factors in the jury system been undertaken." This thesis reports only some aspects of a larger study designed to probe into the effects that some of the variables previously mentioned have upon a jury's decision-making pro- cess. This larger study sought the answers to three questions: (1) What are the effectsof the judge's instructions to the jury to disregard some information presented to them on the jury's decisions about the trial? (2) What, if any, are the 3 effects of informing a jury of its power to disregard the letter of the law and acquit a defendant, even if found guilty, solely on the basis of a different perception of justice (i.e., power 9f nullification)? (3) How does the use of self—presentation strategies by the defendant affect the verdict and other decisions made by the jury? The first of these questions relates to the role of moderator that the judge plays in the courtroom. Indeed, we have mentioned that there exist several extra-legal variables that have an impact upon the decision—making process of the jury. In theory, it is the role of the judge to minimize, as much as possible, the unwanted influence of such variables, and one of the means through which this arbitration is accom- plished is usually the judge's verbal instructions to the jury to disregard improper evidence. The extent to which such instructions are successful in accomplishing their ob— jective was one aspect of the overall study. The second question is related to the behavioral effects produced by the explicit avowal of the judge to the jury re— garding its right to acquit defendants without regard to law and evidence (i.e., the power of nullification). This issue has become increasingly important in recent years as the number of relatively mild crimes arising out of ideological positions has increased (e.g., war protests), and some sectors of the society have become sympathetic to the cause of the offenders. Thus, the problem is one where the actions of an 4 individual, while outside the law, have become accepted by the community. Finally, and most important for this thesis, the study was geared to explore different communication strategies that a defendant may follow in order to impress a jury in a favor- able way. In an extensive study involving 3,576 cases in trial, Kalven and Zeisel (1966) found that, in about one— third of the cases, the judges reportedly disagreed with the decisions rendered by the jury. The judges also felt that the disagreements were partly due to the juries' different evaluations of the defendants. Therefore, it seems crucial to determine the ways in which a defendant may influence the jurors' attitudes toward him. Goffman (1957) has documented a great variety of tech- niques or strategies of_self presentation that people use in order to influence interpersonal evaluations. Following Goffman's (1957) pioneering ideas, a host of writers and re- searchers have studied verbal and non-verbal techniques of self-presentation. Three such techniques directly relevant to this thesis are: self disclosure (Jourard, 1971), deviance avowal (Turner, 1972), and ingratiation (Jones, 1964). The work of these authors will serve as the basis for our theo- retical analysis of communication strategies in the courtroom. Given the breadth of this study and the large amount of data generated, it is difficult to present the theoretical rationales and hypotheses regarding all of the questions dis- cussed above, as well as the empirical results relating to 5 them. Therefore, in this thesis we will present the complete experimental design in order to provide the general context of the study. However, both the theoretical discussion and data analysis will refer mainly to the effects of self-pre— sentation techniques on a jury's attribution of responsibility of a defendant. The structure of this work is as follows. In this chapter we will review the theoretical background of the prob- lem under consideration, and the theoretical hypotheses to be tested. In Chapter II we will describe the experimental design, procedures, subjects and instrumentations. Chapter III will report the results obtained. Finally,in Chapter IV we will discuss the findings and analyze the implications of the study in general. Eguity Theory Mysliwiec (1974) has argued that the fairness of a system of justice depends on the principles of universal law and individualized equity. The principle of universality is encouraged and preserved in the laws of the country, while the principle of equity functions through the jury system. Indeed, in rendering a verdict a jury must apply the general law and, at the same time, take into consideration the special circumstances of a particular case. Given the importance that a jury's decisions have upon the well being of the com- munity, it follows that questions regarding a juror's sense of equity are of crucial social importance. Two of such 6 questions are: What are the relevant legal and extra-legal inputs (variables) that affect a juror's decision-making process? and, how are those inputs weighted and combined in order to arrive at a verdict? Mysliwiec (1974) suggests that these questions can and should be studied within the scientific framework of ag- tribution and equity theories respectively. The contention is that equity theory provides a conceptual equation for arriving at decisions regarding the fate of a defendant based on a variety of perceived rewards and costs experienced by the defendant and his victim in the course of their rela- tionship. Further, attribution theory provides guidance in identifying the relevant inputs and outcomes to be entered in the equity equation. Let us briefly discuss relevant aspects of each of these theories. In social psychology, equity theory has been elabo- rated by Homans (1961), Adams (1965), Blau (1967), and Walster §t_al. (1970) among others. The basic postulate of the theory is that an equitable relationship exists when the participants derive equal relative outcomes from the relation— ship. Adams (1965) expressed this principle in the equation: OutcomesA (rewards-costs) OutcomesB (rewards-costs) InputsA(assets—libilities) = InputsB (assets-liabilities) Thus, the relative outcomes mentioned in the postulate refer to the ratios of outcomes to inputs of the participants A and 7 B. Outcomes refer to the rewards and costs derived from a relationship. Rewards may range from money to affect, and costs may include time invested in the relationship or emo— tional stress. On the other hand, inputs refer to the assets and liabilities which entitle the participants to rewards and costs respectively. Assets may be personal attractiveness and honesty. Liabilities may include evil intentions and non—conformity. It should be made clear that equity is a subjective concept which varies from culture to culture and even from person to person. Thus, in a given relationship, one par- ticipant may perceive the relationship as equitable while the other participant may not perceive it as such. One such explanation for this disagreement might be the participant's perceptions of what inputs and outputs should be considered in the equation, or the relative importance that should be assigned to them. A second general postulate of the theory posits that when individuals find themselves in an inequitable relation- ship, they become distressed and attempt to eliminate their distress by restoring equity. An individual may restore actual equity by altering his own inputs or outcomes, or the inputs or outcomes of the other participant. On the other hand, an individual may re- store psychological equity by changing his perceptions of the inputs and outcomes derived by himself and/or the other participant. 8 This discussion of equity theory is highly relevant to the problem of jury equity if we consider that as Legant (1973:3) points out, in civil law, inequity "exists between the plaintiff and the defendant; in criminal law it lies be- tween the offender and his victim or, as the law sees it, between the offender and the state, or the larger society which the state represents." Moreover, observers of an inequitable relationship have been found to experience distress "in much the same way as do participants in the inequitable relationship and are thereby similarly motivated to restore actual or psycholog- ical equity . . ." (Mysliwiec, 1974:21). Thus,given that the juror is the observer in disputes over legal inequities, we would expect the juror to feel a strong need to restore equity. In such case, the equity equation previously pre- sented can serve as a theoretical model of the juror's sense of equity. Attribution Theory Equity theory, however, does not provide any guidance for identifying the relevant inputs and outcomes of an equit— able relationship. Fortunately, researchers working on at— tribution theory1 have carried out extensive work in two areas relevant to the problem. First, they have determined some of the inputs (assets and liabilities) which entitle the participants in a relationship to rewards and costs. Second, they have studied the process whereby the existence 9 of these inputs is inferred by the observer (e.g., juror) of a relationship. The area of attribution theory which is most relevant to our present concerns is that dealing with attribution 9f responsibility (AR). Indeed, assignment of responsibility of a defendant is the most important means through which a juror restores equity in the relationship between the de— fendant and the plaintiff or the victim. Thus, let us now turn to the theoretical analysis of AR and the discussion of some of the variables which have been found to be impor— tant inputs in the equity equation and therefore to have a crucial effect on AR. Fishbein and Ajzen (1973:149) contend that "the term attribution of responsibility can perhaps best be viewed as a moral judgment." The authors substantiate this view by presenting Heider's model of AR which includes five levels. (1) Association: the person is held responsible for all events that are in any way associated with him. (2) Com— mission: the person is held responsible whenever he is per- ceived to be instrumental in producing the events. (3) Fore- seeability: the person is held responsible if he could have foreseen the consequences of his actions. (4) Intentionality: the person is held responsible only for those events which he intended to produce. (5) Justification: the person is held responsible to the extent that he could not control all of the factors influencing his behavior. 10 Fishbein and Ajzen (1973:150) argue that, from this category system, it follows that a question such as, 'Is the actor respon- sible for the accident?‘ can be interpre- ted in different ways: (1) Was the actor associated with the accident?, (2) Was he instrumental in producing the accident, i.e., did he cause it?, (3) Is he respon— sible in the sense that he could have fore- seen the accident?, (4) Did he intend to cause the accident?, and (5) To what extent was his behavior justified? The relevance of this discussion for the investigation reported in this thesis is manifold. First, a jury sitting in judgment of a defendant usually has to make decisions based on its attributions regarding internal states of the defendant such as intentionality or voluntariness. Second, Heider's model of levels of AR is not only intuitively sound but it closely parallels formal distinctions made in criminal law for differentiating between charges such as manslaughter (unfore- seeable and unintentional crime) second degree murder (the crime is intentional but partially justifiable) and first de- gree murder (the crime is intentional and unjustifiable). Finally, in our research we are interested in exploring some variables which may serve as inputs in the equity equation and therefore may affect the responsibility attribution of jurors. Let us now turn to the conceptual analysis of some of these variables. Liking and Attribution of Responsibility Recent studies on jury decision—making have uncovered a number of variables which can be considered as inputs in 11 the equity equation and therefore affect jury decisions. Some of these include race, sex, income, education, family status (Nagel, 1969), juror-defendant similarity of attitudes (Mitchell and Byrne, 1973), and attractiveness of defendant and victim (Landy and Aronson, 1969). Of these variables, in this work we have chosen to deal only with the defendant's attractiveness variable for two reasons. First, from a pragmatic point of View, attrac— tiveness is a variable which has shown consistent, strong impact on AR, and therefore clarification of its effects is an important practical endeavor. Second, from a theoretical perspective, attractiveness is an important variable in the study of human communication in that it influences and is in- fluenced by social interaction (Newcomb, 1961). An assumption often made in research dealing with equity theory is that attractiveness of an individual (i.e., liking for an individual) is a positive input in the equity equation. This assumption has been supported in studies by Landy and Aronson (1969), Shaver (1970), Sigall and Landy (1972), Mitchell and Byrne (1973), and Nemeth and 80818 (1973). Landy and Aronson (1969) asked subjects to read a brief description of a negligent homicide case which was identical for all subjects except in the description of the victim. In a second experiment, the description of the de— fendant's character was also varied. The results show that simulated jurors view a crime as being more serious if the 12 victim is an attractive person and/or if the defendant is an unattractive one. In subsequent replications of the study by Sigall and Landy (1972) and Nemeth and Sosis (1973) similar results were reported. These studies suggest that a defendant who is liked will be perceived in an inequitable relationship with his victim (especially if the victim is disliked). The reason for such perception is that attractiveness is perceived as an asset which entitles the defendant to greater rewards and/ or diminished costs. Equity then may be restored by attrib— uting less responsibility to the defendant.2 Thus, Hypothesis 1: The greater the liking of a juror for a defendant, the lesser the responsibility that will be attributed to the defendant. If this relationship continues to be supported by em- pirical data, one may ask what are the possible communication alternatives open to a defendant who wants to increase his attractiveness and therefore receive lighter verdicts by a jury. Clearly, the strategies of self presentation that a defendant may choose are greatly restricted by the rules of law and evidence. For instance, while the defendant may want to present an extended account of his positive traits and altruistic activities, such an account could immediately be challenged and curtailed if it is irrelevant to the nature of the trial. 13 However, within these restrictions, a defendant still can exert some degree of control upon the information to be disclosed and the source of such disclosure. In the remain- der of this chapter we will discuss these two variables (content and spurge of disclosure) and the intervening mechan- isms through which they may relate to liking andixiturn to AR. Communication Strategies and Attribution of Responsibility Generally, we may say that the actions of a given in- dividual can be roughly perceived and categorized as being either positive, negative or neutral. When such actions are described in a message, they become the content of the mes— sage, and the same categorization (i.e., positive, neutral and negative message content) can be applied. The questions that concern us in this work are: How do jurors interpret a defendant's disclosure of positive or negative messages about himself? What are the effects of such interpretations on the juror's liking for the defendant? And finally, what are the effects of such liking for the defendant on subsequent attributions of the defendant's responsibility? In previous sections of this chapter we hypothesized an answer to this last question stating that an increase in liking for the defendant will result in decreased AR. We will now review the issues involved in the first two questions and develop some hypothetical answers. A simplistic approach would lead us to believe in a straightforward relationship between the content value of a l4 self-presentation message (e.g., positive or negative) and subsequent liking toward the source: If the source says positive things about himself, his attractiveness will be in- creased in the eyes of the receiver. On the other hand, neg— ative self-presentation would lead to lesser attractiveness. Empirical studies, however, do not bear such over— simplified relationship. Although the content-value of mes- sages does have an impact on interpersonal liking, its main effect is not as crucial as its interaction effect with other variables. One of these variables is the importance or rel— evance of the message content. The interaction is such that if the message is irrelevant or unimportant, no change in liking will occur. Another important variable is the receiver's percep— tions of the source's intentions for delivering the message. If the self-presentation message is perceived as a conscious intent to obtain rewards or avoid punishment by manipulating the feelings of the receiver, an inverse relationship be- tween message content value and source's attractiveness will occur. Often, the context in which the message is sent will be used by the receiver as a cue to decode what the source's intentions are. In the context of the courtroom where the jurors are sitting in judgment of a defendant, it is likely that they will be very sensitive to the "hidden" intentions behind the messages sent by all the other social actors, in- cluding the defendant. Thus, the defendant's negative self— avowal (a liability, in termsof content) may not be taken at 15 face value but be construed as honesty (an asset for the defendant). Conversely, positive self—presentation (asset) may be taken as ingratiation (liability). Several concepts in the social psychological literature bear on this issue and they will be reviewed below. There are several theoretical notions which are rel- evant in the discussion of possible effects of negative self- presentation. Literature in the area of self disclosure reports that self avowal of negative information is con- sistently construed as self-disclosing behavior. Further- more, it has been found that to receive self—disclosing information is socially rewarding (Worthy et_al., 1969); it indicates to the receiver that he is trusted and liked (Jourard, 1971); it tends to create sympathy towards the source, and it enhances the image of the source as an honest person (Jourard, 1971). Berscheid and Walster (1969) have argued that we like those from whom we receive rewards, liking and trust. Therefore, we would expect that a juror who feels trusted and liked by the defendant, will return the liking and trust (Worthy ep_al., 1969), and will be less likely to attribute responsibility for negative actions to the defendant. Jones and Davis (1965) have offered a theory of attribution which further clarifies why a strategy of nega— tive self disclosure (i.e., deviance avowal) would increase the defendant's attractiveness. These authors are concerned with the process whereby peOple infer an individual's 16 internal dispositions from observable events. The basic thesis of the theory is that the significance of an action derives from the alternatives open to the actor. Further- more, given that social roles specify the most likely altern- ative of social behavior, "the performance of social roles tends to mask information about individual characteristics . . ." (Jones gt_al., 1961). Jones and Davis (1965) intro- duce the term correspondence, which refers to the extent to which a person's act and underlying disposition are similarly described by a perceiver's inference. In general, the attri- bution ofeaparticular trait on the basis of a given action will be made more confidently (will be more correspondent) if the action departs from normative expectations. Thus, when a defendant is being tried for an offense, one of his role expectations is that he will not engage in any type of deviance avowal.4 Should such expectation be violated by the defendant, such violation could be interpreted by the jurors as a revelation of the defendant's underlying dispo— sition to be honest. Finally, in a study dealing with source credibility and persuasiveness, Walster, Aronson and Abrahams (1966:325) argue that "any communicator, regardless of his prestige, will be more effective and will be seen as more credible when he is arguing for a position opposed to his own best interest, than when arguing for changes obviously in his own best interest." This proposition was empirically supported in two studies conducted by the authors. 17 At this point it should be clear that the common de— nominator in the theoretical rationales just presented is the notion of a defendant's honesty. It will be seen that this same notion of honesty provides the bases for our theoretical discussion of positive self-presentation by a defendant. Indeed, when a communicator presents self in a positive manner to a receiver who has the power to fulfill an immediate and relevant need of the source, such communication may be con- strued as ingratiation and, as a result, may lower the liking for the source. Jones (1964) describes ingratiation as the attempt to increase one's attractiveness to another in the hope of gain- ing something of value in return. The author discusses three modes of ingratiation: self—presentation of ones positive attributes, conformity to the opinions of another, and flat— tering the other person. The choice of ingratiation tactic depends on the situation. Clearly, in a defendant-juror relationship the defendant is not in a position of using either conformity or other enhancement techniques. If the defendant chooses the strategy of positive self-presentation, such communication may be perceived as dishonest and produce an unsympathetic attitude toward him. In all of the theoretical approaches discussed so far, theorists have made predictions relating positive and/or negative self-presentation and attractiveness which are based on the interaction between the content of the disclosure and 18 the context in which such disclosure is made. Such inter— action then provides the basis from which jurors may infer the defendant's latent dispositions and personal character— istics. Thus, a defendant who avows deviance may be seen as a courageous, honest individual. On the other hand, a de— fendant who presents himself in a positive manner may be seen as a dishonest manipulator. However, besides such interpretations and inferences as to why a communication is made, there exists the aspect of what the communication says, or, in other words, the con- tent of the communication itself. It is clear that if a juror were to consider only the content of the disclosures made by a defendant, predictions regarding the relation between strategies of self—presentation and attractiveness would be opposite to those previously advanced. Let us elaborate on this argument. Working within the framework of attribution theory, Kelley (1967) has postulated that an individual will attribute the causality of a particular event to an entity if (1) the event occurs when the entity is present and does not occur when the entity is absent (i.e., the principle pf covariation), (2) the event is consistent over time, (3) the event is con- sistent when the mode of interaction with the entity varies, and (4) there is consensus among observers regarding the nature of the event. 19 Based on these notions, one would conclude that the more instances of deviant behavior provided by a defendant to a jury, the more the defendant will be identified (i.e., labeled) as a deviant individual and the greater the dislike for him. Furthermore, negative information about the de— fendant may also decrease the juror's perceived similarity with the defendant, and therefore decrease liking. This same kind of rationale could be applied to a de- fendant's positive self-presentation. That is, the more in- stances of positive behavior provided by the defendant to the jury, the more likely that he will be identified as an honor- able individual who has made a mistake, and therefore deserves a break. In order to probe into the validity of these contra- dictory hypotheses, it is necessary to separate out the ef- fects produced by message content (i.e., avowal type) from the effects produced by the defendant's self—disclosure (which may result in inferences regarding his motives for such disclosure). One way of doing so is by manipulating the source pf disclosure. Indeed, by allowing the informa— tion to be presented by an individual other than the defen— dant himself, it is possible to break up the context which allows for jury inferences regarding ulterior motives of the defendant's avowal type and, at the same time, isolate the effects of the information content. 