AK ANALYSIS 0? PRDXEHICS AND 'SMQMSCEMENG BEHAWGR 0F REUDMST MB NON~RE€EBMST ADULT SOCIAL GF'FENDERS FROM BLACK. CHICANO AND WERE [RENE POPULAHONS . A Dburtahon ' for flu Dag M: of pit. D MICHIGAN STATE UNIVERSITY Ronaid L. Braithwaite I974 ———*i willwilllugllllgllllgllul ‘ LIBRAR Y Mich? Qtate V ”f? "Kiri THES‘S This is to certify that the . 5’ ‘ thesis entitled An Analysis of Proxemicfi Sell-Vt" Disclosing Behavior of Recidivist and Non- Recidivist Adult Social Offenders From Black, Chicano and, White Inmate Populations presented by Ronald L . Brai thwaite has been accepted towards fulfillment of the requirements for Ph. D. (1H,“... in Education MM Major professor Date Jail: £52,. 1974- ABSTRACT AN ANALYSIS OF PROXEMICS AND SELF-DISCLOSING BEHAVIOR OF RECIDIVIST AND NON-RECIDIVIST ADULT SOCIAL OFFENDERS FROM BLACK, CHICANO AND WHITE INMATE POPULATIONS BY Ronald L. Braithwaite The basic purpose of this study was to investigate self-reported differences in self-disclosure (SD) and proxemic distance parameters of Black (B), White (W), and Chicano (C) inmates; and differences between recidivist (R) and non-recidivist (NR) inmates from each of these ethnic groups. It is generally accepted in the literature of counseling and interpersonal processes that self-disclosure (SD) is one of the most important and desirable process variables. Self- disclosure apparently functions to enhance client growth, insight and behavioral change. Self-disclosure has been treated as a process variable having many conditional contingencies which may all serve as factors affecting the timing, amount, kind, motive and (A ,{9 (Om (é selectivity of personal disclosures. Jourard and his Ronald L. Braithwaite associates (1958, 1964) have postulated and supported the general notion that openness and transparency via self—disclosure to significant others are a necessary condition for the actualization of healthy personality. The basic design of this study was a survey and quasi—experimental "posttest only" design recommended for educational research by Campbell and Stanley (1963). All subjects were male social offenders incarcerated at the State Prison of Southern Michigan, and were selected from a computer-generated printout which indicated a grade reading level of 4.9 or above for the subjects. Each subjeCt in the study was randomly selected from his race group with a probability of inclusion proportional to total race group size. The sample consisted of 15 Black non—recidivist (NR's), 15 Black recidivists (R's), 15 White NR's, 15 White R's, 10 Chicano NR's and 10 Chicano R's. The three dependent measures used in this evaluation were: (3) self-reported scores on a modified version of the Jourard Self-disclosure Questionnaire (SDI); (b) be- havioral measures of proxemic distance and area need parameters; and (g) the Paranoia sub-scale score on Ronald L. Braithwaite the Minnesota Multiphasic Personality Inventory (MMPI). The data were collected over a four-week period. All §§ were administered the SDI and the proxemic test within a one-hour time frame. The investigator was cognizant of the general climate of suspicion and distrust among inmates; hence, an atmosphere of good will and rapport was initiated by the investigator in an effort to decrease faking and response set. The SDI sought to identify differentials among and within levels of race and NR-R variables for total SD scores, as well as sub-scores for the following seven disclosure targets (target persons): closest inmate friend, work supervisor, counselor or psychologist, correctional staff, family and friends on the outside, any inmate and education staff. The SDI requested the inmates to respond to 21 personal statements concerning how comfortable they felt or would feel about discussing or revealing the topic of the statement with the seven target persons. Four response options ranging from "would lie," "would not discuss," "would discuss in general terms only," to "would discuss in full and complete detail," were presented for each statement. Ronald L. Braithwaite The basis hypothesis tested in this study focused on main effects and two- and three-way interactions. The hypotheses predicted that significant differences would be observed among the three race groups and between NR-R groups. Repeated measures analysis of SD for targets crossed with race indicated significant differences. The differences were graphically displayed to further depict the patterns of SD. A hypothesized significant negative correlation between total SD and the Paranoia sub-scale score of the MMPI was not supported by the results for any of the race groups. The investigation also resulted in statistically significant differences among the races in proxemic area. Scheffe post hoc tests indicated that the differences were attributable to differences between the Chicano and White inmate groups. A hypothesized significant negative correlation between total SD score and proxemic area for race groups was supported only by the data derived from the White group. The data strongly supported, for all gs, the anticipated greater need for rear proxemic distance as compared to frontal distance. A multivariate and univariate analysis of variance and covariance program developed by Finn (1967) was used to compute and analyze the data. An a priori alpha level was set at .05. AN ANALYSIS OF PROXEMICS AND SELF-DISCLOSING BEHAVIOR OF RECIDIVIST AND NON-RECIDIVIST ADULT SOCIAL OFFENDERS FROM BLACK, CHICANO AND WHITE INMATE POPULATIONS BY (A Ronald LD/Braithwaite A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Personnel Services and Educational Psychology