20 We are aware that such a move may produce some influ- ence of other variables such as the source's credibility, appearance, etc. However, such variables can be experiment- ally controlled so that, at least in theory, the problem is ameliorated. Furthermore, such a move greatly increases the practical significance of the study since most of the infor- mation pertaining to a case in trial is usually presented by witnesses rather than the defendant. Thus, we can summarize the alternative hypotheses pre- sented in this section in Figure 1.1. . a Disclosure Source Self Other Negative Xll < X 12 21 1 X22 Avowal Type Positive i aCell means refer to attribution of responsibility scores. Figure 1.1 Summary of Hypothesis Relating Communication Dimensions and Attribution of Responsibility. For purposes of clarity, the interaction presented in Figure 1.1 is presented in a two-part hypothesis: Hypothesis 2: There is an interaction effect between disclosure source and avowal type such that, (a) Under conditions of negative avowal, less responsibility will be attributed to a defendant if 21 deviance is d18closed by the defendant himself rather than by a witness. (b) Under conditions of posi— tive avowal, less responsibility will be attributed to a defendant if the positive information is disclosed by a witness rather than by the defendant himself. Intervening Variables and Attribution of Responsibility It is important to note that this hypothesis takes for granted that a linear relationship between liking and AR exists. This assumption is not so troublesome because it has been expressed in Hypothesis 1 which will be empir— ically tested. However, we have not accounted for the var- ious intervening mechanisms posited in the theoretical rationales in this section. Such intervening variables were basically four: honesty, similarity, labeling and sympathy. These variables are posited as the channel of influence from the communication strategies to the liking for the defendant. Therefore, their empirical study is crucial if we want to be able to support the theoretical rationales from which they were derived. In other words, if we did not measure and analyze the effects of these intervening variables indepen- dently, the following consequences would occur: (1) a sig— nificant relation between communication strategies and AR could be explained through a number of theoretical rationales; however, none of those rationales could be empirically support— ed; (2) if we found the relation between communication and AR 22 not significant, we would have no way of knowing whether the experimental manipulations actually triggered the psycholog- ical mechanisms we had expected to set in motion. Accordingly, in studying the effects of these inter— vening variables on AR we will again take the relationship between liking and AR for granted. Further, while we assume that the intervening variables exert some of their influence on AR through liking, alternative rationales and hypotheses regarding the relationship between these variables and AR are presented below. Hypothesis 3a: The greater the perceived similarity of a juror with the defendant, the less the AR. Similarity refers to the process whereby a receiver perceives self as similar to the source. Thus, if the infor- mation provided in a communication is deviant in nature (e.g., "I am a homosexual"), we would expect that the receiver will perceive himself as less similar to the source (unless, of course, the receiver also shares homosexual tendencies). Both Shaver (1970) and Berscheid and Walster (1969) have reasoned that, when an individual perceives self as similar to another individual who is undergoing a stressful situation, such perception is threatening to the perceiver in the sense that it heightens the possibility that he could find himself in the same situation. In a juror—defendant relationship, if a juror perceives the defendant to be similar to self, the dissonance aroused by the situation may be 23 lessened by attributing less responsibility to the defendant. Hypothesis 3b: The greater the positive label- ing of a defendant on the part of a juror, the less the AR; and the greater the negative labeling the greater the AR. The process of stereotyping has been posited in var- ious perceptual theories as a mechanism for providing struc— ture to our perceptions of an environment in constant flux. Such structure, deeply rooted in our language, is basically a category system which allows us to differentiate and label objects in the environment. For instance, a person who admits being an ex-convict may create negative attitudes towards him/her from the aud- ience, simply because of the information transmitted is neg- atively valued. Once an object or a person has been placed in a certain category, such a label serves as a basis for subsequent evaluations toward the object or person. In the juror-defendant relationship, if a juror, on the basis of some information, labels the defendant under a positive cate- gory, say, as a victim of the circumstances, his general attitude toward the defendant should be more positive and result in less AR. It is also intuitively sound to expect that if juror perceives the defendant as a deviant (negative attitude) person, greater responsibility will be attributed. Hypothesis 3c: The greater the sympathy toward a defendant on the part of a juror, the less the AR. 24 Sympathy refers to the extent to which the communica- tion produces sympathetic feelings from the receiver toward the source. For instance, a message indicating that its source has experienced some negative outcomes or unjust suf- fering is likely to produce sympathetic feelings on the part of the receiver. In some of the studies dealing with the relationship V between liking and AR, the variable of liking has been man— ipulated by presenting a characterization of the defendant as a person having suffered various undesirable outcomes. For instance, one of Landy and Aronson's (1969) manipulations of a defendant's attractiveness includes the description of whether the defendant suffered loss of sight in one eye during the course of a manslaughter offense (attractive condition) or did not suffer any injury (unattractive condition). As Sigall and Landy (1972) point out, the attractiveness of an individual and the feelings of sympathy towards him are two different variables. In terms of equity theory, sympathy results from the perception that a person has incurred some costs. From this, it follows that the actor deserves increased rewards or decreased punishment. In the case ofaipresumed criminal, this outcome would be reflected in less AR. Hy— pothesis 3c is, of course, a straightforward derivation of this discussion. Hypothesis 3d: The greater the perceived honesty of the defendant, the less the AR. 25 Honesty refers to the extent to which the receiver feels a source was honest in his communication. For instance, as discussed earlier, negative self disclosure may be per- ceived as more honest than self-ingratiating information. The rationale again is that honesty can be viewed as a positive input in the equity equation, and therefore a person who is perceived to be honest should not be attributed too great a responsibility for negative actions. By the same token, a person who, after having committed a crime, is also perceived as dishonest, will be held responsible to a greater extent. Let us close this chapter by presenting a model (Figure 1.2) which summarizes the variables and relationships presented in this chapter. It should be noted that the model does not, and is not intended to provide for, unequivocal determination of all possible interrelationships among the variables. The model is intended to serve as representation of the causal flows of influence which we have hypothesized in this work. DEEPHEENT\ERLMHES INHfiNENDREVARDflflES DEMEDEMPVARD¥EES labeling Shmflsrity ‘ . . 9 : Awmml'hfie ++> S th ++> 'Lflung'a++lAR! Disclosure Source EHEEL y anesty Figure 1.2. A Model of the Relationship Between Message Strategies and Attribution of Responsibility. 26 FOOTNOTES lAttribution theory derives from the writings of Heider (1958). The basic premise underlying the theory is that a person has a need for perceptual order or a need to find the causes of events (changes) occurring in the environment. The process whereby the person searches for such causes is called attribution, and the goal of this process is to identify dispositional properties of entities, either in the environment or in an individual, to which the changes in the environment could be attributed. 2It should be noted that the study reported in this work deals with the judgment of a presumed criminal by a jury. Thus, the AR necessarily refers only to negatively valued acts. 3Jourard (1971), has defined self disclosure as "the process of making the self known." In order to have a more specific definition, by self disclosure we will understand the trans- mission of information which is very private and intimate in nature. 4An extreme case of this expectation has been formalized in the law which grants a defendant the right to avoid dis— closure of self-incriminating information. CHAPTER II METHOD Design In order to test the hypotheses presented in the pre- vious chapter (as well as other hypotheses not presented in this work),a 13—ce11 study of a simulated trial was designed (see Table 2.1). Subjects (Ss) were asked to imagine them- selves as jurors and render a judgment regarding the respon- sibility (dependent variable) of a defendant in a murder trial. In the case presented to the 85, the defendant (Johnny Marco) was accused of murdering his father while attempting to de— fend his sister from the father's attacks. In this study, four independent variables were manipu- lated in the overall design. First, the source of disclosure was manipulated by either having the defendant self—disclose personal information (self disclosure condition), or by having a witness present the same information (other disclosure condition). Second, the avowal type was manipulated by pre- senting messages either positive (positive avowal condition) or negative (negative avowal condition) in their content. There was also a control group where neither positive nor negative information was presented (neutral condition). 27 28 .mnobESC Haoo pocmflmmm ucwmoummu mach“ ocp CH mumbesc one Q .xuflaflnflmcommou mo soapscflupum some mcu ma oHnMHHm> usuccmmmw CHME mafia ma NH HH onocmH #02 OH m o v N cofla IMOHM IHHHDZ a a m m H cusp laugh IHHHSZ Hm30>¢ Hm30>¢ m>fluflmom Hmsokm m>flpmmoz Hmunsmz mHsmoaomHQ Hmcuo oHocmH #02 muocmH mHOcmH mHocmH muocmH #02 “OZ Hmzo>< ammunmz Hm30>< m>fluflmom Hmso>< m>flummmz musmoHomao mama m.suspm may we cmammo assumso one .H.m magma 29 Third, the stricken testimony variable was manipulated by having the judge instruct the jury to disregard certain information (ignore condition) or by deleting this instruc- tion from the transcript presented to the 85 (not ignore condition). This manipulation occurred within both the positive and negative avowal conditions, and only in the self- disclosure condition. The actual transcripts identifying the messages presented to the Ss in every combination of in- dependent variables discussed so far are presented in Appendix C. Finally, the power of nullification variable was manipu- lated as the presence (nullification condition) or absence (n9 nullification condition) of the following sentence: However, it is in your power as a jury to go against the instructions given to you in this case, if you deem it necessary in order to best serve the interests of justice. This sentence was inserted in the final instructions of the judge to the jury, and presented to the 85 as part of the experimental stimulus. Note that in the other disclosure condition all 55 were exposed to the nullification treatment. The first 10 cells in the design represent combinations of the 3 independent variables: avowal type, stricken testi- mony, and power of nullification. The last three cells re— flect the manipulation of the source of disclosure. Not all logical combinations of the independent variables were studied since this would have implied a 20 cell design difficult to manage and interpret. 30 Again, it should be stressed that in this thesis we will concern ourselves only with a partial analysis of this design.1 Such analysis includes data from cells 3, 7, 9, ll, 12, and 13, and is presented in Table 2.2. Table 2.2. The Design Under Analysis. Disclosure Source Self Other . a o Negative 3 11 a h B Q Positive 7 12 3 8 a Neutral 9 13 i aAll cell numbers correspond to those specified in Table 2.1. Note that in this design, the presence of the nulli— fication and the absence of the judge's instructions vari- ables are kept constant for all cells. Therefore, the only independent variables being manipulated are source of dis- closure and avowal type. Subjects The 260 882 who participated in the study were en- rolled in nine undergraduate courses in the Departments of Psychology, Communication, and Urban Development at Michigan State University. One hundred and thirty-eight of the Ss 31 were male, 120 were female, and two did not indicate sex. Since N=20 for every cell in this study, 6—ce11 design to be analyzed in this dissertation is based on a total of 120 85 except for some variables where missing data were found and 12 Ss had to be randomly dropped so that cell sizes would be equal to 18. Such variables are clearly specified in the section dealing with the analysis of the data. Procedures The experimental sessions were conducted during regu— lar class periods. The 85 were told that their participation in the study was voluntary, and they were generally very cooperative; only an average of two students in each class refused to participate and left the classroom. The experimenter3 (a female graduate student in the Department of Communication), after introducing herself, ex- plained to the SS the procedures to be followed. Such expla- nation took approximately two minutes and was standard for all classes. Good morning; my name is Nancy Richardson. I am a research assistant in the Communication Depart- ment where we are presently engaged in some research regarding communication issues in the legal system. Today, you will be asked to read carefully the doc- uments I am about to hand out. Then, I will play a tape recording, and finally, you will answer some questions attached at the end of the handout. You should remember two important points: First, in this study we will ESE trick you, lie to you, or otherwise deceive you in any way. Therefore, we ask you to answer the questionnaire in all honesty. If at the end of the session you have a2y questions, I will be happy to answer them for you. 32 The second important point is that you should read carefully and thoroughly each and every page of the document since every bit of information is very important. I will now hand out the questionnaires. The instructions in the booklets are clear and self- explanatory. A booklet containing the experimental stimulus, ma- nipulations, and measurements was randomly distributed with— in each class (see Appendix A). At this point the experimenter's instructions were as follows: Please begin to read the documents as soon as you receive them. Read until page 10 and then stop. It will take some people longer to read than others. Those who finish first, please do not talk to your neighbor, please just wait quietly until the other members of the class have finished. The experimental stimulus consisted of: (a) an introduction of the issues involved in the study, and the instructions as to how to proceed with the experiment (the introduction pointed out some shortcomings of the jury system as it pre- sently functions and stressed the importance of the 88' re— sponses as inputs to be considered in future recommendations to change it); (b) a questionnaire asking the same kind of information asked from real jurors in pre-trial examinations (this questionnaire had the double purpose of collecting information about the Ss, and bolstering the realism of the experiment); (c) a transcript of a case which had been edited and condensed into 5 pages of written information, and 12 minutes of tape—recorded arguments. 33 The written part of the transcript contained the fol- lowing information. (1) A statement of the court clerk re— garding the charges against the defendant and the defendant's plea of not guilty to the charge of second degree murder. (2) The opening statements (summarized) of the prosecution and the defense. The prosecution would attempt to show that Johnny Marco fatally stabbed his own father in cold blood. Prosecution asks for a verdict of 2nd degree murder. The defense would attempt to show that Johnny had never had a decent home life and that, on the night of the killing, he had been goaded by his father beyond endurance and had final— ly stabbed him in the heat of a struggle. Defense asks for a verdict of manslaughter. (3) General information about the defendant, the victim, and the defendant's sister who was also involved in the case. (4) Background testimony of the policeman who arrested the defendant and the coroner who ex— amined the victim. The policeman testifies that, at the time of the arrest, Johhny had told him "I killed him (Johnny's father), but he asked for it." The coroner testified that the deceased had suffered three wounds in his chest, one of which pierced the heart and was fatal. (5) The defendant's testimony (in the self disclosure condition), or the testi— mony of a friend of the defendant (in the other-disclosure condition) of either positive, negative or neutral informa— tion about the defendant. All these combinations of testi- mony are presented in Appendix C. However, as an example, 34 let us present the self—disclosure, negative avowal con- dition: Defense: Johnny: Defense: Johnny: Defense: Johnny: Defense: Johnny: Defense: Prosecution: Defense: Judge: Johnny: Defense: Johnny: Johnny, how old are you? Nineteen years old. Where did you attend high school? George Washington High School. Johnny, where do you live? 1405 East 103£d Street. Who lived with you at that address? My father and my sister. How would you describe your relations with your father? Your Honor, I object. That question is not relevant to this case. Your Honor, I believe it is relevant. I will attempt to show a history of hostility and provocation between Johnny and his father. The court will withhold ruling on the objection and will allow defense counsel a few minutes to demonstrate the rele- vance of this line of questioning. Johnny, please answer the question. Well, things were not that good between my father and me. We had some rough times together. He really wasn't home that much, and neither was I. I spent most of my time with the Raven's street gang. We pulled the robbery of the neighborhood grocery store last year. Johnny, how long have you lived at the 1405 East 103£d Street address? For the last eight or nine years. (trial continues) ’P’ 35 After all of this information was read by the 85 (this reading took about 10 minutes), they listened to a tape re— cording (Appendix B) containing the closing statements of the defense (7 minutes) and the prosecution (5 minutes). The experimenter's instructions were: I will now play the tape recording of the closing statements in this trial. Please listen very carefully, since you may be getting new in— formation. You will first hear the defense attorney's closing statement and then the prose— cuting attorney's closing statement to the jury. The main points in the defense's statement were: (1) that Johnny had had a poor upbringing complicated with family problems and Violent surroundings; (2) that Johnny loved his sister very much and therefore was highly provoked when his father attacked her; (3) that Johnny's father was drunk (as usual) and violent the night of the murder. More— over, since Johnny's father was a strong man, Johnny had to frighten him away from his sister using a knife and, in the heat of a struggle stabbed his father to death. The prosecution made the following points: (1) that Johnny was a violent, socially disruptive individual who had taken "the easy way out" in life; (2) that Johnny should have been able to handle his father's rage without pulling out a knife because his father was defenseless and clumsy with drunkenness; (3) that Johnny's motivation was not to protect his sister but rather to injure his father, and that the three wounds Johnny inflicted to his father were not acciden— tal. 36 After the tape had been played, the 55 were asked to read the judge's final instructions, and then to respond to a questionnaire. These activities were usually carried out in a period of 20 to 25 minutes. At this point, the experimenter's instructions to the Ss were: The next section is the transcript of the judge's instructions to the jury. Please read them very carefully. You will then be asked to fill out the last section of the document and answer some questions. Please do not look back into the document——before page I0-:w3en you are answering the questions. This is very important. Please answer every question. You will notice that there are usually three boxes in the left hand column for your answer. Ignore the numbers above the boxes. They are there for coding purposes only. We have provid- ed three boxes in the event you wish to answer a question 100%, in which case you would use all three boxes. If you answer less than 100%, place a zero in the first box (examples put on black- board). Please fill in every box. If you have any questions, raise your hand and I will try to assist you. When you are finished, please raise your hand and I will pick up your booklet. If you must leave at that time, please remember not to discuss this document with people outside this class. Once again, please read the judge's instuc- tions and the questions carefully, and give your answers the seriousness and thought they warrant. Thank you. In the experimental stimulus the judge instructed the jury as follows: As to the question of guilt, it has already been established that the defendant killed his father. It is up to you as a jury member to de— cide whether or not he is guilty of second—degree murder. To help you make your decision, we repeat here the charges against the defendant: 'It is charged that the defendant willfully and deliber- ately stabbed his father, causing him to die, 37 although the act of murder was not premed- itated nor planned, and that the act was not committed under sufficient mitigating circum— stances to relieve him of criminal respon— sibility before the law.‘ You may find him guilty of this offense or of the lesser offense of manslaughter, or may find him not guilty as charged. You have been instructed, during the course of this trial, as to the applicable law in this case. You have also been advised that if Johnny is legally guilty of killing his father, he must be found guilty of either second degree murder or manslaughter. Your deliberations are con— fidential and your conclusions cannot be challenged. After reading these instructions, 85 responded to a questionnaire. Upon completion of this instrument, the ex- perimenter answered questions and later dismissed the 85. Other than explaining the nature of the variables involved in the study, no debriefing took place since there were no concealed intentions, manipulations or measurements. The total amount of time required to run the experiment in a class was from 50 to 60 minutes. The Trial The experimental stimulus used in the study was an actual case originally tried in the state of New York (The State of New York vs Johnny Burdizk). This case was later adapted by Sears (1965), and Zillman and Cantor (1974). For this study, we added the experimental manipulations to the Zillman and Cantor adaptation. 38 The Pretest In order to test the appropriateness of the case as experimental stimulus, as well as the experimental procedures and instrumentation, an extensive pretest was conducted with 96 undergraduate students attending 6 introductory courses in the Department of Communication at Michigan State University. The results of the pretest showed that the case was indeed interesting and believable for the Ss who consistently showed great involvement and seriousness during the experi- mental sessions. At the end of the sessions, very frequently the Ss asked the experimenter what had been the outcome of the actual case, and this was also taken as an index of their interest. On the bases of the pretest, some questions in the questionnaire were reformulated so that they would become more understandable. Since it was found that the 85 required a long time to fully answer the questionnaire, four open- ended questions were dropped. These questions had been de- signed to measure how well the 85 had comprehended the evi- dence presented in the case. While such information was valuable, the time required for its gathering exceeded the time availability of the 85. Also, six new multiple choice questions were added: subjects' felt freedom of decision (question 24), perceived importance of judge's instructions (question 25), perceived defendant's openness (question 26), 39 and three new manipulation checks (questions 30, 31, and 32). In the pretest it was also observed that some 88 looked back on the case description in order to answer some of the questions. This observation lead to an increased emphasis by the experimenter that such a practice was not permitted. In the actual study the Ss apparently complied with these instructions. Subjects raised no questions regarding the clarity of the instructions or the stimulus material during the pre- test. Accordingly only two changes were made. First, a paragraph in the prosecutor's closing statement was deleted because it made reference to the defendant being a member of a street gang. Such deletion was due because of the pos— sible unknown interaction effects that it could have with the experimental manipulation of avowal type. Second, the order of presentation of the prosecution and the defense closing statements was reversed to comply with the actual order that would be followed in a courtroom. Thus, in the final experimental stimulus the prosecution presented its opening statements first and the defense second. Then, the defense presented its closing statements first and the pros— ecution last. Finally, the pretest served as training for the experimenter who became proficient in following the ex— perimental procedures before the actual study was started. 40 After all these changes, the distribution of time in each run of the experiment was approximately the following: Initial verbal instructions and dis— tribution of the written materials ....... 5 minutes Time for reading pages 1 to 10 ........... 12 minutes Time for playing the prosecution speech. ................. . ..... ..... ...... 6 minutes Time for playing the defense speech ...... 7.5 minutes Verbal instructions about how to answer the questionnaire.. ..... . ....... .. 2 minutes Time for answering the questionnaire and coding the responses................. 23 minutes Total Time ...... . ..... ................... 55 minutes The Questionnaire All of the intervening and dependent variables pre— sented in our theoretical model (Figure 1.2) were measured with the questionnaire presented in Appendix A. In Table 2.3 we present a summary of the variables measured, their location in the questionnaire, and the type of scale used for their measurement. We will now turn to the presentation of the results obtained in the study. Table 2.3. Summary of Nbasured variables. 41 Identification variables Question Number Scale 1. Sex A Male/female 2. Age B Year of birth 3. Father's occupation C Open ended 4. Family size D Number of brothers and sisters 5. Order of birth E First born Later born Only child 6. Family economic status F Considerably above average Somewhat above average Average Somewhat below average Considerably below average 7. Father and Mother education G Less than 8 grades 8 grades 9-11 grades 12 grades Graduated high school Some college College degree Advanced degree 8. Perceived fairness of jury system H Open ended Dependent variables 1. Attribution of responsibility: (a) Accident 1 0—100 percent (b) Responsibility of Johnny 2 0-100 percent (c) Responsibility of sister 3 0-100 percent (d) Responsibility of father 4 0—100 percent (e) Guilt 5 0—100 percent Table 2.3 (cont'd.) Dependent variables (f) verdict (g) Sentence (h) Moral justification of crime Liking (a) Desire to meet (b) Closeness (c) Admiration (d) Potential liking (e) Initial liking Intervening variables 1. 2. Similarity Honesty labeling (a) Negative (b) Positive Sympathy Control variables Confidence in responses Adjustnent to role-playing Perceived freedom of choice Perceived openness 42 Question Number 6 13 14 15 16 17 12 10 18 19 20 21 24 26 Scale Guilty of second degree murder Guilty of manslaughter Not guilty 0 to 30 years inprisonnent 0—100 percent 7 intervals 7 intervals 7 intervals 7 intervals 7 intervals 7 intervals 0-100 percent l-lOO percent 5 choices 5 choices 7 intervals 7 intervals 7 intervals 7 intervals 43 Table 2.3 (cont'd.) Dependent variables Question Number S9213 5. Perceived ingratiation 11 0—100 percent Manipulation Checks 1. Attention to experinental task (a) Retention of information 22 Johnny's age (b) Retention of information 23 Year of incident 2. Positiveness of content 32 7 intervals 3. Negativeness of content 31 7 intervals 4. Retention of positive content 30 True/false 5. Retention of negative content 27 True/false 6. Stricken testinony 28 True/false 7. Nullification 29 True/false 8. Importance of judge's instructions 25 7 intervals 44 FOOTNOTES 1Further information concerning the data not presented in this work can be obtained from Dr. Edward L. Fink, Depart- ment of Communication, Michigan State University, East Lansing, Michigan, 48824. 2A Total of 269 85 answered the questionnaires. Five of them were dropped because more than 5 questions were left unanswered. Four additional subjects were randomly dropped in order to equalize the cell sizes. 3The collaboration of Ms. Nancy Richardson is highly appreciated. 4It was felt that adding this paragraph would encourage honesty and participation of the subjects, and that this advantage outweighed the possible danger of a sensitizing effect. CHAPTER III RESULTS In this chapter we will present the results obtained in the study. Unless indicated otherwise, the data analyzed correspond to the six-cell design presented in Table 2.2, and to the four—cell design which can be constructed by dropping from the analysis the two control groups in the neutral avowal condition. Means and standard deviations of all the variables presented in this chapter can be found in Appendix D. The coding of the variables will be indicated as necessary. The data analytic techniques employed to test the hypotheses are product moment correlations, analyses of variance, and canonical correlations. The standard level of significance of p i .05 is adopted. The presentation of results will be organized as follows. First, possible biases in the assignment of sub- jects to the experimental groups will be examined. Second, Ss' background characteristics as possible contaminating in— fluences on the main dependent variable will be discussed. Third, attention will be focused on the Ss' perceptions of (a) the experimental variables, and (b) the intervening vari— ables. Finally, data concerning each of the three sets of 45 46 hypotheses previously advanced will be presented. Equivalency of Subjects in the Experimental Groups In order to investigate possible biases in the random procedure followed in assigning Ss to treatment groups, the correlations between seven subject background variables and each of the independent variables are calculated. Such cor- relations for the total sample (N=26O)l in the l3—cell design are presented in Table 3.1. Table 3.1. Correlations Between Subjects' Background Vari- ables and the Independent Variables.a Disclosure Source Avowal Type (Linear) Sex -.053 -.029 Age —.O30 .083 Family size .008 .089 Birth order -.054 .034 Family economic status .036 -.O39 Father's education -.015 .018 Mother's education -.039 -.002 aMissing cases omitted; 257 i N < 258 for all correlations. Analysis of this table reveals no significant2 cor- relations between the Ss' sex, age, family size, birth order, family economic status, and parents' education with the in- dependent variables. In fact, the average intercorrelation 47 is only .038, and the highest correlation in the table is .089. Thus, we conclude that there is little reason to sus- pect any subject bias in the overall study. Influence of Subjects' Background Characteristics on Attribution of Responsibility Given that AR is the main dependent variable, special care is taken in probing various possible influences on it. Previous work in the area (e.g., Stephan, 1973) has pointed out some significant associations between AR and variables such as sex and age. Accordingly, simple correlations be- tween subject background measures and various indicators of AR are calculated. These correlations are presented in Table 3.2 which shows weak relationships for the most part. None of the cor- relations concerning birth order, family economic status, and father's education are statistically significant.3 Furthermore, mother's education correlates significantly only with the indicator of sentence (r=—.l37; N=200; p i .05) and, given the low magnitude of the other correlations of this identification variable, such finding could be expected by chance alone. A similar situation occurs with the variable age, which is significantly related only to the not guilty verdict (r=-.168; N=200; p i .05). A stronger pattern of results is present in the var- iable family size which correlates significantly with both attributed responsibility (r=—.136; N=200; p i .05) and degree of'guilt(r=.l46; N=200; p i .05). Both correlations indicate 48 . wCOHu.MHm.H.H8 .Ho. w.m um pcMOHMHcmwmo .mo..w a pm pcmoauacmamn Ham sou mmm w.z.w smm “Banshee mummo maummaza nsma.- Mao.- aoo.- omo. mac. moo. ammc.n mocmucmm omo. mmo.- moo. mmo.u mHo.- anH.- mac. museum uoz sao.- 400.- ass. was. mac. was. mmo. umpaasmamcmz omo.u «so. mmo.- mac. Gso.- sac. omma.n swans: mwummo cam who. Hos. mao.- Has. nsva.u moa. mmo. pause «Ho. mmo.u moo.- mmo.- NNH.- mmo. peas. swarms spaaanamcoammm omo.- mac. moo.- oao.u smo.u omo. oso. umumam suaaanamcoawmm wmo.- mac. amo.- ooa. nGMH.- mmo. mmo.- ucwacmmmo suHHHnHmcoammm sao.- Hmo.- omo. omo.- was. one. cams. nauseous coflumosom coaumonom mmm nacho cmflm mom Rom Hugues. magnum auuam assess mo COfiuanHuue Mo muoumoaocH one moacwflHm>_pcsonxomm .muommnsm comzumm chflucHwHHoo m.§:aa&§ .N.m OHQMB 49 that the greater the family size of the subject, the more leniency toward the defendant. These findings are unexpect- ed from the lack of previous research, and are difficult to explain because of the low magnitude and contradictory direction of the remaining correlations. It should also be noted that even the two significant correlations explain a very low percentage of variance (about 2%). The strongest relationship shown in Table 3.2 is that between sex of the subject and various indicators of AR. The correlations between sex and both accident (r=.l96) and second degree murder verdict (r=-.182) are significant at p i .01; the correlations with sentence (r=—.155) and responsibility of father (r=.l47)are significant at p i .05. These results show a tendency for women to be more tolerant toward the defendant and less tolerant toward the defen- dant's father than males. Such pattern of results is in accordance with findings reported in the literature (e.g., Stephan, 1973). Since these correlations only account for a 3% to 4% of the variance, and given that Ss were randomly assigned to experimental groups regardless of their sex (see previous section), we conclude that there is little chance of a significant influence of subjects' personal characteristics on AR. Subjects' Perceptions of the Experimental Variables In order to find out if the Ss perceived and inter— preted the manipulations of independent variables in the way 50 it was planned, analyses of variance are performed on the following variables: (1) perceived positiveness of content, (2) perceived negativeness of content, (3) perceived ingra- tiation, and (4) perceived openness. These analyses are presented in Tables 3.3 through 3.6 which are discussed below. Perceived Positiveness of Content - Along with other measures in the questionnaire, 83 were asked to state "How good do you think it is to be a member of a Boy Scout TrOOp and participate in a neighborhood clean-up campaign?" This question was answered on a seven point scale ranging from "not at all good" (one) to "very good" (seven). The grand mean (5.3) in Table 3.3 shows an overall tendency of the 88 to perceive the content of the messages as fairly positive. Furthermore, it is clear that the source of disclosure does not have an appreciable effect on the perceived positiveness of the messages, and this is reflected in the total means of 5.3 for the self disclosure (SD) and 5.2 for the other disclosure (OD) conditions. However, sig- nificant differences due to avowal type are found. For the neutral avowal condition (NoA) the total mean is only 4.7, while for the negative avowal (NA) condition the mean is 5.3, and for the positive avowal (PA) condition it is even higher at 5.8. These differences result in a significant (F=7.l3; df=2/102; p-: .001) main effect for avowal type in the six-cell design. This effect is also present when the control groups are removed from the analysis in the four-cell Table 3.3. 51 Means and Analyses of Variance for Perceived Positiveness of Content, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Total Negative 5.1 5.3 Positive 6.1 5.8 Neutral 4.9 4.7 Total 5.3 5.3 Analysis of Variance Six—Cell Design (N=108): Source of Variance df MS F p Disclosure Source (A) l .45 .28 .600 Avowal (B) 2 11.68 7.13 .001 A x B 2 2.23 1.36 .260 Residual Error 102 1.64 Total 107 Four—Cell Design (N=72; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) l 0.00 0.00 .00 Avowal (B) l 5.56 4.26 .04 A x B l 3.56 2.73 .10 Residual Error 68 1.30 Total 71 a I O I o o I Higher means indicate greater perceived pOSitiveness of content . N=18 per cell. 52 design (F=4.26;df=l/68; p i .04). Thus, we conclude that the messages designed to pre- sent the defendant in a positive manner were perceived by the Ss as generally positive. Furthermore, such messages are rated most positively by those 85 who had been exposed to them, less positively by Ss who had heard negative messages about the defendant; and least positively by Ss in the control groups. Perceived Negativeness of Content — A similar proced- ure to the one outlined above for positive messages is fol- lowed regarding negative messages about the defendant. SS were asked "How bad do you think it is to be a member of a street gang and to rob a store?" This question was answered on a scale ranging from one, "not at all bad" to seven, "very bad." The general perception of such actions as negative is reflected in the grand mean (5.3) shown in Table 3.4, and also in most of the other means presented in the same table which show only small differences among them. The notable exception is the mean corresponding to the SD-NA cell (4.8), which is much lower than the others, and therefore produces identical marginally significant main effects (F=3.80; df= 1/68; p i .055) for both disclosure source and avowal type, when the control groups are left out of the analysis. How- ever, the effect of this low mean is lessened by the uni- formity of other means and the within-cell variances when 53 Table 3.4. Means and Analyses of Variance for Perceived Negativeness of Content, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 4.8 5.8 5.3 Positive 5.8 6.0 5.9 Neutral 5.8 5.6 5.7 Total 5.5 5.8 5.6 Analysis of Variance Six-Cell Design (N=108): Source of Variance df MS F p Disclosure Source (A) l 2.37 1.53 .220 Avowal (B) 2 3.12 2.01 .139 A x B 2 3.79 2.44 .092 Residual Error 102 1.55 Total ’ 107 Four—Cell Design (N=72; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) l 6.13 3.80 .055 Avowal (B) l 6.13 3.80 .055 A x B l 3.13 1.94 .168 Residual Error 68 1.61 Total 71 aHigher means indicate greater perceived negativeness of content. N=18 per cell. 54 the control groups are included in the six-cell design. Here, no significant effects are observed. We conclude that the messages designed to present the defendant in a negative manner are indeed perceived as negative by Ss in all conditions, except for those 85 who had heard the defendant disclose such negative messages be- fore. These Ss View such information as less negative than their counterparts in the other experimental conditions. Perceived Ingratiation — One of the hypotheses in the study is that when a defendant presents himself in a posi- tive manner, this communication will be construed as ingra— tiating, and a negative attitude toward him will ensue. Furthermore, it is anticipated that the same positive mes- sages, when presented by a person other than the defendant, will not be construed as ingatiation and, therefore, no negative feelings will be produced. In order to check whether or not perceived ingratia- tion was produced by the experimental manipulations, 88 were asked "To what extent did Johnny's testimony impress you as an attempt to present himself in a favorable way?" The scale used by the 85 to answer this question ranged from zero to 100 percent, where higher percentages indicate greater per- ceived ingratiation. These results are presented in Table 3.5. Taking the total mean for the NoA condition (44.5) as a baseline for comparison, it can be observed that the total mean for the 55 Table 3.5. Means and Analyses of Variance for Perceived Ingratiationé by Disclosure Source and by Avowal Type. Disclosure Source Avowal Type Self Other Total Negative 39. 44.6 .0 Positive 62. 66.1 64.4 Neutral 41. 48.1 .5 Total 47. 52.9 .3 Analysis of Variance Six-Cell Design (N=108): Source of Variance df MS F p Disclosure Source (A) 1 725.93 .73 .396 Avowal (B) 2 5432.58 5.43 .006 A x B 2 31.18 .03 .969 Residual Error 102 1000.22 Total 107 Four-Cell Design (N=72 ; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) 1 325.13 .29 .590 Avowal (B) 1 9045.13 8.17 .006 A x B 1 15.13 .01 .907 Residual Error 68 1106.61 Total 71 a . . . . . . . Higher means indicate greater perceived ingratiation. N=18 per cell. 56 NA condition (42) is somewhat lower, and the total mean for the PA condition is much higher (64.4), resulting in a sig— nificant main effect for avowal type for both the six (F=5.43; df=2/102; p i .006) and four (F=8.l7; df=l/68; p i .006) cell designs. A more detailed analysis of the means in the positive avowal condition shows that the defendant's SD does produce a moderately high degree of perceived ingratiation (62.8). However, contrary to expectations, such perception of ingratiation increases to 66.1 when the source of the positive information was a person REESE than the defendant. Thus, we conclude that, in an absolute sense, the experimental manipulations did not produce a powerful sense of ingratiation in the 85, even if the differences between experimental groups are significant. Furthermore, whatever the perceived ingratiation produced, it was mainly due to the positive nature of the messages' content, and the source of disclosure did not have an overriding effect which would have been reflected in a significant interaction. Perceived Openness - The same way that the interplay of SD and positive avowal is thought to result in perceived ingratiation, the interaction between SD and negative avowal is hypothesized to result in perceived openness. Table 3.6 shows the means and analysis of variance for perceived openness as measured by the question "How open do you think Johnny has been in this trial?", which was answered on a scale ranging from one ("not at all open") to seven ("very 57 Table 3.6. Means and Analyses of Variance for Perceived Openness, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 5.1 4.2 4.6 Positive 4.3 4.4 4.3 Neutral 3.8 4.3 4.1 Total 4.4 4.3 4.3 Analysis of Variance Six—Cell Design (N=108): Source of Variance df MS F p Disclosure Source (A) 1 .15 .08 .783 Avowal (B) 2 2.78 1.43 .245 A x B 2 4.93 2.53 .084 Residual Error 102 1.95 Total 107 Four—Cell Design (N=72 ; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) 1 2.72 1.63 .206 Avowal (B) l 1.39 .83 .365 A x B l 4.50. 2.70 .105 Residual Error 68 1.67 Total 71 aHigher means indicate greater perceived openness. N=18 per cell. 58 much open"). Generally, the means obtained center around 4.3 or the middle of the scale, which seems to be the base- line of comparison from which there are two notable depar- tures. First, in the SD—NA condition a great increase of perceived openness can be observed (i=5.l), a result which conforms to previous expectations. Second, the SD-NoA con- dition shows a decrease of the mean to 3.8. This reduction is also logical because of the very nature of the experi- mental condition which presents the defendant with a minimal amount of disclosure. In summary, in this section we have presented the results concerning the Ss' perceptions of the experimental variables. These results suggest that Ss perceived as fairly positive those messages designed to present the defendant in a positive manner, and as negative those messages designed to present the defendant under a negative light. Further— more, a moderately high degree of perceived ingratiation was observed under conditions of positive avowal and regardless of the source of disclosure. Finally, under conditions of negative avowal, greater openness of the defendant was per— ceived in the self disclosure-negative avowal condition than in any of the other experimental conditions. Subjects' Perceptions of the Intervening Variables Statistical procedures similar to those discussed above are applied to data regarding the intervening variables posited by our theoretical rationales. Such analyses, pre— sented in Tables 3.7 to 3.10, are discussed below. Table 3.7. 59 Means and Analyses of Variance for Perceived Honesty, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 68.3 54.0 61.1 Positive 67.1 58.5 62.8 Neutral 67.3 54.2 60.7 Total 67.5 55.5 61.5 Analysis of Variance Six—Cell Design (N=120): Source of Variance df MS F p Disclosure Source (A) 1 4308.01 7.22 .008 Avowal (B) 2 48.48 .08 .922 A x B 2 92.01 .15 .857 Residual Error 114 596.69 Total 119 Four—Cell Design (N=80; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) 1 2610.61 4.43 .039 Avowal (B) 1 56.11 .10 .758 A x B 1 165.31 .28 .598 Residual Error 76 589.00 Total 79 aHigher means indicate greater perceived honesty. N=20 per cell. 6O Perceived Honesty - Perceived honesty was measured on a zero to 100 percent scale of response to the question "To what extent do you perceive Johnny as being an honest person?" Greater percentages indicate greater perceived honesty. In Table 3.7 it can be seen that while the different manipulations of avowal type show virtually no differences between them, the manipulation of disclosure source has a powerful impact which is reflected in significant findings for both the six—cell (F=7.22; df=l/1l4; p i .008) and the four—cell (F=4.43; df=l/76; p i .039) designs. This pattern of results leads us to conclude that the Ss seem to interpret honesty simply as the defendant's verbal participation in the trial. Further evidence for ,this conclusion lies in the 13 point difference between the means for the control groups in the neutral condition. In- deed, even when the defendant presents himself in a neutral fashion, he is still rated as being moderately honest (67.3); and this rating contrasts sharply with the rating given when the same neutral information is provided by another indi— vidual (54.2). Perceived Similarity — The variable of perceived sim— ilarity was measured by the question "How similar do you feel to Johnny?" 85 responded to this question using a seven point scale ranging from "not at all" (one) to "very much" (seven). 61 The grand mean (i=2.7) in Table 3.8 clearly shows the low degree of identification with the defendant felt by the 85, a result which can be expected given the fact that the defendant is being tried for murdering his father. However, even within this low level of perceived similarity, a sign- nificant interaction effect is found in both the six (F=3.34; df=2/ll4; p i .043) and the four (F=6.20; df=l/76; p i .015) cell designs. These results show that 55 perceive themselves as being more similar to the defendant when he discloses either negative or neutral information about self, and when ehheh disclosed positive information about the defendant. Positive and Negative Labeling - In order to find out the effects that the experimental manipulations had upon the stereotyping process of the 85 toward the defendant, 85 were asked: "To what extent do you agree with the following statement? Johnny's testimony increased my general opinion of him“; the response was recorded on a scale ranging from one for "strongly disagree" to five for "strongly agree." The question relating to negative evaluation was "To what extent did Johnny's testimony impress you as information EE' vealing his criminal disposition?" The scale used for this measurement was from zero to 100% where greater percentages of response indicate more negative evaluations. The results relating to these questions are presented in Tables 3.9 and 3.10. Table 3.9 shows that source of dis- closure has only a slight effect on the juror's positive 62 Table 3.8. Means and Analyses of Variance for Perceived Similarity, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 3.3 1.9 .6 Positive 2.3 3.1 .7 Neutral 3.2 2.8 .6 Total 2.9 2.5 .7 Analysis of Variance Six-Cell Design (N=120): Source of Variance df MS F p Disclosure Source (A) l 6.53 1.74 .190 Avowal (B) 2 .31 .08 .921 A x B 2 12.16 3.23 .043 Residual Error 114 3.76 Total 119 Four-Cell Design (N=80; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) 1 2.11 .57 .454 Avowal (B) l .11 .03 .863 A x B l 23.11 6.20 .015 Residual Error 76 3.73 Total 79 aHigher means indicate greater perceived similarity. N=20 per cell. 63 Table 3.9. Means and Analyses of Variance for Positive Evaluation, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 3.1 3.4 .3 Positive 3.1 3.4 .3 Neutral 2.9 2.7 .8 Total 3.0 3.2 .1 Analysis of Variance Six-Cell Design (N=108): Source of Variance df MS F p Disclosure Source (A) 1 .75 .95 .333 Avowal (B) 2 2 37 2.99 .055 A x B 2 .78 .98 .378 Residual Error 102 .79 Total 107 Four-Cell Design (N=72; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) 1 2,00 2 87 .095 Avowal (B) l O 00 O 00 1.000 A X B 1 .06 .08 .780 Residual Error 68 70 Total 71 a I 0 ' c I 0 Higher means indicate greater pOSitive evaluation. N=18 per cell. 64 Table 3.10 Means and Analyses of Variance for Negative Evaluation, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 42.2 31.2 36.7 Positive 30.0 30.8 30.4 Neutral 27.7 27.8 27.8 Total 33.3 29.9 31.6 Analysis of Variance Six—Cell Design (N=108): Source of Variance df MS F p Disclosure Source (A) 1 306.70 .34 .562 Avowal (B) 2 764.23 .84 .433 A x B 2 394.29 .44 .648 Residual Error 102 905.78 Total 107 Four—Cell Design (N=72; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) 1 465.13 .45 .507 Avowal (B) 1 715.68 .68 .411 A X B 1 630.13 .60 .440 Residual Error 68 1044.86 Total 71 aHigher means indicate greater negative evaluation. N=18 per cell. 65 evaluations, the general trend being a more positive evalua- tion of the defendant under conditions of other—disclosure. The avowal type manipulation has a marginally significant effect only when the control groups are included in the analysis (F=2.99; df=2/102; p i .055). The fact that this main effect totally disappears when the control groups are taken out of the analysis points to the conclusion that ehy information about the defendant, regardless of its nature, leads to an increased positive attitude toward the defendant. Table 3.10 presents the results regarding negative attitudes toward the defendant. It shows no statistically significant main effects or interactions. However, it is important to note the strong negative evaluation of the de- fendant under conditions of SD—NA (42.2, as compared with a grand mean of 31.6). This result, which is congruent with theoretical expectations, does not produce any significant effects because of high within-cell variability. Finally, it should be noted that the correlation be- tween the two variables discussed in this section is r=.21 (see Table E.3), a theoretically inconsistent result which will be discussed in Chapter IV. Sympathy - In an attempt to find out the effects of the experimental variables on the sympathy felt by the 85 to- ward the defendant, Ss were asked: "To what extent do you agree with the following statement? Johnny's testimony made me feel sympathetic towards him." The scale of response had 66 five points ranging from one for "strongly disagree" to five for "strongly agree." Table 3.11 reports the analyses performed on this variable. It shows no significant effects produced by the manipulation of experimental variables. It can be seen that disclosure source is, in this case, irrelevant as a predic— tive variable of sympathy. Results for avowal type show only a minimal amount of variance between groups in the direction of a slight amount of sympathy when positive content is dis- closed and little sympathy under conditions of negative dis— closure. In summary, in this section we have presented the results concerning the effects of the experimental manipula- tions on the intervening variables. It was found that per- ceived honesty is highly influenced by disclosure source but not by avowal type. A low level of perceived similarity re- sults from the self-disclosure of negative actions or from the other—disclosure of positive actions. Thirdly, while positive attitude toward the defendant is fostered by any kind of information (positive or negative) about him, neg- ative labeling is produced only when the defendant avows deviant behaviors. Finally, none of the experimental vari- ables has a significant effect on sympathy toward the defendant. We will now turn to the presentation of the results concerning each of the three sets of formal hypotheses 67 Table 3.11. Means and Analyses of Variance for Smypathy, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 3.2 3.1 3.1 Positive 3.4 3.6 3.5 Neutral 3.3 3.2 3.3 Total 3.3 3.3 3.3 Analysis of Variance Six-Cell Design (N=108): Source of Variance df MS F p Disclosure Source (A) 1 .00 .00 1.000 Avowal (B) 2 1.01 .95 .389 A x B 2 .19 .18 .833 Residual Error 102 1.06 Total 107 Four-Cell Design (N=72; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) l .06 .05 .823 Avowal (B) 1 2.00 1.81 .183 A x B 1 .22 . .20 .655 Residual Error 68 1.10 Total 71 aHigher means indicate greater sympathy. N=18 per cell. 68 advanced in Chapter I. Three statistical procedures will be used in the analysis of the data: simple product-moment cor— relations, canonical correlations, and analysis of variance. Of these, only the canonical analysis merits some discussion in advance of the presentation of the results. Canonical Correlation Analysis Canonical correlation analysis is a statistical tech— nique through which it is possible to assess the nature of a relationship between two sets of variables, where each of the sets has theoretical meaning as such, and may be char- acterized by more than one underlying dimension. This tech— nique can be viewed as combining principles of both factor analysis and multiple regression. Its goal is to find a linear combination of variables in each set, for which the correlation between the two composites (i.e., canonical var- iates) is maximum. If no significnat linear association be- tween the two sets of variables exist, no significant canonical variates can be found. On the other hand, it is possible to find several significant sets of canonical var- iates. The correlation between two corresponding canonical variates from different sets is called the canonical correla— tion, and its meaning is analogous to a simple product- moment correlation. The significance of a canonical correla- tion is tested through a x2 test statistic which takes into account sample size so that the same canonical correlation is 69 more significant as the sample size increases. The assumptions of canonical correlation statistical analysis are essentially those underlying simple correlation: (a) normality of distributions, (b) homogeneity of variance, (c) independence of observations, and (d) linear association between variables. Given the difficulty of interpreting the results yielded by this technique, in the presentation of our results we will examine only the firt set of canonical variates, which is appropriate given that our main interest is the optimal prediction of one set of variables by another. Furthermore, in an effort to clarify the meaning of our findings we will follow the procedures outlined by Van de Geer (1971:156-170) for interpreting canonical results in the form of causal models. Thus, Figure 3.1 represents, in a summary form, the general causal model hypothesized in Chapter I. This model includes the different variable sets to be entered in the canonical analysis (error terms have been left out for simplicity). Now, given that canonical corre- lation can only handle two sets of variables at a time, Figure 3.1 also shows six possible pair-wise comparisons among the sets.4 Of these comparisons, (l), (2), and (3) correspond exactly to Hypotheses one, two and three respec- tively. In the discussion below, each of these comparisons will be presented as an isolated causal model so that the manipulation of the data in each of the cases will be clear. 70 (2) _: : - ” <3) " ’"””3‘ ~ \ '/’ I” (l) ‘ \‘s. l. -‘v i «0“ ‘\‘\\\ 5 i - ' ‘ \x \ I ‘ 1 .- _ \ x _ lbqerhmxmal Zhuerwafing Lfldng | AR ! variables variables Indicators I Indicators f o . . 1‘1 1, \. I" .\.‘ \ (4) ’. \_\ (5) {/1 ‘s k I - I I .. ~-\ \ ' - _.- / \ ’!;/ (6) Figure 3.1. Summary Version of the General Causal Model of the Study and Pair—wise Comparisons Between Sets of Variables. Hypothesis 1 The greater the liking of a juror for a defendant, the less the responsibility that will be attributed to the defendant. In order to test this hypothesis, simple and canon- ical correlations between different indicators of liking and AR are calculated. Table 3.12 presents product-moment correlations between all of the indicators of these vari— ables. Because of the greater stability accruing to a greater sample size, such correlations are based on the re- sponses of the total sample for the study (N=26O).5 A relevant result, calculated from this table, is the average intercorrelation between the five indicators of 71 .mcoflflmhoo in non 8N .v. z w mmm “ooppfleo mommo mcammdzc omm. omm. mow. mow. mma. mmo.| mmo.l moa.| vao. meo.l NoH.| HHH. HHN.I Awaxv mcflxmq HMGHHCH ooo.H moo. one. mmv. oom. wmo.u omH.I omm.| moo. mmo. mma.| mHH. mmm.| Amaxv mcaxaq Baucmuom ooo.H How. ham. mmm. mmo.| oHN.I mmm.l oma. mmo. vmm.| ooa. NHN.I Amaxv soapmnaeom ooo.H omw. vmm. Hwo.l nvm.l omm.u ova. mmo.| vam.| mom. mom.l AHHxV mmmcwmoHU ooo.H mma. moo. oma.| mmo.| mmo. voo.s mmH.| mmo. mma.| Aoaxv com: on chance ooo.H Hmo.u onH.| mmv.l Hem. ooo.| omm.| moH. omm.l Amxv wpaflso uoz ooo.a Nmo.l omo. nmo. omo. moo. mva. oma.| Amxv Hmpcmsmamcmz ooo.H omm. me.t ooa.| omm. omm.l who. Ahxv moons: ooo.H omm.| ovo.l one. mnm.| oom. onv Haasw ooo.a va. mmm.l mmm. How. Amxv Monumm mUHHHchcommmm ooo.H BAH. mmo. who. wav Hmumam suuaafimaommmm ooo.H mom.| on. Amxv uccpcmmmo humaflcflmcommmm ooo.H mv.ml Amxv pcmofloom coo .H Awe 888m max max flax oax mx ox bx ox mx ox mx mx ax mdflfiw 86.8. 93 now. muflaacflmcommmm mo coausnaunu< mo muonmoflocH com mcflxaq mo mHOHMOHocH cmmzumm chHumHmHHou .Na.m canoe 72 liking (Xlo to ) which is r=.442. Such correlation points X14 toward the underlying theoretical meaning of this set of in- dicators, but also points out some degree of independence be- tween the indicators which were designed to tap different dimensions of the general concept of liking. It should also be noted that eh; these correlations are positive and signif- icantly different from zero at p i .01. Data presented in Table 3.12 show strong support of Hypothesis 1. All five correlations between sentence and indicators of liking are negative (as expected); four of them are significant5 at p i .01, and the fifth is significant at p i .05. A similar pattern of results is observed in the correlations between responsibility of defendant and liking, where three of them are significant at p i .01, and the re- maining two are significant at p i .05. Furthermore, all five correlations between liking and guilt are in the right direction, and three of them reach significance at p i .01. Finally, except for one, all 13 correlations between liking and the three alternative verdicts (second degree murder, manslaughter and not guilty) are in the right direction, and six of them are significant at p i .01. These correlations, however, should be considered very cautiously since responses to the three alternative verdicts are obviously interdepend- ent. The process of discussing several intercorrelations separately is cumbersome at best and deceptive at worst. Canonical correlation is a statistical technique especially 73 suited to overcome this problem. Accordingly, Table 3.13 presents the canonical weights and canonical correlations between indicators of liking and indicators of AR for the four and six-cell designs. The product-moment correlations from which thistatdeais computed are presented in Tables E.l and E.2 in Appendix E. Results for the four—cell design shown in Table 3.13 indicate a canonical correlation of r=.646 which accounts for 42% of the variance and results in x2 of 67.44, which is significant at the .004 level. Similar results are obtained for the six-cell design which yields a canonical correlation of r=.572, accounting for 33% of the variance. The chi square value in this analysis is equal to 72.88, significant at p = .001. The comparison of the four and six-cell designs allows us to study the stability of the correlations. When these coefficients do not change a great deal, this is reason to feel more confident that they are good estimators of the true parameters of the population, however, in order to be conservative, the smaller coefficient should always be taken as the better estimate.6 Hypothesis 2 There is an interaction effect between disclosure source and avowal type such that, (a) Under conditions of negative avowal, less responsibility will be attributed to a defendant if deviance is disclosed by the defendant himself rather than a witness. 74 Table 3.13. Canonical weights and Canonical Correlations Between Indicators of Liking and.Indicators of.Attribution of Responsibility for Four Cell and SixeCell Designs.a Eigenvalue Degrees Canonical (variance 2 of Signif- Design Correlation explained) x Freedom icance Four cells (Ne72) .646 .418 67.44 40 .004 Six Cells (N=108) .572 .327 72.88 40 .001 Canonical weights Four Cells Six Cells Indicators of liking (N=72) (N=108) Desire to Meet -.377 -.531 Closeness .399 .503 Adndration .148 .191 Potential Liking .825 .850 Initial Liking .142 -.148 Indicators of Attribution of Responsibility, Sentence -.286 -.124 .Accident .278 —.162 Responsibility Defendant -.123 -.042 Responsibility Sister .175 .003 Responsibility Father .132 —.006 Guilt -.434 -.516 .Murder'verdict .022 -.613 .Manslaughter'verdict -.438 -.615 ; aF'irst canonical variate only. 75 (b) Under conditions of positive avowal, less responsibility will be attributed to a defendant if the positive information is disclosed by a witness rather than the defendant himself. In order to test such interaction effect, analyses of variance are performed on five different indicators of AR. These indicators were: responsibility of defendant, attri- bution of guilt, verdicts of murder and not guilty, and sentence. The results of such analyses are presented in Tables 3.14 to 3.18 respectively. Note that these tables include the results obtained for both the four and the six— cell designs. The latter have been included so that the reader may have as complete information as possible. However, the discussion below will concentrate only on the results for the four-cell design which bear directly on the hypoth- eses presented above. Table 3.14 displays the results obtained for the at- tribution of responsibility variable. To measure this vari- able, the Ss were given a scale of zero to 100% in which to answer the question "How reeponsible do you think Johnny is for the death of his father?", where greater percentages represent increased attribution of responsibility. It will be seen that the manipulation of the experimental variables has essentially no effect on this indicator of AR. Such absence of effects is reflected in both the uniformity of mean values and the extremely low F ratios presented in this table. 76 Table 3.14. Means and Analyses of Variance for Responsi- bility, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 56.3 55.5 55.9 Positive 55.5 58.7 57.1 Neutral 56.8 51.5 54.5 Total 56.2 55.2 55.8 Analysis of Variance Six—Cell Design (N=120): Source of Variance df MS F p Disclosure Source (A) 1 27.08 .03 .874 Avowal (B) 2 89.26 .08 .920 A x B 2 178.68 .17 .846 Residual Error 114 1069.53 Total 119 Four-Cell Design (N=80; Excluding Neutral Avowal Condition): Source of Variance df . MS F p Disclosure Source (A) l 28.80 .03 .872 Avowal (B) l 31.25 .03 .867 A x B 1 80.00 .07 .789 Residual Error 76 1105.37 Total 79 aHigher means indicate greater attribution of responsibility. N=20 per cell. 77 Using the same type of scale as in the previous ques- tion, 85 were asked "To what extent do you think Johnny is legally guilty of killing his father?" The results of the analysis are presented in Table 3.15 which shows that the expected interaction effect failed to materialize. Instead, there was a main effect for avowal type which reached an F=2.66, significant at p 3.10; df=l/76. The effect consisted in greater attribution of guilt under conditions of positive avowal, and lesser attribution when negative avowal occurred. Subjects were also asked to render a verdict which was to be chosen from three alternatives: second degree murder, manslaughter, or not guilty. Since each of the alternative verdicts is analyzed independently, knowledge of the results obtained for any two verdicts would automatically determine the outcome of the third. Thus, Tables 3.16 and 3.17 present the results concerning the second degree murder and the not guilty verdicts. It should be noted that these two tables are statistically interdependent. Furthermore, since the dependent variable is dichotomous, significance levels should be evaluated cautiously. Means in the table of second degree murder show that Ss have a slight tendency to be more lenient when negative avowal has occurred, and that the source of disclosure has negligible effects. On the other hand, means in the table of not guilty verdict (Table 3.17) show the presence of the expected interaction. Indeed while 30% of the 88 who chose Table 3.15. 78 Means and Analyses of Variance for Guilt, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 56.7 51. 53.9 Positive 70.7 63. 67.2 Neutral 70.6 51. 61.2 Total 66.0 55. 60.8 Analysis of Variance Six-Cell Design (N=120): Source of Variance df MS F p Disclosure Source (A) 1 3276.08 2.49 .117 Avowal (B) 2 1767.26 1.34 .265 A x B 2 527.48 .40 .670 Residual Error 114 1314.21 Total 119 Four-Cell Design (N=80; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) 1 787.51 .59 .443 Avowal (B) 1 3524.51 2.66 .107 A x B 1 9.11 .01 .934 Residual Error 76 1325.67 Total 79 aHigher means indicate greater attribution of guilt. N=20 per cell. 79 Table 3.16. Means and Analyses of Variance for Second Degree Murder Verdict, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 10 10 10 Positive 15 20 18 Neutral 25 05 15 Total 17 12 14 Analysis of Variance Six-Cell Design (N=120): Source of Variance df MS F pb Disclosure Source (A) 1 .08 .61 .437 Avowal (B) 2 .06 .47 .624 A x B 2 .18 1.42 .246 Residual Error 114 .12 Total 119 Four—Cell Design (N=80; Excluding Neutral Avowal .Condition): Source of Variance df MS F pb Disclosure Source (A) 1 .01 .10 .751 Avowal (B) 1 .11 .91 .342 A x B 1 .01 .10 .751 Residual Error 76 .12 Total 79 aMeans indicate the percentage of subjects who rendered this verdict. N=20 per cell. bSince dependent variable is dichotomous, alpha levels may not be exact. 80 Table 3.17. Means and Analyses of Variance for Not Guilty Verdict, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 3O 20 25 Positive 05 3O 18 Neutral 15 15 15 Total 17 23 19 Analysis of Variance Six-Cell Design (N=120): Source of Variance df MS F pb Disclosure Source (A) 1 .08 .48 .488 Avowal (B) 2 .11 .70 .499 A x B 2 .33 2.10 .127 Residual Error 114 .15 Total 119 Four—Cell Design (N=80; Excluding Neutral Avowal Condition): Source of Variance df MS F pb Disclosure Source (A) 1 .11 .68 .412 Avowal (B) l .11 .68 .412 A x B l .61 3.71 .058 Residual Error 76 .17 Total 79 aMeans indicate the percentage of subjects who rendered this verdict. N=20 per cell. bSince dependent variable is dichotomous, alpha levels may not be exact. 81 this alternative verdict were in the SD-NA and OD-PA groups, 20% of these 85 were in the OD—NA condition and only 5% were in the SD—PA group. This interaction results in an F=3.7l which is marginally significant at p=.058 with df=l/76. Finally, results shown in Table 3.18 again weakly reflect the predicted interaction. While the differences between the means do not reach statistical significance, it is important to note that such means represent suggested number of years in prison for the defendant; therefore, to the extent that these results may be generalizable to an actual court setting, differences of one or two years may have tremendous social importance. From this discussion we can conclude that the evidence supports Hypotheses 2a and 2b only partially. This conclu- sion is strengthened by the evidence yielded by a canonical analysis which used the independent variables and their interaction as one set, and the indicators of AR as another set. This evidence is presented in Table 3.19. The can- onical correlation obtained was .439 which explains 19% of the variance. The corresponding x2=33.59 is significant at p i .09. Also important is the finding that source of dis- closure is the most powerful predictive variable (weight is .91). Hypothesis 3 Stemming from the theoretical model presented in Chapter I, the following set of hypotheses was advanced: 82 Table 3.18. Means and Analyses of Variance for Sentence, by Disclosure Source and by Avowal Type.a Disclosure Source Avowal Type Self Other Total Negative 5.6 7.3 6.5 Positive 5.8 4.8 5.3 Neutral 4.0 5.4 4.7 Total 5.1 5.8 5.5 Analysis of Variance Six-Cell Design (N=120): Source of Variance df MS F p Disclosure Source (A) 1 16.13 .41 .523 Avowal (B) 2 30.86 .79 .459 A x B 2 21.56 .55 .579 Residual Error 114 39.31 Total 119 Four-Cell Design (N=80; Excluding Neutral Avowal Condition): Source of Variance df MS F p Disclosure Source (A) 1 3.20 .07 .798 Avowal (B) l 26.45 .55 .462 A x B l 36.45 .75 .388 Residual Error 76 48.42 Total 79 a O O I 0 Means indicate suggested number of years in prison. N=20 per cell. 83 Table 3.19. Canonical Weights and Canonical Correlation Between Experimental Independent Variables and Indicators of Attribution of Responsibil- ity.a Eigenvalue Canonical (Variance 2 Degrees of Correlation Explained) x Freedom Significance .439 .193 33.59 24 .092 Independent Variables Canonical Weight Source of Disclosure (A) .908 Avowal Type (B) .203 A x B -.367 Indicators of Attribution of Responsibility Sentence -.553 Accident -.517 Responsibility Defendant —.464 Responsibility Sister .070 Responsibility Father -.797 Guilt .063 Murder Verdict .386 Manslaughter Verdict .399 aFour-cell design only (N=72). First canonical variate only. 84 Hypothesis 3a: The greater the perceived simi— larity of a juror with the de- fendant, the lesser the AR. 3b: The greater the positive labeling of a defendant, the lesser the AR; and the greater the negative label- ing, the greater the AR. 3c: The greater the sympathy toward a defendant the lesser the AR. 3d: The greater the perceived honesty of the defendant, the lesser the AR. The validity of this set of hypotheses is analyzed through the use of correlational techniques similar to those presented in the previous section. Correlations between these variables (identified as intervening variables in the theoretical model) and indicators of AR for the total sample (N=260) are presented in Table 3.20. Support for Hypothesis 3a is found in the first column of Table 3.20. Indeed, data confirm that perceived simi- larity results in less attribution of responsibility (r=.230; p i .01),5 less attribution of guilt (r=.232; p i .01), fewer murder verdicts (r=—.173; p i .05), more not guilty verdicts (r=.263; p i .01), and lighter sentences (r= i .172; p i .05) as well as increased attribution to accidental forces (r=.l35; p i .05). Results regarding Hypothesis 3b are less conclusive. While most correlations shown in columns two and three of Table 3.20 are in accordance with the expectations stated in the hypothesis, only a few of these correlations are Ho. mo. V V m um useoewflcmem a pm pcmoacuaccmo n .oom W z W mmm “oouuHEo mommo mcflmmazc comm.u Och.u ccc.u umcm.u nmcH.u mocmucmm ccc. mac. accfi. ch. emcm. confine uoz mma. mcc. cmH.- cad. mac.u empamsmamamz .u .I .I .I .I no no omcm nmca mcc pace nmca a z % cHH.- ccc.- cNH.- cmH.- ommm.- Basso cmca. oHHN. ccH. cca. cmc. nuance Q muaaflcflmcommom mmc. mmc. occH. cmc.u Hmc.u umpmam suaaflnam:0dmmm cca.u ccc.n mec.u mHH.u comm.) pamocmmmc suflaanam:0dmmm ummm ncca mac acme nmma. ncmoaoo< mcflamcmq mcaaonmq monocom >3umm§>m o>flucmmz m>HuHmom muflncaflfiam m.mnoumoaccH sunfish newcommom Mo coflusncuuud pom onQMHHm> moflco>uoucH cmmBuom mGOHumHOHHOU .om.m manna 86 statistically significant. For instance, positive evalua- tion of the defendant does lead to decreased murder verdicts ( =-.161; p i .05), increased accidental attribution (r=.158; p i .05), and lighter sentences (r=.205; p i .01). However, negative labeling does not correlate significantly with any of these variables. Instead, contrary to expectations, negative labeling is seen to correlate positively with not guilty verdicts (r=.168; p i .05). Support for Hypothesis 3c is somewhat stronger. Sympathy leads to increased accidental attribution (r=.l47; p i .05), fewer murder verdicts (r=-.162; p i .05), and lighter sentences (r=-.216; p i .01). Finally, Hypothesis 3d is generally supported since honesty correlates significantly (at the .01 level) with accidental attribution (r=.325), murder verdict (r=.273), and sentence (r=-.356). It should be noted that while some of the results concerning attribution of responsibility to the defendant's sister and father are significant, they are reported here only as additional information but will not be discussed in this work. In the discussion so far we have treated each of the intervening variables separately from each other. Such fragmented analysis has shown that negative evaluation is a weak predictor of AR, and that similarity, sympathy and honesty correlate only mildly with the dependent variables. 87 In order to explore this pattern of results further, and, at the same time, to probe the aggregate effect of all the intervening variables, a canonical analysis is performed. This analysis, presented in Table 3.21, yields a canonical correlation of .601 for the four-cell design which explains 36% of the variance. The x2 for such correlation is 57.36 and is significant at p i .037. The results for the six—cell design show a canonical correlation of .530 which explains 28% of the variance, and a x2 = 67.61 sig- nificant at p 5 .004. The meaningfulness of these results may be question— able due to the lack of a theoretical common denominator in the set of intervening variables. In order to overcome this problem the data were transformed following procedures out- lined by Van de Geer (1971), so that the effects of each of the intervening variables on the construct AR could be ob- served. The models presented in Figure 3.2 are the result of such transformations. An analysis of these models reinforces the finding that negative evaluation of the defendant has a negligible effect on AR. Furthermore, they show that honesty has the greatest influence on AR. In regard to the indicators of AR, the high coefficients obtained for sentence and accident indicate their reliability as indicators of AR and clarify why these variables systematically correlate with the inter— vening variables as shown in the discussion of Table 3.21. 88 Table 3.21. Canonical weights and Canonical Correlations Between Indicators of Intervening variables and Indicators of .Attribution of Responsibility for Four Cell and Six- Cell Designs.a Eigenvalue . Canonical (variance 2 Degrees of Signif— DeSign Correlation explained) X Freedom icance Four cells (N=72) .601 .361 57.36 40 .037 Six Cells (N:108) .530 .281 67.61 40 .004 Canonical weights Intervening Variables Four Cells (Ne72) Six Cells (N=108) Honesty .634 .742 Similarity .436 .274 Sympathy . 270 .345 Negative evaluation -.029 -.115 Positive evaluation .166 .135 Indicators of Attribution of Responsibility Sentence -.457 -.419 Accident .597 .367 Responsibility Defendant -.l75 -.236 Responsibility Sister .212 .185 Responsibility Father .030 .186 Guilt .087 .362 Nnrder'verdict —.069 -.427 Manslaughter verdict -.108 -.221 aFirst canonical variate only. 89 Sentence 6 =, . 80 Accident Responsibility - . 59),. Defendant Honesty . 63 . . . x , Responsibility Similarity y, AR “.7, Sister Sympathy Responsibility Negative Father Evaluation 63Guilt Positive . E luati on 18 Murder Verdict \ Manslaughter Verdict (a) Four—Cell Design (N=72) Sentence Accident Responsibility Honesty .74 93", Defendant . . . .27 Responsibility Similarity - M Sister . \\ \ Sympathy \\~1~58\\ Responsibility N . -.l \ ‘ Father egative Evaluation / \20x _ . '73 GLlilt . 28 Murder Verdict . 14 Positive Evaluation . Manslaughter Verdict (b) Six-Cell Design (N=108) Figure 3.2 Causal Nbdel of the Relationship Between Intervening Variables and AR Variables . Correlations Between Independent Variables , and Error Terms of Indicators of AR not Included. 90 Summary In summary, in this chapter we have presented results concerning: (a) equivalency of Ss in the experimental groups, (b) influence of Ss' background characteristics on AR, (c) Ss' perceptions of the experimental variables, (d) Ss' per— ceptions of the intervening variables, (e) relationship be— tween liking and AR, (f) relationship between disclosure source and avowal type, and AR, and (g) relationship between honesty, similarity, sympathy and attitudes, and AR. In the next chapter we will discuss these results and analyze their theoretical and pragmatic implications. 91 FOOTNOTES 1Throughout this chapter the total sample size is used in the presentation of those results which derive greater statistical stability and/or contextual meaningfulness from a larger sample size. The analysis of possible biases in the random assignment of 85 to groups is a case in point since 88 were simultaneously assigned to all 13 experimental conditions of the overall study. 2All significance levels presented in this section are based on N=200 and, therefore, represent conservative estimates, given that the actual sample size is about 2600 3Significance levels calculated on N=200. 4There exists a statistical technique developed by Joreskog (1970) which allows for the testing of a complex causal model as a whole, instead of breaking it down into various submodels. This technique was not used in the analysis of the data because the author had no access to any computer facilities where the program necessary for the analysis was operational. 5Significance levels are calculated on N=200. 6The coefficients presented in the four and six-cell designs are always similar to some extent because two- thirds of the 88 being compared are the same people. However, assuming that the experimental manipulations do have an effect, the addition of the 36 control 85 in the six-cell design should increase the error variance of the calculations. CHAPTER IV DISCUSSION Summary This work is concerned with the effects of some com- munication strategies used by an alleged offender standing trial on the decisions rendered by a simulated jury regarding the offender's degree of responsibility. The theoretical underpinnings for the study come from attribution and equity theories. Equity theory provides a conceptual equation for arriving at decisions regarding the fate of a defendant based on a variety of perceived rewards and costs experienced by the offender and his victim in the course of their relation- ship. Attribution theory provides some guidance in identify- ing the relevant inputs and outcomes to be entered in the equity equation. One of such inputs is the juror's liking for the defendant. Such liking can be greatly influenced by communication variables such as the content of the information regarding the defendant's behavior and by the source of such information. The interaction of these two variables also af- fects the jurors' stereotyping of the defendant, their feelings of sympathy for and similarity with the defendant, and their perceptions of his honesty. 92 93 In order to test the relationship between these vari- ables and attribution of responsibility (AR), 120 undergrad— uate students enrolled in various social science courses were asked to imagine themselves as jurors, and render a judgment regarding the responsibility of a defendant accused of murdering his father while attempting to defend his sister from the father's attacks. In a 2 x 3 design, 85 heard either the defendant (self disclosure condition) or a witness (other disclosure condition) disclose actions per- formed by the defendant which were either negative (negative avowal condition), positive (positive avowal condition) or neutral (neutral avowal condition). Subsequent analysis examined: a) The influence of these experimental manipulations on the Ss' AR to the defendant; b) The relationship between AR and four intervening variables: perceived honesty of the defendant, labeling of his communication, feelings of simi- larity with the defendant, and feelings of sym— pathy toward the defendant; c) The relationship between liking for the defendant and AR. Results indicate: a) Strong suppprt for the hypothesis that liking for a defendant leads to less AR. In fact, liking indicators shared as much as 42% of the variance C) 94 of a set of indicators of AR; Mixed support for the hypothesis that there is an interaction between disclosure source and avowal type such that under conditions of neg- ative avowal, less responsibility will be at- tributed to a defendant if deviance is disclosed by the defendant himself rather than a witness. And, under conditions of positive avowal, less responsibility will be attributed to a defendant if the positive information is disclosed by a witness rather than the defendant himself. While none of the evidence regarding this hypothesis is significant at an acceptable level, the expected interaction did materialize in the analysis of the "not guilty verdict" and suggested sentence, but failed to materialize in the analysis of at— tributed responsibility, guilt, and the "second degree murder verdict." Further, a canonical analysis yielded a canonical correlation of .44 between the experimental variables and the indica- tors of AR. Although this correlation indicated that 19% of the variance is shared by both sets of variables, such coefficient is significant only at p i .09; Support was also found for the hypotheses that similarity, sympathy, perceived honesty and positive 95 evaluation of a communication lead to decreased AR. A canonical correlation of .60 between the set of intervening variables and a set of indi- cators of AR was found. Such correlation is sig— nificant at p < .037 and explains 36% of the var- iance. Analysis of the canonical weights of the predictor variables indicate that perceived honesty was the best predictor of AR, coming and that labeling processes were virtually useless in predicting AR. Discussion The discussion of the results is organized in the fol- lowing manner: first, we will make a general statement about the theoretical framework for this study (i.e., equity and attribution theories); second, we will systematically dis- cuss the theoretical implications of such framework regarding (l) the victim-related variables, (2) the defendant-related variables (e.g., AR, and liking and its dimensions), and (3) the theoretical rigor of the equity model; third, we will review the experimental manipulations and the way they relate to each of the intervening variables: (4) perceived honesty, (5) positive and negative labeling, (6) similarity, and (7) sympathy; fourth, we will analyze the generalizability of the findings; fifth, based on the previous sections we will propose some suggestions for future research; and sixth, we will end this work with some general conclusions. 96 The Theoretical Framework In the theoretical analysis it was pointed out that jurors may seek to apply the law in equitable terms; that is, in the process of attributing the responsibility of an offender, they use the following equation derived from equity theory: Outcomes A _ Outcomes B Inputs A Inputs B In the present study, A is the defendant, B is the victim, and their inputs and outcomes adopt the following values: luirflmmimioflkspmmfibIUty _ D§uh Infbrmation about the defendant _ Information about the victim It should be noted that both the information about the victim and his death were kept constant in this study. As for the defendant, the attribution of his responsibility is the dependent variable, and the information about him (communica- tion strategies), as well as the way this information may be interpreted (liking, honesty, labeling, similarity and sym- pathy), represent the independent and intervening variables respectively. From this theoretical analysis the following implica- tions can be derived: 1) Regarding the victim—related variables: 97 The servity of the offense (i.e., outcome to the victim), the attractiveness of the victim (i.e., the sum total of his inputs), and the relationship between these var— iables represent important determinants of both the responsi— bility attributed to, and the perceived attractiveness of the defendant. Therefore, given that in this study the perceptions of the victim-related variables were not measured, their rela- tion to the independent and intervening variables will remain unknown and subject to future study. In practical terms this means that we cannot be sure that: (a) the Ss perceived the victim-related variables as constant, and, furthermore, that 85 with different perceptions were not systematically distrib- uted in the experimental groups; and (b) the great severity of the offense presented (murder of the defendant's father) did not reduce the variance of the dependent variables, thus moderating the relationship advanced in the hypotheses. 2) Regarding the defendant—related variables: Attribution of responsibility: As mentioned in Chapter I, there are at least five levels which can be dis- tinguished in the process of AR: association, commission, foreseeability, intentionality and justification. Because of the nature of the case used in this work, intentionality and justification represent the crux of the dependent variable. Unfortunately there are no data to support the experimental assumption that the Ss would perceive their task as that of determining whether the crime committed by the defendant was 98 (un)intentional and/or (un)justifiable. Furthermore, the fact that there are two distinct levels of attributions in- volved in the determination of AR leads to the theoretical need to make explicit the sets of necessary antecedent con- ditions which trigger each of these processes. This sug- gests another line of action for the systematic development of theory and research. Liking and its Dimensions: As for the defendant's inputs, one outstanding finding in the present study is the importance that liking for the defendant has on the AR pro- cess. This finding not only corroborates similar evidence reported in previous studies but it also furthers our under- standing of the relationship in at least two ways. First, previous studies such as those carried out by Landy and Aronson (1969), and Nemeth and Sosis (1973) have manipulated the attractiveness of a defendant by varying characteristics such as family status and defendant's occupation. Thus, in the Nemeth and Sosis (1973:227) study, the attractive defen- dant is presented as "a likeable, middle-class. upstanding citizen with no criminal record," and the unattractive as "a working class citizen who has marital difficulties and has a criminal record." In the present study we have explored how attractiveness can also be influenced by communication variables, such as source and content of information. More- over, we have shown that even under conditions of extreme un- attractiveness (i.e., a low class, socially disruptive teen- 99 ager who has killed his father), liking for a defendant is still an extremely powerful determinant of AR. Secondly, we have theorized that liking for a defen— dant can be conceptualized as indicating the sum-total of his assets. Furthermore, we have posited four variables which might be construed as affecting liking, namely perceived honesty, similarity, sympathy, and labeling. Although the relationship between these intervening variables and liking was not presented as a formal hypothesis in this study, such relationship was considered in the causal model presented in Chapter I (see Figure 1.2). Results presented in Table F. l of Appendix F show significant (p i .001) canonical correla- tions of .67 and .68 between the intervening variables and liking. These correlations account for 46% of the variance in both the four and the six—cell designs. 3) Regarding the Theoretical Rigor of the Equity Model: If one were to take the equity formula previously presented at face value, the logical conclusion would be that we are confronted with a systematic, axiomatized, well develop— ed theory. In fact, one would be tempted to consider this formula as the expression of a general law because of the apparent clarity of its components and the precise relation— ships among them. Regretfully, this is not the situation because of the following reasons: a) The variables which intervene in the equa- tion have not been clearly and unequivocally 100 operationalized and thus, their identifica- tion and precise measurement is not feasible. b) The mathematical relationships expressed in the equation such as the additive models which define numerators and denominators, the ratio models which define each term,and the identity sign which relates both terms have not been systematically derived and tested. Therefore, no serious mathematical models are posited; rather, the whole formula represents a conceptual model geared to sen- sitize people to the ideas of exchange theor— ists. The lack of systematic theory building has led to a situation where any behavior can be perceived as an input or an outcome, an asset or a liability, a cost or a reward. Furthermore, there are no clearcut predictions of the response of a subject under certain conditions; that is, equity can be achieved by changing actual or perceived behaviors in any of the components of the equation. In this work, an effort has been made to overcome some of these problems by: a) Providing a specific referent for the inputs of the defendant (i.e., liking). 101 b) Exploring the relationship between such input and a few variables (e.g., perceived honesty, similarity, sympathy and labeling). c) Determining the relationship between liking and the output to the defendant (i.e., re- sponsibility attributed). d) Attempting to study these variables and relationships in a controlled experimental manner, and through the manipulation of communication variables. The first three of these objectives have been ful- filled to a great extent. The achievement of the fourth has been hindered by Operational flaws which are discussed in the following section. The Independent and Intervening Variables If, as we have seen, variables such as honesty, simi- larity, sympathy, and labeling have an impact on AR, the question arises as to how different communication acts influ- ence these variables. This question is especially important since its answer may help explain the lack of significant re— sults concerning the relationship between communication strat- egies and AR. Evidence regarding this question was presented early in Chapter III. In summary, it was found that: 102 a) The messages intended to have a positive content were not perceived as such by the control 85, while 85 exposed to these messages did find them positive. b) The messages intended to have a negative content were perceived as negative by the control Ss, but not so negative by the Ss exposed to them. c) The manipulation of source of disclosure does not seem to have had a strong effect given that Ss perceived the defendant equally open in the self disclosure and the other disclosure conditions. 4) Perceived Honesty: Given the lack of strength of the experimental manipulations reflected in the previous results, it is only logical that such manipulations would pro- duce only a fair degree of perceived honesty. Furthermore, these weak results show a significant main effect due to source of disclosure (SD), which means that the Ss perceived the defendant equally honest regardless of the content of the disclosure. The implication is that negative SD did not increase perceived honesty, and that positive SD did not pro- duce a feeling of dishonest ingratiation. Similarly, the weak perception of ingratiation pro- duced was mainly due to the type of avowal regardless of the source of disclosure. This situation contributed to 103 some absurd results. For instance, those 85 in the positive avowal-other disclosure condition report that the defendant's testimony had been perceived as 66% ingratiating, despite the fact that the defendant had not made any explicitly positive statement at all! While the failure of the experimental manipulations to trigger the expected psychological processes could be interpreted as lack of support for our theoretical expecta- tions, the previous findings lead us to conclude that because of the weakness of the experimental manipulations, this study cannot be considered as a completely fair test of such hy- potheses, and that more adequate evidence should be gathered before judging their conceptual validity. Despite these operational flaws, the results of this work have provided conceptual considerations which have en- riched the initial theoretical positions. In Chapter I we reasoned that a defendant who avows deviance may be perceived as an honest individual whereas a defendant who presents himself in a favorable light may be perceived as attempting to gain the favors of the jury by presenting his nice qual- ities. While this rationale does make sense, this study has made clear that a defendant who avows deviance could also be perceived as sick or dumb by others. Thus, it may be that honesty is a characteristic dif- ficult to perceive on the basis of information content alone, or even when such information is placed in a particular context. It may be that honesty is a characteristic that 104 people do not assign very readily to other people. It may also be that such attribution necessarily presupposes actual interaction between both parties. Furthermore, this sug- gests that honesty may be perceived as a personality char- acteristic, so that an individual with a criminal history would be perceived as dishonest regardless of the type of information that he subsequently provides. In this respect, some researchers (Ekman and Friesen, 1969) suggest that peOple infer honesty from the nonverbal demeanor of the source. This interpretation, firmly grounded on everyday experience and on systematic observations, sug- gests that an individual who is presented only with a type- written transcript of a defendant's testimony will have a difficult time trying to decide whether the defendant is an honest person or not. All of these arguments clearly point toward the need for designing a study in which people can engage in actual interaction for a reasonable period of time, so that some of the conditions necessary for attribution of honesty such as spontaneity and voluntariness can be present. Finally, it should be mentioned that the relationship "perception of message leads to perceived honesty" is not necessarily an irreversible one. On the contrary, it seems plausible that the perceived honesty of a person should have an impact on the perceptions of messages received from that person. For instance, it would be expected that if a person 105 perceived as honest presents negative information about self, he/she would be evaluated quite negatively, but if he/she presented positive information, the evaluation would be very positive. On the other hand, if a person perceived as dis— honest presented positive or negative information, the eval- uation of him/her would probably be somewhere between the positive and the negative ends of the scale, as a cautious response on the part of the receiver. This interaction is graphically depicted in Figure 4.1. Interpersonal Evaluation + : Honest Individual ///oDishonest 0 Individual __ O — + Avowal Type Figure 4.1. Interaction Between Perceived Honesty, Avowal Type and Interpersonal Evaluation. 106 Moreover, from an interpersonal perception standpoint, a message which apparently goes against the best interests of the source may not be taken at face value by the receiver, and instead, may be perceived as manipulative information. This argument stresses the many levels of perception of in- tent between people (cf., Hastorf et al., 1970), and jointly taken with the argument of reciprocity in the relationship, points out a conceptual complexity between honesty and inter- pretation of information, not fully realized at the beginning of this work. 5) Positive and Negative Labeling: Generally weak and insignificant results were observed regarding the positive and negative labeling pro- duced by the various manipulations of disclosure source and avowal type. More positive evaluation occurred whenever there was any information about the defendant regardless of its nature or source. As for the negative evaluation measure, Ss in the self disclosure-negative avowal condition reported greater perceived criminality. In our theoretical analysis, positive and negative labeling reflected our interest in the process of stereotyping an individual based on limited information about him. It was thought that negative information about an individual would quickly lead to a negative categorization and subse- quent dislike for him. These expectations were weakly re- flected in the results of the study (see Tables 3.9 and 3.10), 107 and did not reach significance because of the relatively large amount of error variance. This may be indicative of measurement error and/or the presence of other independent variables which account much better for the variance of these variables. In this case, we believe that both of these factors affect the results. First, the question "to what extent did Johnny's testimony impress you as revealing of his criminal disposi- tion?" can be considered only as a crude approximation to the concept of stereotyping. A similar question had been found troublesome in the pre—test and accordingly it had been rephrased. Still, Ss may have found this question rather complex, and therefore, its measurement may have been unreliable. Second, the question "To what extent did Johnny's testimony increase your general opinion of him?" seems to measure a generalized attitude toward the defendant. One might hypothesize that such a general attitude could be a result of the stereotyping process. But such a result would very likely be influenced by a host of other factors besides the content of the messages exchanged in a relationship. These factors would become relevant as the relationship de- velops over repeated instances in social intercourse. Further empirical support for the contention that these questions entail faulty measurement is provided by the positive correlation between them. Indeed, since our , \} . . l ‘ 3|(‘ 1 1x: ‘ x l 108 theoretical rationale advanced that the positiveness or negativeness of the message would produce positive or neg- ative labeling, one would expect that the answers to these two questions should be inversely correlated. Since the actual result is r=.21, there is reason to believe that the questions are not reliable indicators of the labeling process. 6) Similarity is a troublesome variable in that, as Shaver (1970) points out, there are at least two crucial aspects to it. First, situational similarity refers to the likelihood that a juror will find himself in a situation similar to that of the defendant. Second, personal similar- ity refers to the extent to which the juror perceives him- self as being similar to the defendant as a person. If we take these brief definitions as the necessary conditions that trigger perceived similarity between two people, it follows that these conditions should be reflected in any message intended to produce feelings of similarity. It is our contention that the very weak perceptions of similarity with the defendant reported by the 85 are quite logical in that: (a) none of the messages used in the experimental manipulations fulfilled the conditions mentioned above; and (b) the experimental situation which presented the defendant killing his father is so dramatic that it is plausible to assume few Ss would imagine them- selves involved in a similar situation. Moreover, the 109 defendant was reported to have lived such an unusual life that any feelings of personal similarity might have been prevented. Contrary to our initial theorizing, the results of this study have shown that it is not plausible to expect that, in general, if a message carries a content which is positive, this will increase feelings of similarity on the part of the receiver. Conversely, there is no reason to expect that if the message is negative in content, this will automatically increase perceived dissimilarity on the re- ceiver's part. All of these arguments lead us to suggest that in future studies similarity should be manipulated in one of the following manners: a) Pre-test 85 on several important topics, and during the experiment have confederates systematically agree or disagree with the Ss' positions, thus creating different experimental conditions. b) Compare a situation where a friend of the juror committed a crime versus a situation where the offender is a stranger. This would allow us to manipulate similarity in an indirect subtle manner. While these suggestions are not the only possible ones, they make relevant the points that similarity can be llO manipulated through communication, and that it is an impor— tant determinant of AR. In this connection, the theoretical rationale would be that jurors' decisions may reflect their need to enhance their own control over their environment, to feel that their world is ordered, patterned, and therefore predictable. Thus, if a juror sees self as similar to a defendant, or feels possible that he might face the same situation some day, it would be expected that he will be more lenient in his AR. 7) Sympathy: The absence of significant results re- gardimgthe variable sympathy illustrates what appears as a clear operational flaw in the study. Indeed, in the initial theoretical analysis we proposed that sympathy could be the result of a message indicating that its source has experi- enced some negative outcomes or unjust suffering. It is quite obvious that the negative message presented to the 85 does not have any of these characteristics. If anything, it indicates that its source has experienced some underserved positive outcomes by robbing a store without any subsequent punishment. Thus, sympathy is not produced by the experimen- tal manipulations simply because there is nothing in the information to be sympathetic about. One strategy of communication which would involve sympathetic information is the "sad tale" which has been dis- cussed by Scott and Lyman (1968:52) as a technique for ac- counting for present behavior through an "arrangement of lll facts that highlight an extremely dismal past, and thus 'explain' the individual's present state." This strategy is currently very relevant because it is closely connected with a recent trend in sociological theory which explains deviant behaviors not as the free ex- ercise of a person's will but as a result of an unfair soci- ety which prevents the full development of some of its mem- bers, thus forcing them into socially disruptive behaviors (cf., Simmons, 1969). The adoption of such a view by some of the experimental Ss would explain whatever degree of sym- pathy for the defendant was reflected in the results. From this discussion it follows that the whole sym- pathy issue, and in fact the whole AR issue, are determined to a great extent by the attitude toward deviancy that the perceiver has. More generally, it may be that this attitude is dependent upon personality characteristics such as intern- al-external locus of control (cf., Joe, 1971). At the theo- retical level, the implication of this discussion is that this variable should be built into the design of future studies. At the methodological level, one implication is that great care should be taken in the phrasing of AR ques— tions. For instance, the question "why did he do it?" implies personal responsibility while the question "what caused him to do it?" implies external responsibility. A recurrent issue in all of the findings presented in this section is the lack of power demonstrated by the 112 manipulation of independent variables in producing any ef— fects on the intervening mechanisms. Such weak effects rep- resent a plausible explanation for the insignificant findings linking communication strategies and AR. Clearly, if the manipulations were not perceived by the Ss in the way they had been intended, there is no reason to expect the theoret- ical hypotheses to be supported by the evidence. In the pre- vious pages we have tried to analyze why the experimental manipulations were ineffective. In the following section we will deal with several other shortcomings of the study. Generalizability of Findings There are some obvious limitations to the generaliz- ability of our findings. First, the Ss were college students who are probably different from the populations from which actual jurors are drawn in terms of their age, education, and personality characteristics. Second, the Ss rendered their judgments in the absence of any group deliberations. Third, Ss knew that their task was hypothetical and that their judgments would not have an effect on a defendant. Fourth, the simulated jury technique used in the investigation departs from actual court proceedings in several ways (e.g., Ss were in a classroom; they read and listened only to a summary of the case; they knew they were participating in an experiment). While these are all definite shortcomings of the study, there are reasons to believe that their unwanted influence may 113 not be so powerful. First, various studies which have uti- lized the simulated jury technique drawing the 85 from dif- ferent populations have reported findings similar to those reported in field studies in which actual jurors have been interviewed. Furthermore, recent studies (Miller and Siebert, 1974), which used both students and actual jurors as 83, report only a slight tendency towards leniency on the part of the students. Second, there is evidence to suggest that actual jurors arrive at their decisions before deliberation starts (Kalven and Zeisel, 1966). It has also been found that, in some experimental settings, decisions rendered by a group do not differ significantly from decisions rendered by individuals (Stephan, 1973). Third, the experimental instruc- tions strongly emphasized to the jurors the importance of their performance, and the possibility that their responses could have some effects on future reforms of the trial pro- cess. Finally, within the resources available to the inves- tigator, every effort was made to bolster experimental realism. Accordingly, details from an actual trial were used as the experimental stimulus, and the experimental manip- ulations were designed with legal assistance. Given these arguments, and given the difficulty of doing research on actual jurors, we conclude that the simu- lated jury technique provides an adequate starting point for understanding jury processes, and that results obtained by using this technique should be generalized with caution. 114 A related, yet different kind of limitation refers to the context in which the study was carried out, and its re- lationship with the theoretical requirements imbeded in the different rationales that we wanted to test. This would translate to the questions: does this experiment trigger the necessary antecedent conditions which have been found in the literature on equity, attribution, self-disclosure, ingra- tiation, liking, etc.?. Is it reasonable to apply concepts, definitions and findings from these bodies of knowledge to such a structured context? Aside from the usual arguments of artificiality level- ed against experimental designs, these questions ask whether or not the scientific test of the theories was a fair one. Unfortunately, there is not a clear-out answer. On the one hand, it is quite clear that the context in which previous studies have been carried out has been very different from the context of this study. A few of the dif- ferences are: a) Dyadic interaction versus listening and reading. b) Unstructured situations versus structured ones. c) Free flow of information versus restriced flow. d) Informal situations versus formalized norms. e) Spontaneous versus contrived interaction. On the other hand, the theoretical fields of concern to this study are ambiguous in their conceptual requirements. 115 Thus, while all of them assume some sort of interaction as a basis, most of them do not present a clear-cut, agreed upon set of antecedent conditions. Therefore, there was no a priori reason to believe that this study would deny such conditions. In conclusion, it seems to us that some of the find- ings of this study are generalizable, especially those con— nected with the relationship between the intervening vari- ables and liking, and between liking and attribution of responsibility. Other findings, especially those regarding the com- munication strategies, cannot be generalized since we have partially explained their ambiguity on the grounds that there were some operational and experimental flaws, and that the context of the study was very restricted. Here we are con- fronted with variables which more clearly require other con— texts (e.g., self disclosure and ingratiation). In any case, the fact that we have used this argument to defend our theo- retical position, even in the face of contradictory results, only points out the need to clarify and systematize the antecedent conditions involved in these theories. Suggestions for Future Research The troublesome pattern of results found in this study points out the need for further research designed to extend, clarify, and evaluate several theoretical and methodological aspects left ambiguous in this work. Throughout this chapter 116 we have already made several suggestions for future research. In this section, an effort will be made to systematize such suggestions and include other relevant ones. 1) Regarding the victim-related variables: From our discussion of these variables we conclude that a line of research which can be pursued is the concep- tual replication of this study at varying levels of severity of offense. Walster (1966) found that 85 tend to assign greater responsibility for an occurrence to some appropriate person when the consequences of such occurrence are more severe. As we repeatedly pointed out, the present study used an experimental stimulus in which the offense was ex- tremely severe. This situation probably diminished the vari- ability of the dependent variable and suppressed higher cor- relations with the experimental variables. Thus, we would expect the variables used in the present study to be even more powerful in determining AR when the offense committed by the defendant is not as severe as killing one's father. It is necessary to study the juror's perceived rela- tionship with the victim. This could be done by taking a balance theory approach which would also systematically deal with the relationships juror-defendant and defendant-victim, which are both very relevant in an exhaustive equity theory. These relationships can be easily depicted in a standard co— orientation model: Several hypotheses relevant to AR could be derived from this model. Two of them are: Juror / v ' \ Defendant f Victim . \'\/ Figure 4.2 A Coorientation Model of the Juror-Victim-Defendant Relation- ships. /.*- a) The more attractive the victim, the higher the responsibility that will be attributed to the defendant. b) The more equitable the relationship between defendant and victim, prior to the commission of the offense, the greater the drive to re-balance the relationship by attributing greater responsibility to the defendant. 2) Regarding the defendant-related variables: First of all, it is necessary to clarify the different levels of AR and specify the antecedent conditions that lead to each of them. Then, it is necessary to develop and validate a taxonomy of messages and situations which will fulfill those conditions and therefore, trigger the AR process. 118 Such messages need not be confined within the context of the legal setting but may easily include similar situa- tions in which an individual finds himself in the position of having to account for his behavior to another individual or social group (i.e., "behavioral accounting"). In this wider context, some relevant work has been done by various authors. For instance, Sykes and Matza (1957) have discussed several "techniques of neutralization" or justifications for deviant behavior which are seen as valid by delinquents but not by the legal system or the society at large. Five major types of such techniques include: denial of responsibility, denial of injury, denial of the victim, condemnation of the condemners, and appeal to higher loyalties. This discussion has been complemented by Scott and Lyman (1968) who discuss several types of excuses such as: appeal to lack of knowl- edge, appeal to biological drives, scapegoating, and appeal to accidents. The crux of the problem is, of course, to find out which of these accounts are more effective under various conditions. As for the intervening variables, we have already pointed out certain antecedent conditions that trigger per- ceived honesty (i.e., any message that conveys spontaneity, voluntariness), similarity (i.e., any message that agrees with same beliefs, attitudes or values of the receiver), sympathy (i.e., any message that indicates negative outcomes or unjust suffering of its source), and positive/negative 119 labeling (i.e., any message that states the commission of desirable/socially disruptive behaviors on the part of its source). This approach to the study of communication variables has far-reaching implications for communication science. For instance, the notions just discussed reflect a reorienta- tion of thinking in the sociology of language which views linguistic behavior not as an expression of individual pri- vate states but as an indicator of future actions (i.e., language as coordinator of social action). From this per- spective, the explanation of a particular type of human be- havior follows from "an analysis of the integrating, control- ling, and specifying function a certain type of speech fulfills in socially situated actions" (Mills, 1970:472). The study of behavioral accounting also has implica- tions for other theories in social psychology. For instance, in a discussion of equity theory and legal structures, Ma- caulay and Walster (1971) have identified some of the tech- niques through which a harm-doer may restore equity with his victim. Interestingly enough, some of the techniques used to restore psychological equity are exactly those techniques of neturalization discussed by Sykes and Matza (1957). Accordingly, the study of the latter techniques would be a substantial contribution to equity theory. 3) Regarding the mathematical rigor of the 'relationships: 120 As we have pointed out, the relationships derived from equity and attribution theory lack mathematical rigor partly because of measurement problems and partly because of imprecise integration models. Both problems have been at- tacked by Anderson (1974), who in the last few years has de- veloped an information integration theory which posits the existence and operation of a general cognitive algebra. In a review of this work Anderson (l974:3) states that processes for integrating information follow simple rules of ordinary algebra in a wide variety of situations. This cog- nitive algebra appears to be a general property of the mind, since it is Operative in widely different substantive areas." Based on these ideas, Anderson presents several models with different degrees of mathematical sophistication: additive, multiplying, multilinear, ratio, several types of averaging models, and others. The conclusion of Anderson's survey is that there is enough evidence in the literature to support his claims. This approach opens great new possibilities in the study of social judgment, and these possibilities should also be probed in future research. 4) Regarding other related fields of study: The results of this research also have impli- cations for areas such as self disclosure, ingratiation and source credibility, and suggest some areas that need further clarification. For instance, in this study, the defendant 121 was presented as self disclosing either negative, positive or neutral information. Such a manipulation assumed a linear relationship between avowal type and the predicted variables (e.g., AR, liking). However, evidence in the literature on both self disclosure and ingratiation suggests that the nature of the relationship may be curvilinear. Cozby (1972) reports that Ss express more liking for medium disclosers than for low disclosers, but also feel a strong dislike for extremely high disclosers. Furthermore, Derlega, Harris, and Chaikin (1973) had 66 female Ss interact with a female con- federate who disclosed either very superficial information, highly intimate information of a conventional nature, or high— ly intimate information of a deviant nature. Ss like the confederate significantly less under the deviant-high self- disclosure condition. On the other hand, Jones (1964) suggests that potential ingratiators usually consider the possibility of detection and therefore tend to adopt a level of ingratiation which will not make the other person suspicious. Clearly, the assumption is that high levels of ingratiation may backfire and result in strong dislike for the ingratiator. Regarding the source credibility variable, in Chapter I we reported the finding of Walster gp_al., (1966) that "any communicator, regardless of his prestige, will be more effective and will be seen as more credible when he is argu- ing for a position opposed to his own best interest, than 122 when arguing for changes obviously in his own best interest." The results of this study seem to indicate that their con- clusion might not be as generalizable as they claim. That is, whether a position is Opposed or in favor of the source's best interests is something to be decided by the receiver ratmarthan inferred from the message. In other words, second- guessing the source's intentions is a possibility that should be considered. Secondly, it would appear that the phrase "regardless of his prestige" might not be applicable when considering sources with negative credibility. Thus, further research is needed in order to clarify the relationship between self disclosure, ingratiation, source credibility and liking, and the subsequent effects of this relationship on AR. Conclusion A goal in this work has been the merging of theoret- ical and applied concerns so that its results may be useful for other theory construction in communication, and practical applications in the legal community. Regarding the theoretical aspects of this research, an attempt has been made to clearly specify, whenever possible, alternative and sometimes contradictory hypotheses derived from different conceptual frameworks. Regarding the pragmatic aspects of this work, we have tried to contribute some knowledge about the workings of the legal system. 123 Although the experiment presented in this disserta- tion refers to a very specific setting such as a simulated jury, its theoretical underpinnings are much broader in scope. We have chosen to View the study of human communica- tion from the standpoint of exchange-based theories such as equity, attribution and identity-negotiation. The basic postulate in these theories is that human behavior (e.g., communicative behavior) will be affected by the actor's per- ceptions of the outcome structures of social situations (e.g., a court-room trial). In other words, as shown by Weinstein et al. (1968), Weinstein and Deutschberger (1964), Weinstein (1966), Archibald and Cohen (1971), Brown and Garland (1971), Schneider (1969), and many others, "differences in the struc- ture of available outcomes will affect the interpersonal strategies (i.e., communication behaviors) used in pursuing them" (Weinstein §p_al., 1968). On the other hand, social actors are usually aware of these tactics and strategies, and therefore seek to interpret the communication behavior of others in light of the possible motives (hidden agenda) that could produce such behaviors. These theoretical ideas have been the motivation and main thrust of this work. We believe that they reflect a truthful conception of the nature of man. Unfortunately, we have not succeeded in designing a study where all of the rich- ness and truth that we see in them would be adequately tested. Yet, we also believe that some of the findings are supportive 124 of such a conception. Thus, at the very end, we are forced to, once more, face and deal with the basic truth that any scientist must accept: regardless of the results, science is a painful process where one never dares to believe in that which one has found. APPENDIX A The Experimental Questionnaire 125 Court Procedures Study MICHIGAN STATE UNIVERS ITY Spring, 1974 (Study II) 126 Court Procedures Study PART I: INTRODUCTION This project is being carried out as part of a series of studies investigating the process of decision-making in the jury. The goal of this series of investigations is to understand how jury members come to their decisions as they receive information about a case as well as to suggest and try out needed innovations in present court procedures. The present jury system is under attack for a variety of important reasons. First of all, the insistence upon an absolute decision--that is, the choice between "guilty" and "not guilty"-—is considered unrealistic and unfair because in most real situations a person cannot be justifiably con- sidered "totally guilty beyond any doubt" or "totally inno- cent beyond any doubt." As individuals, we tend to assign to ourselves and to others varying degrees of guilt and in- nocence, and we assign such guilt or innocence with varying degrees of certainty. And yet jury members are not permitted to do this. This arbitrary restriction on the jury's de- cision leads to many difficulties in arriving at a decision, and also, many legal experts feel, to many unjust decisions. Therefore, some experts would like to have jury members be able to qualify their votes by indicating the degree of certainty they feel about the defendant's guilt or innocence. 127 Secondly, the insistence upon a unanimous vote of the jury has some very undesirable consequences. First, it some- times renders it impossible for a jury to come to a decision. If not, it often adds considerable time to the already cumber- some deliberation process. Secondly, it forces the jury members to discuss the trial and to go over each member's decision. This brings a great deal of social pressure to bear on the few members whose decisions are deviant from the group's judgment. Often, a unanimous decision is reached not as a result of complete agreement about the proper out— come of the case, but as a result of the social pressure and time pressure acting upon all members to agree with each other. Some innovators are suggesting that jury members vote separately and that the results be tabulated by a non-involv- ed party. Under this system, the percentage of agreement neces- sary for a verdict would be pre—determined. Thirdly, some critics argue that jury members should have a greater share in the interpretation of the laws they are dealing with. Many times a jury member finds himself in Opposition to the law which has been broken--that is, he feels that although an action may have been against the law, it was morally justified. Although this is a difficult issue to deal with, it reflects a problem that is quite frequent these days, and some believe that the jury's views on the moral justification of a defendant's actions should not be entirely beyond the realm of legal considerations. 128 Finally, the present jury procedures have been crit- icized for not allowing jury members to make decisions be- yond determining the guilt or innocence of the defendant. Many critics would like to see the individual jury members make recommendations about the sentencing or other punishment of the convicted defendant. The committee which is supporting these studies is considering making some of these recommendations regarding new court procedures, which might remedy some of the short- comings of the present situation. The purpose of today's study is to try out some of these proposed innovations to see how feasible they are and what implications they have for the decision-making process, before any definite recommendations are made. To do this, we are asking each of you to act as a jury member for a criminal case we will present to you. We want to find out how a jury member comes to decisions on the basis of the information about a case. For this reason, after you hear the testimony on the case, you will be asked to generate a set of judgments on the basis of the informa— tion you have received. The case you will be working with is a classic one. It has been condensed and edited by a group of professors at the Columbia University Law School for use in courses in criminal law, and is now being used at several law schools in this country. All the names have been changed, and the actual jury's verdict has not been made public by the editors. 129 But the remainder of the records--the testimony of the var- ious witnesses, the arguments made by the prosecuting and defense attorneys, and the statements made by the judge-—has been left intact by the editors, except that it has been shortened considerably through summarization of the less important points. We have been careful to provide you with materials selected from the court records of a real trial, and we think it is a case you will find interesting. In the next few pages you will find the background in— formation on the case. Later on, you will hear the prose— cution and the defense's closing statements. Finally, we will ask you to fill out a questionnaire regarding the de— cisions you have made on the basis of the information pro- vided. Please feel free to report your judgments in all honesty since your answers will be strictly confidential and will be made public only in the form of grouped, anonymous averages. Remember that you are now a juror evaluating an actual 2%. We thank you for your cooperation. 130 Court Procedures Study PART II: JUROR IDENTIFICATION FORM Below you will be asked to answer a set of questions which are representative of the lines of questioning pursued by most attorneys in the process of selecting a jury. A. Sex (please check) male female B. Date of your birth day month year C. Your Father‘s occupation: (or was, if dead or retired) (specify the kind of work he does, not where he works) D. The number of brothers and sisters you have is: (Fill in box) E. In your family are you the: (please check) First Born Later Born Only Child F. In terms of income or wealth of families in your com— munity, do you think your family is: considerably above average somewhat above average average somewhat below average considerably below average G. How far did your father and mother go in school? (Check one for each) FATHER MOTHER less than 8 grades 8 grades 9-11 grades 12 grades graduated high school some college college degree An advanced degree (Masters, Ph.D., or professional such as law or medicine) |||||l|| llllllll fair do you think the jury system is? CC m 0 £ 131 Court Procedures Study PART III: TRANSCRIPTS FROM TRIAL 132 State of New York vs. Johnny Marco I. Opening Statements l. glgpk: (Introduced and briefly described the case.) "Johnny Marco is charged with 2nd degree murder of his father, Frank Marco, on the night of August 9, 1957. It is charged that the defendant willfully and deliberately stabbed his father, causing him to die, although the act of murder was not premeditated nor planned. The defendant has pleaded not guilty to the charge of 2nd degree murder. He admits killing his father, but pleads that it was a case of simple manslaugh- ter in that the act was committed under sufficient mitigating circumstances to relieve him of criminal responsibility be- fore the law. The trial of the State of New York vs. Johnny Marco is now in session." 2. Prosecuting Attorney: (The attorney for the State briefly summarized the events leading up to the trial.) The prosecution would attempt to show that Johnny Marco, on the night of August 9, 1957, fatally stabbed his own father in cold blood; and that although this act could not be considered premeditated in the usual sense of the word, it was in cold blood and not either in defense or in rage and while in pos- session of his faculties. Prosecution asks for a verdict of 2nd degree murder. 3. Defense Attorney: (The defense made a rather long statement about the circumstances preceding the alleged act 133 of murder.) The defense admits that Johnny stabbed his father, but pleads extenuating circumstances. It would at- tempt to show that Johnny had never had a decent home life, that he had been brought up cruelly and without love, that his father and indeed the rest of society had always treated him as an animal, had not given him even the barest necessi— ties, and had never even attempted to teach him to be a civi- lized human being. That on the night of the killing he had been goaded by his father beyond endurance and had still con- trolled himself, that his father became openly violent, not only toward Johnny but also toward his young sister, and that in attempting to defend himself Johnny struggled with his father and finally stabbed him in the heat of the struggle. Defense asks for a verdict of manslaughter, and acquittal on the charge of second degree murder. II. General Information Frank Marco, born 1910, died 1957, the deceased and father of the defendant. Johnny Marco, 19 years at the time of the trial, defendant and son of the deceased. Angelina Marco, 15 years at the time of the trial, daughter of the deceased and sister of the defendant. III. Background Testimony Witness: Captain Anthony Buonauro, New York City Police Department. Captain Buonauro testified that on the night of 134 August 9th he was called to the Marco's apartment by neigh- bors. He knocked on the door which was opened by Johnny (whom he identified in court by pointing to him). In the small apartment he found Mr. Marco lying on his face on the floor with a switchblade knife beside him. There was blood on the knife and Mr. Marco was dead. Johnny's sister Angelina was crying loudly and Johnny was covered with blood. When Captain Buonauro asked Johnny what had happened, the boy re- plied, "I killed him, but he asked for it." On cross examin- ation, Captain Buonauro stated that when Johnny was told of being arrested he shouted, "No, I'm not going to jail," turned around and tried to leave by the fire-escape. Finally Captain Buonauro and Patrolman Kegel together were able to capture Johnny and take him into customy. Witness: Lowell B. Waterman, coroner. Dr. Waterman testified that the deceased had suffered three wounds in his chest, one of which pierced the heart and was fatal. The consulting physician supported Dr. Waterman's testimony and there was no cross examination. The following are excerpts from the actual trial con— cerning the testimony of the defendant. Defense: Johnny, how old are you? Johnny: Nineteen years old. Defense: Where did you attend high school? Johnny: Defense: Johnny: Defense: Johnny: Defense: Prosecution: Defense: Judge: Johnny: Judge: Defense: Johnny: 135 George Washington High School. Johnny, where do you live? 1405 East 103£d Street. Who lived with you at that address? My father and my sister. How would you describe your relations with your father? Your Honor, I object. That question is not relevant to this case. Your Honor, I believe it is relevant. I will attempt to show a history of hostility and provocation between Johnny and his father. The court will withhold ruling on the objection and will allow defense counsel a few minutes to demonstrate the relevance of this line of questioning. Johnny, please answer the question. Well, things were not that good between my father and me. We had some rough times to- gether. He really wasn't home that much, and neither was I. I spent most of my time with my Boy Scout troop. We organized the neighborhood alley clean-up last year. The jury is instructed to ignore Johnny's last statement. Johnny, please confine your answers to questions that are asked. Johnny, how long have you lived at the 1405 East 103£d Street address? For the last eight or nine years. (trial continues) 136 Now we will give you further relevant information on the case. This will be in the form of a tape recording of parts of the prosecution and defense attorneys' closing re- marks in court, and it highlights the important points of evidence they raised. You will hear some new evidence being presented in the tapes and some references made to testimony which you have not read, this is natural since we have short— ened various portions of the trial in order to present it to you in a short time. This tape, of course, is not an original, since the making of tapes for purposes other than those direct- ly connected with the proceedings of the case itself is pro- hibited. However, in order to represent, as best we can, the actual situation of a jury member in court, this reconstructed taped speeches are more suitable than merely a printed copy of the testimony. They strictly follow the written court record, and quote it verbatim. PLEASE DO NOT TURN THE PAGE AND START ANSWERING THE QUESTIONNAIRE UNTIL THE EXPERIMENTER HAS PLAYED THE TAPE RE- CORDING COMPLETELY. (If other people have not finished read- ing the case up to this point, you may have to wait a few moments for the tape to be played.) 137 After the prosecution and the defense rest their cases, the judge instructed the jury as follows: As to the question of guilt, it has already been established that the defendant killed his father. It is up to you as a jury member to decide whether or not he is guilty of second-degree murder. To help you make your decision, we repeat here the charges against the defendant: 'It is charged that the defendant willfully and deliberately stabbed his father, causing him to die, although the act of murder was not premeditated or planned, and that the act was not committed under sufficient miti— gating circumstances to relieve him of criminal re- sponsibility before the law.' You may find him guilty of this offense or of the lesser offense of manslaughter, or may find him not guilty as charged. You have been instructed, during the course of this trial, as to the applicable law in this case. You have also been advised that if Johnny is legally guilty of killing his father, he must be found guilty of either second degree murder or manslaughter. Your deliberations are confidential and your conclus- ions cannot be challenged. 138 Court Procedures Study PART IV: JUROR'S DECISIONS Now you are to make a judgment on the basis of the information available. Again, you are asked to imagine your— self in the position of a member of a jury, and the evidence you have heard is all you have available. Please answer the questions honestly and carefully as your responses may be very important to future trial procedures. Thank you. 1. To what extent would you say that the death of the victim, Mr. Marco was an accident? Answer any percent from 0 to 100, where O is not at all accidental and 100 is totally l6 l7 l8 accidental 2. How responsible do you think Johnny is for the death of his father? Answer any percent from O to 100, where 0 is not at all responsible and 100 is totally 19 20 21 responsible 3. How responsible do you think Johnny's sister is for the death of her father? Amswerany percent from 0 to 100, where 0 is not at all responsible and 100 is totally 22 23 24 responsible 4. How responsible do you think Johnny's father is for his own death? Answer any percent from O to 100, where 0 is not at all responsible and 100 is totally 25 26 27 responsible o\0 o\0 o\0 o\0 10. 11. 139 To what extent do you think Johnny is legally guilty of killing his father? Answer any percent from O to 100, where 0 is o\0 not at all guilty and 100 is totally 28 29 30 guilty . . 31 What 18 your verdict on the case? r——- Guilty of second degree murder l ' Please enter a one in the box at the right of the alternative you choose and a zero in the other two boxes. Guilty of manslaughter and 32 not second degree murder Not guilty 33 Assume that Johnny has been found guilty and that you are to assign a prison sentence to 34 35 him. If the option of parole does not exist, what sentence, if any, would you assign to him? Answer in number of years in prision. The maximum number of years for second degree murder is usually 30. To what extent do you perceive Johnny as being an honest person? 36 37 38 Answer any percent from 0 to 100, where O is not at all honest and 100 is totally o\0 honest To what extent do you think Johnny was morally justified for his actions in this case? Answer any percent from 0 to 100, where 0 i539 4O 41 not at all morally justified and 100 is totally morally justified o\o To what extent did Johnny's testimony impress you as information revealing his criminal disposition? 42 43 44p Answer any percent from O to 100, where o is not at all revealing and 100 is completely o\° revealing To what extent did Johnny's testimony impress you as an attempt to present himself in a favorable way? 45 46 47 Answer any percent from 0 to 100, where 0 is not at all favorably and 100 is completely o\0 favorably 140 In the next set of questions, choose the number in the scales that best corresponds to your feelings for each particular question, and enter it in the corresponding box at the immediate right of the question. NOT AT ALL VERY MUCH 48 12. How similar do you :1 2 3 4 5 6 7! ‘ J feel to Johnny? ‘ 13. How much would you , ‘49 like to really meet '1 2 3 4 5 6 7 ‘ Johnny? - 14. How close do you feel 50 to Johnny? {1 2 3 4 5 6 7 15. How much do you 51 admire Johnny? . ———- I1 2 3 4 5 6 7 I 16. If you got to know 52 Johnny better, how I r—— much do you think 1 2 3 4 5 6 7, 1 i you would like him? . . 17. How much did you 53 initially like ——- Johnny? 1 2 3 - 4 5 6 7 | 18. To what extent do you agree with the following statement? "Johnny's testimony increased my general opinion of him." 1 - strongly disagree Please enter 2 - disagree your answer in 3 - neither agree nor disagree 54 the box at the 4 - agree ' right 5 - strongly agree 19. To what extent do you agree with the following statement? "Johnny's testimony made me feel sympathetic towards him." 1 - strongly disagree Please enter 2 - disagree your answer in 3 - neither agree nor disagree 55 the box at the 4 - agree right 5 - strongly agree 20. 21. 22. 23. 24. 25. 141 Generally, how confident do you feel of your responses to all the questions presented above? Please enter your answer in the box at the right NOT AT ALL, VERY MUCH 56 r— CONFIDENT 1 2 3 4 5 6 7 CONFIDENT To what extent do you think you put yourself in the role of a juror in making the prior judgments? Please enter your answer in the box at the right NOT AT ALL VERY MUCH 7 —_ I ~— 1 2 3 4 5 6 7 1 Please enter your answers in the boxes at the right. U1 KO What is Johnny's age? .__...... U1 (I) {___l In what year did the incident take place? 60 61 19 In the following questions please answer with the scale given How free did you feel in arriving at a verdict? Please enter your answer in the box at the right NOT AT ALL VERY MUCH 62 FREE 1 2 3 4 5 6 7 FREE How important were the judge's instructions in arriving at your decisions? Please enter your answer in the box at the right NOT AT ALL VERY MUCH IMPORTANT IMPORTANT 63 r__. 1 2 3 4 5 6 7 142 26. How open do you think Johnny has been in this trial? Please enter your answer in the box at the right NOT AT ALL VERY MUCH OPEN OPEN 1 2 3 4 5 6 7! 64 In the following questions, choose the answer (true or false) that you consider correct, and enter the corre- sponding number in the box at the right. 27. In his testimony, Johnny mentioned being a member of a street gang. 1 - true 65 2 - false i 28. The judge instructed the jury to ignore part of Johnny's testimony. l - true 66 2 - false 29. The judge instructed the jury of their power to disregard his instructions concerning applying the law. 1 - true 67 2 - false 30. In his testimony, Robert mentioned Johnny being a member of a Boy Scout troup. l - true 68 2 - false 143 31. How bad do you think it is to be a member of a street gang and to rob a store? Please enter your answer in the box at the right NOT AT ALL VERY BAD BAD 69 l 2 3 4 5 6 7 32. How good do you think it is to be a member of a Boy Scout troup and participate in a neighborhood clean-up campaign? Please enter your answer in the box at the right NOT AT ALL VERY GOOD 7O GOOD 1 2 3 4 5 6 7 We will now ask you to transcribe your answers for questions 1 through 22 and 28 through 32a to the answer sheet provided in the following page. the instructions for doing this are as follows: 1. The number directly above each square at the right margin of the questions corres— ponds with the numbers in bold print on the following page. For instance, in ques- tion one there are three squares numbered l8, 19, 20 which match the same numbers in the second row of the answer sheet. 2. After matching the square number with the bold print number on the answer sheet, you are to black out the alternative that cor- responds with the number inside the box. For example, if your answer for question one had been 075%, you would black out alternative zero for number 18, alternative seven for number 19, and alternative five for number 20. 144 REMEMBER THE FOLLOWING: a) b) C) d) If a box is blank because the answer was a two—digit number or for any other reason, black out the zero alternative in the scoring sheet. Make sure you use the scoring pencil provided to you. Otherwise, your scores will not be read by the computing machine. Never black out more than one alternative in any given number in the scoring sheet. When you have finished, check to see if you filled out every number in the scoring sheet. Yours should have started at number 18 and finished at number 70. THANK YOU FOR YOUR COOPERATION DO NOT WRITE BELOW THIS LINE 717 72 73 74. 75 76 77 78 79 80 APPENDIX B Transcript of the Tape-recorded Statements 145 Defense Johnny Marco has lived all his life at one of the worst slums in the city. His mother died when he was very young and since then he has lived with a drunken father who has beaten him, belittled him and squandered the little wel- fare money he received on liquor instead of supporting his children. Johnny has gone through life wearing old and ragged clothing, often not sure of where his next meal was coming from. He has lived in an area where violence is a commonplace happening. In Johnny's area, violence is some- times the only means to prevent one's self from being bul- lied and pushed around. Johnny's life has consisted of a series of hard knocks, one after the other. His environment has continual- ly acted as a force, eroding his patience and challenging his stability. We cannot be surprised that he reacted vio— lently when faced with the extremely provoking circumstances of the night in question. Imagine this poor, mistreated boy coming home that night to find his drunken father attempting to abuse his sister Angelina, an innocent, defenseless, young girl. We know that Johnny generally tended to pro— tect his sister whenever it was necessary. As Mr. Long from the New York Welfare Department pointed out, the only way he could think of to get Johnny really angry was to say something negative about his sister. Johnny really loved his sister very much, and he idealized her womanhood and 146 her purity, the way it's done in the Italian tradition, a tradition which he had learned so much about from his late grandmother. Knowing that he held such ideals, it should not be surprising that his first reaction was to protect his sister from abuse and dishonor by attempting to stOp his father. He tried to protect her without intending to hurt his father. But he had no way to stop his father without attacking him physically. He was driven by the urgent and overwhelming desire to protect his sister. He could have had no better motivation. The prosecution would have us believe that Angelina was in no real danger when Johnny came home. You've heard Mr. Pucchini's testimony. Frank Marco habitually came into his bar in the evening and often during the day as well, as he held no steady job. He drank heavily. Mr. Pucchini said that Mr. Marco had to be helped home about 2 or 3 times a week. And that he occasionally became aggressive, "fighting- like," as he put it, when he had a little too much. Mrs. Pertelly who lives on the same street as the Marcos has testified that the women in the area knew that they had better avoid Frank Marco when he was drunk. More than a few of them had to reject his unwanted advances when he was in this condition. Mrs. Samsio, the Marco's upstairs neighbor, has testified that she has seen Angelina in at least one occasion with bruises on her face as if she had been beaten. 147 Mrs. Samsio also said that she overheard violent arguments between Frank Marco and his daughter usually when he came home drunk. You remember that Mr. Pucchini said that on the night of the stabbing, Frank Marco had a few beers and about a dozen shots of rye, and that he seemed a little more drunk than usual when he left. There cannot be a doubt that in this condition, Frank Marco was a threat to his daughter. The prosecution has asserted that even if Johnny felt it was necessary to defend his sister, he had no excuse for pulling a knife on his father. This is unfair. We've heard that Frank Marco weighed over 200 pounds and that he was considered the strongest man in the neighborhood before his drinking became heavy. Johnny weighs hardly more than 135. He tried to subdue his father with his bare hands, but he was no match for his father. He could not have pro- tected his sister without threatening him without some weapon. Johnny has consistently reported that he drew his knife only after failing to stop his father with his bare hands. The prosecution is trying to make a big issue out of the fact that on cross examination, Angelina admitted that amid all of the emotion and excitement she couldn't be certain whether or not she could remember Johnny trying to subdue his father first before pulling his knife. Angelina was in critical danger on that night and suffered through one of the most traumatic experiences a young girl can go 148 through. Being cross—examined was also a highly emotion— arousing experience for her. To Angelina, the whole evening was a nightmare and she admits that she is no longer certain about many things which happened. We should not take this understandable vagueness of Angelina's memory of the whole evening as evidence against Johnny on this one particular point. The prosecution has tried to show that Johnny meant to kill his father when he pulled his knife by mentioning that Johnny was experienced with a knife. But all the boys in the area carry knives, according to the social worker, Mr. Withum, and he has a great deal of experience with these boys. The knives are used as status symbols and play objects rather than as weapons. Johnny had owned a knife for years without ever using it as a weapon. We should not take Johnny's experience as a knife-handler as evidence that he intended to use his knife on his father. On that night, when Johnny attempted to protect his sister, his father, in a drunken state, advanced on the boy and they began to fight. Johnny risked being severely beaten as well as not being able to protect his sister; so he pulled out his knife, not with injuring or killing in mind, but to scare his father off. Now, instead of realizing his folly and standing back as a normal, sober man would do, Frank Marco lunged at his son and tried to get the knife away. At this point Johnny's fear was intense. He knew his father's violent nature and 149 saw the extent of his anger. During the scramble for pos- session of the weapon, Johnny suddenly felt he was fighting for his life. He was sure that if his father ever got the knife away, would not hesitate to use it on him, so in a moment of panic, he stabbed blindly at his father. We cannot blame him for resorting to the only means of defense he saw left to him. This boy's defense of his sister and of himself against a man who constituted a real threat to the physical and moral safety of his own children, this we are asked to consider as murder. The prosecution cannot ask this of us. Johnny's fear was understandably high. He showed remarkable control until the danger became unbearably great. Johnny may have been guilty of bad judg- ment in ever using the knife, but we should not charge him with the responsibility for murder. Johnny wasaiyoung, re- sponsible boy with human emotions, defending his sister and himself against a drunken, violent father. We have no just- ification for treating this boy like a ciminal. We may con- sider Johnny Marco a well-meaning but unfortunate boy whose impulsive actions in defense of his sister led to a senseless tragedy, but we cannot call him a murderer. I ask for a verdict of "not guilty." 150 Prosecution Johnny Marco was a bad boy. He was constantly getting in trouble; bullying people and taking what he wanted with little regard for others. He disrupted school and he beat other boys. His teacher, Miss Calderare, testified that he was a continual behavior problem. Often Johnny solemnly re- fused to reply to her questions, and once he shouted an ob- scene remark at her. Now this is hardly the picture of a peaceful boy. Nor are the frequent threats, the "I'll get you later," that he often put to those who dared to disagree with him. Many other boys have lived in poor families and not killed their fathers, how can this boy be excused just be- cause his life has not been easy? Instead of rising above his difficulties and becoming a decent citizen, he has taken the easy way out. Now going through the major events once more: Johnny came home one night and presumably found his father yelling at his sister. According to the defendant, he tried to stOp his father, and he struggled with him. Up to this point, his behavior sounds like what other boys might have done, but then ladies and gentlemen, then Johnny took a switchblade knife from his pocket. He stated that he merely wanted to scare his father off, but did he need a switchblade knife to do that? Johnny Marco, a healthy, young man of 19, asks us to believe that he could only stop a defenseless man, clumsy 151 with drunkenness, by pulling a switchblade knife on him. What is even more surprising is that the knife was not used to threaten his father, or even to wound him slightly. No, the knife was plunged three times into the man's body. Once straight into his heart. According to the police reports, there were no bruis- es on Mr. Marco's body, nor on Johnny, indicating a struggle had taken place. Now we must remember that Johnny was no clumsy novice with a switchblade knife. He couldn't have stabbed his father three times by accident. He stabbled the unarmed man and can give no better reason than "he just didn't think." Throughout this trial we have maintained that Angelina Marco was in no danger when her brother came home. And yet if we suppose for the moment that she was in danger, John- ny's actions still cannot be seen as justified. He clearly had other, better and safer, means to protect his Sister. If he really wanted to protect his sister, he could easily have done so without the knife. If his sister really was in danger, Johnny could at least have interferred with his father long enough for her to run away. And at this point, having accomplished what he had set out to do, he could have run away, too. He would have had no reason to stay and fight. If his motivation was really to protect his sister, and not to injure his father, his task would have been quite 152 simple. But, Johnny remained to fight. Johnny had other, completely nonviolent alternatives if his goal was really to protect his sister. This incident did not happen in isolation, it happened in an apartment house on a hot summer night, with all the windows open. And Johnny could easily have summoned help from the neighbors. But he didn't even try. Five of the Marco's immediate neigh- bors have testified in this trial, and not one has mentioned hearing any loud voices from the Marco's apartment before they heard the screams of the dying Frank Marco. Johnny did not choose the only alternative open to him; he chose the most violent, and the most senseless way he could of stopping an argument between his sister. The defense asks you to excuse him because there was provocation, but was there sufficient provocation? No. Clearly, neither Johnny's life nor that of his sister was threatened. Nor was it shown that his father was a cruel man who tortured his son. It was hardly a matter of self defense. Nor was it an accident. No, ladies and gentlemen, there was some provocation but nothing of a severity sufficient to excuse a boy for killing his own father. I ask for a verdict of guilty of second degree murder. APPENDIX C Experimental Manipulations of the Independent Variables 153 (Cells 1 and 2) Self, negative avowal Ignore condition Johnny: Well, things were not that good between my father and me. We had some rough times together. He really wasn't home that much, and neither was I. I spent most of my time with the Raven's street gang. We pulled the robbery of the neighborhood grocery store last year. Judge: The jury is instructed to ignore Johnny's last statement. Johnny, please confine your answers to the questions that are asked. Defense: Johnny, how long have you lived at the 1405 East 103£d Stteet address? Johnny: For the last eight or nine years. (trial continues) 154 (Cells 3 and 4) Self, negative avowal Not ignore condition Johnny: Well, things were not that good between my father and me. We had some rough times together. He really wasn't home that much, and neither was I. I spent most of my time with the Raven's street gang. We pulled the robbery of the neighborhood grocery store last year. Defense: Johnny, how long have you lived at the 1405 103£d_Street address? Johnny: For the last eight or nine years. (trial continues) Johnny: Judge: Defense: Johnny: 155 (Cells 5 and 6) Self positive avowal Ignore condition Well, things were not that good between my father and me. We had some rough times together. He really wasn't home that much, and neither was I. I spent most of my time with my Boy Scout troop. We organized the neighborhood alley clean-up last year. The jury is instructed to ignore Johnny's last statement. Johnny, please confine your answers to the questions that are asked. Johnny, how long have you lived at the 1405 103£d Street address? For the last eight or nine years. (trial continues) Johnny: Defense: Johnny: 156 (Cells 7 and 8) Self positive avowal Not ignore condition Well, things were not that good between my father and me. We had some rough times together. He really wasn't home that much, and neither was I. I spent most of my time with my Boy Scout troop. We organized the neighborhood alley clean-up last year. Johnny, how long have you lived at the 1405 East 103£d Street address? For the last eight or nine years. (trial continues) Johnny: Defense: Johnny: 157 (Cells 9 and 10) Self neutral avowal Not ignore condition Well, things were not that good between my father and me. We had some rough times together. He really wasn't home that much, and neigher was I. Johnny, how long have you lived at the 1405 103£d Street address? For the last eight or nine years. (trial continues) Robert: Defense: Robert: 158 (Cell 11) Other negative avowal Not ignore condition Well, things were not that good between Johnny and his father. They had some rough times together. His father really wasn't home that much, and neither was he. Johnny spent most of his time with the Raven's street gang. They pulled the robbery of the neighborhood grocery store last year. Robert, how long have you lived at the 1405 103£d Street address? For the last eight or nine years. (trial continues) Robert: Defense: Robert: 159 (Cell 12) Other positive avowal Not ignore condition Well, things were not that good between Johnny and his father. They had some rough times together. His father really wasn't home that much, and neither was he. Johnny spent most of his time with his Boy Scout trOOp. They organized the neighbor- hood alley clean—up last year. How long have you lived at the 1405 East 103£d address? For the last eight or nine years. (trial continues) Robert: Defense Robert: 160 (Cell 13) Other neutral avowal Not ignore condition Well, things were not that good between Johnny and his father. They had some rough times together. His father really wasn't home that much, and neither was he. How long have you lived at the 1405 East 103£d Street address? For the last eight or nine years. (trial continues) cases, 161 (Cells 1, 3, 5, 7, 9, ll, 12 and 13) Nullification condition After the prosecution and the defense rest their the judge instructed the jury as follows" As to the question of guilt, it has already been established that the defendant killed his father. It is up to you as a jury member to decide whether or not he is guilty of second-degree murder. To help you make your decision, we repeat here the charges against the defendant: 'It is charged that the defendant willfully and deliberately stabbed his father, causing him to die, although the act of murder was not premeditated or planned, and that the act was not committed under sufficient mitigat- ing circumstances to relieve him of criminal respon- sibility before the law.‘ You may find him guilty of this offense or of the lesser offense of manslaughter, or may find him not guilty as charged. You have been instructed, during the course of this trial, as to the applicable law in this case. You have also been advised that if Johnny is legally guilty of killing his father, he must be found guilty of either second degree murder or manslaughter. 162 However, it is in your power as a jury to go against the instructions given to you in this case, if you deem it necessary in order to best serve the interests of justice. Your delibera- tions are confidential and your conclusions cannot be challenged. cases , 163 (Cells 2, 4, 6, 8 and 10) No nullification condition After the prosecution and the defense rest their the judge instructed the jury as follows: As to the question of guilt, it has already been established that the defendant killed his father. It is up to you as a jury member to decide whether or not he is guilty of second—degree murder. To help you make your decision, we repeat here the charges against the defendant: 'It is charged that the defendant willfully and deliberately stabbed his father, causing him to die, although the act of murder was not premeditated or planned, and that the act was not committed under sufficient mitigating circumstances to relieve him of criminal respon- sibility before the law.‘ You may find him guilty of this offense or of the lesser offense of manslaughter, or may find him not guilty as charged. You have been instructed, during the course of this trial, as to the applicable law in this case. You have also been advised that if Johnny is legally guilty of killing his father, he must be found 164 guilty of either second degree murder or man- slaughter. Your deliberations are confidential and your conclusions cannot be challenged. APPENDIX D Means and Standard Deviations of the Variables in the Study Table D.1. Means and Standard Deviations of the variables in the Study. variable N=260 N:108 N:72 X SD ‘X’ SD 3? SD Sex .46 .50 Year of Birth 52.22 3.70 Family Size 2.95 1.98 Birth Order .76 .52 SES 1.41 .88 Father Education 4.75 1.85 Mother Education 4.41 1.53 Accident 47.22 30.83 49.53 31.69 48.82 31.98 Responsibility Defendant 56.99 29.88 55.16 32.22 55.46 33.34 Responsibility Sister 12.06 18.15 11.92 18.91 10.99 18.90 Responsibility Father 60.24 28.84 59.09 29.99 58.25 31.21 Guilt 62.60 34.22 62.03 35.99 62.12 36.33 2nd Degree Murder .15 .35 .13 .34 .14 .35 Manslaughter .70 .46 .69 .46 .65 .48 NOt Guilty .15 .36 Sentence 5.55 5.96 5.40 6.34 5.80 7.09 Desire to Meet 4.06 2.28 4.33 2.36 4.36 2.47 Closeness 2.99 1.75 3.18 1.79 3.32 1.84 Admiration 2.36 1.54 2.50 1.59 2.51 1.61 Potential Liking 3.65 1.49 3.70 1.39 3.75 1.36 Initial Liking 3.28 1.50 3.40 1.48 3.44 1.46 Similarity 2.44 1.73 2.56 1.85 2.49 1.85 Table D.l (cont'd.) Positive Evaluation Negative Evaluation Sympathy HOnesty Perceived Positive- ness of Content Perceived Negative- ness of Content Perceived Ingratiation Perceived Openness Honesty Similarity Responsibility Guilt 2nd Degree Murder verdict NOt Guilty verdict Sentence 166 x SD 3.15 .87 34.17 30.54 3.43 .96 60.45 24.21 5.30 1.40 5.53 1.39 50.65 31.39 4.34 1.40 N:8O x SD 61.94 24.55 2.64 1.98 56.47 32.64 60.54 36.47 .14 .35 .21 .41 5.85 6.89 2’ SD 3.10 .91 31.63 29.80 3.30 1.02 61.28 24.23 5.66 1.27 5.23 1.35 50.34 32.58 4.33 1.41 N=120 2' SD 61.53 24.70 2.68 1.97 55.69 32.08 60.74 36.40 .14 .35 .19 .40 5.46 6.22 XI 3.25 33.56 3.30 60.62 5.56 SD .84 32.03 1.04 24.96 1.17 1.33 34.53 1.31 APPENDIX E Additional Intercorrelations Among Variables Amaxv ooflxuo ooo.H was. mam. mom. vmw. woo.u mac. NAH.- mam. 40H. smm.u men. mam.- BUSH ooo.a mum. mom. Hoe. msm.u moH.- oov.- oam. mos. oam.- mam. vsm.- Amaxc ooflan Hoaucmuom ooo.H maw. mmm. mmo.- omm.- Hmm.u oma. omo. Hum.n umm. aom.- Aaaxv coflumuflEOa ooo.H moo. omo.u omH.- omm.- oom. Hmo.u Hmv.u How. Hsm.- loaxc mnmomnoao ooo.a owe. «Na.- smo.u Hma. mmo. mmm.- omm. oom.- onv poo: ou mHHme ooo.H Hmm.u cam. uao.u omo. Hmo.- mma. omH.- Amxv ./ Hounmsmamqflz ”m ooo.H 5mm. omm.- soa.- ooo. mmv.- oao. Asxv noongz common tam ooo.H smm.u mwfi.u msv. mom.u new. onv uafloo ooo.H mHo. on¢.- Hom. mom.- Amxv nonuom suflfiuoflmooonmm ooo.H oom. «mo.u m4o.- 14x1 nounflm suuaflounooommm ooo.H mfiv.u mow. Amxv poooomuoo 333884.91 ooo.H mov.n Amxv uooofloom ooo.H laxv mocouoom max max flax oax ox ox Ax ox mx 4x mx Nx fix .Amhuzv >Uflaflnflmsommmm mo coflustuuu¢.mo mHODMOHOOH paw OQHxHq mo mHODMOHUCH cmmzpmm mGOHumamHHOU .H.m OHQOB ooo.m mmm. mum. moo. moo. mmo.- oom.- mom.- omo. omo. mom.u mom. mom.u Ammxo ocmxmq Homomcm ooo.m oom. mom. omo. omm.u omm.- mmo.- mmm. moo. oom.u oom. mom.- Ammxo ocmxmu HMflgom ooo.m mom. mom. mmo.- oom.- omm.- mmm. omo.- mmm.u omm. mmm.u Ammxo aomumumatd ooo.m ooo. omo.u omm.- mmm.- oom. omo.- mmm.n omm. oom.u lomxo nonconomo ooo.m omm. mum.- moo.- mom. moo. omm.- mmm. mom.- Ammo you: on oumnoo ooo.m mom.- mmm. moo. moo. mmo. omm. mom.- Ammo “mummzmamcmz 8 ”m ooo.m omm. oom.- mmm.- omm. mom.- mom. Ammo noonoz oommoo tam ooo.m mmm.u omo.- moo. mmm.- mmm. ono ommoo ooo.m moo. mom.- oom. mom.n Ammo mocoom fiflmflncoonmm ooo.m oom. omo.- omo. Ammo moonmo 3:34.8QO ooo.m mmm.- mmo. Ammo pooocoooo mummmomnooonmm ooo.m mmm.; Ammo homomoom So; :8 888m mmx mmx mmx omx mm mm mm om mm ox mm mm mm . loomnzo mommmomnooonom mo compsomnuu< mo mHOpwomosH pow mcmxoq mo mqumomwcH coozpom maoammaouuou .N.m OHQmE 169 Table E.3. Intercorrelations of Intervening variables for Four Cell and SiXeCell Designs. X1 X2 x3 X4 X5 HOnesty (X1) 1.000 .099 .155 .197 .232 Negative labeling (X2) .099 1.000 .121 .212 -.052 Similarity (X3) .155 .121 1.000 .267 .134 Positive labeling (X4) .197 .212 .267 1.000 .428 Sympathy (x5) .232 -.052 .134 .428 1.000 Four-Cell Design (N=72) V X X X X Al 2 3 4 5 Honesty (x1) 1.000 .052 .163 .145 .188 Negative labeling (x2) .052 1.000 .116 .215 .050 Similarity (x3) .163 .116 1.000 .160 .069 positive labeling (x4) .145 .215 .160 1.000 .383 Sympathy (x5) .188 .050 .069 .383 1.000 Six-Cell Design (N=108) APPENDIX E Canonical Weights and Canonical Correlations Between Indicators of Intervening Variables and Indicators of Liking 170 Table F.l. Canonical weights and Canonical Correlations Between Indicators of Intervening variables and Indicators of Liking.a Eigenvalue Canonical (variance 2 Degrees of Signif- Design Correlation explained) x Freedom icance Four Cells (N=72) .682 .466 55.15 25 .000 Six Cells (NEJOB) .676 .458 82.17 25 .000 Canonical weights Intervening variables FOur Cells (N:72) Six Cells (N=108) HOnesty .517 .578 Similarity .599 .580 Sympathy .084 —.059 Negative Evaluation .088 .058 Positive Evaluation .240 .331 Indicators of Liking Desire to Meet .298 .196 Closeness .069 .170 Armuration .491 .471 Potential Liking .303 .398 Initial Liking .241 .123 aFirst canonical variate only. REFERENCES 171 REFERENCES Adams, J. 1965 "Inequity in social exchange." Pp. 267-300 in L. Berkowitz (ed.), Advances in Experimental Social Psychology, Vol. 2. New York: Academic Press. Anderson, N. 1974 “Cognitive algebra: Integration theory applied to social attribution." Pp. l-lOl in L. Berkowitz (ed.), Advances in Experimental Social Psychology, Vol. 7. New York: Academic Press. Archibald, W., and R. Cohen 1971 "Self-presentation, embarrassment, and facework as a function of self-evaluation, conditions of self- presentation, and feedback from others." Journal of Personality and Social Psychology 20:287-297. Berscheid, E.,and E. Walster 1969 Interpersonal Attraction. Reading: Addison-Wesley. Blau, P. 1967 Exchange and Power in Social Life. New York: Wiley. Brown, B., and H. Garland 1971 "The effects of incompetency, audience acquaintance— Ship, and anticipated evaluative feedback on face- saving behavior." Journal of Experimental Social Psychology 7:490—502. Cozby, P. 1972 "Self disclosure, reciprocity and liking." Socio- metry 35:151—160. Derlega, v., M. Harris, and A. Chaikin 1973 "Self-disclosure, reciprocity, liking and the deviant." Journal of Experimental Social Psychol- ogy 9:277-284. 172 Ekman, P., and W. Friesen 1969 "Nonverbal leakage and clues to deception." Psychiatry 63:88-106. Fishbein, M., and I. Ajzen 1973 "Attribution of responsibility: A theoretical note." Journal of Experimental Social Psychology 9:148-153. Goffman, E. 1957 The Presentation of Self in Everyday Life. New York: Doubleday. Hastorf, A., D. Schneider, and J. Polefka 1970 Person Perception. Reading, Mass.: Addison- Wesley Heider, F. 1958 The Psychology of Interpersonal Relations. New York: Wiley. Homans, G. 1961 Social Behavior: Its Elementary Forms. New York: Harcourt, Brace, and World. Joe, v. 1971 "Review of the internal-external control construct as a personality variable." Psychological Reports 28:619-640. Jones, E. 1964 Ingratiation. New York: Appleton-Century-Crofts. Jones, E., and K. Davis 1965 "From acts to dispositions: The attribution process in person perception." Pp. 219—264 in L. Berkowitz (ed.), Advances in Experimental Social Psychology, Vol. 2. New York: Academic Press. Jones, E., K. Davis, and K. Gergen 1961 "Role playing variations and their informational value for person perception." Journal of Abnormal and Social Psychology 63:302-310. Jones, E., D. Kanouse, H. Kelley, R. Nisbett, S. Valins, 1972 and B. Weiner Attribution: Perceiving the Causes of Behavior. Morristown, N.J.: General Learning Press. JOreskog, K. 1970 "A general method for analysis of covariance structures." Biometrika 57:239-251. 173 Jourard, S. 1971 Self-Disclosure: An Experimental Analysis of the Transparent Self. New York: Wiley. Kalven, H., Jr., and H. Zeisel 1966 The American Jury. Boston: Little—Brown. Kelley, H. 1967 "Attribution theory in social psychology." Pp. 192—238 in D. Levine (ed.), Nebraska Symposium on Motivation. Lincoln: University of Nebraska Press. Landy, D., and E. Aronson 1969 "The influence of the character of the criminal and his victim on the decision of simulated jurors." Journal of Experimental Social Psychology 5:141- 152. Legant, P. 1973 "Equity theory and the law: Suggestions for future research." Paper presented at the American Psycho— logical Association Convention, Montreal. Macaulay, S., and E. Walster 1971 "Legal structures and restoring equity." Journal of Social Issue 27:173—188. Miller, G., and F. Siebert 1974 "Effects of videotaped testimony on information processing and decision-making in jury trials." Progress Report 1, Department of Communication, Michigan State University. Mills, C. 1970 "Situated actions and vocabularies of motive." Pp. 472—480 in Stone, G., and H. Faberman (eds.), Social Psychology through Symbolic Interaction. Waltham, Mass: Ginn—Blaisdell. Mitchell, H., and D. Byrne 1973 "The defendant's dilemma: Effects of jurors' attitudes and authoritarianism on judicial decisions." Journal of Personality and Social Psychology 25:123- 129. Mysliwiec, S. 1974 "Toward principles of jury equity." Yale Law Journal 83:1023-1054. Nagel, S. 1969 The Legal Process from a Behavioral Perspective. Homewood, Ill.: Dorsey Press. 174 Nemeth, C., and R. Sosis 1973 "A simulated jury study: Characteristics of the defendant and the jurors." Journal of Social Psychology 90:221-229. Newcomb, T. 1961 The Acquaintance Process. New York: Holt Rine— hart and Winston. Schneider, D. 1969 "Tactical self-presentation after success and failure." Journal of Personality and Social Psychology 13:262—268. Scott, M., and S. Lyman 1968 "Accounts." American Sociological Review 33:46—62. Sears, D. 1965 "Biased indoctrination and selectivity of exposure to new information." Sociometry 28:363—376. Shaver, K. 1970 "Defensive attribution: Effects of severity and relevance on the responsibility assigned for an accident." Journal of Personality and Social Psychology 14:101-118. Sigall, H., and I. Landy 1972 "Effects of the defendant's character and suffer- ing on juridic judgment: A replication and clari- fication." Journal of Social Psychology 88:149-150. Simmons, J. 1969 Deviants. Berkeley, Cal: The Glendessary Press. Stephan, C. 1973 "Sex prejudice in jury simulation." Paper pre- sented at the Annual Meeting of the American Sociological Association, New York. Sykes, G., and D. Matza 1957 "Techniques of neutralization: A theory of de- linquency." American Sociological Review 22:664-670. Thibaut, J., L. Walker, and L. Allan 1972 "Adversary presentation and bias in legal decision making." Harvard Law Review 86:386-401. Turner, R. 1972 "Deviance avowal as neutralization of commitment." Social Problems 19:308-321. 175 Van de Geer, J. 1971 Introduction to Multivariate Analysis for the Social Sciences. San Francisco: W. H. Freeman and Co. Walster, E. 1966 "Assignment of responsibility for an accident." Journal of Personality and Social Psychology 3:73-79. Walster, E., E. Berscheid, and G. Walster 1970 "The exploited: Justice or justification?" Pp. 173-204 in J. Macauley and L. Berkowitz (eds.), Altruism and Helping Behavior. New York: Academic Press. Walster, E., E. Aronson, and D. Abrahams 1966 "On increasing the persuasiveness of a low pres- tige communicatior." Journal of Experimental Social Psychology 2:325-342. Weinstein, E. 1966 "Toward a theory of interpersonal tactics." Pp. 394-398 in C. Backman and P. Secord (eds.), Problems in Social Psychology. New York: McGraw Hill. Weinstein, E., and P. Deutscheberger 1964 "Tasks, bargains, and identities in social inter- action." Social Forces 42:451-456. Weinstein, E., L. Beckhouse, P. Blumstein, and R. Stein 1968 "Interpersonal strategies under conditions of gain or loss." Journal of Personality 36:616-634. Worthy, W., A. Gary, and G. Kahn 1969 "Self-disclosure as an exchange process." Journal of Personality and Social Psychology 13:59-63. Zillmann, D., and J. Cantor 1974 "Rhetorical elicitation of concession in per- suasion." Journal of Social Psychology 94:223- 236. M'H1111)11111))»flfljlflffljl'lfljfliflfiJ’I‘ITI'ES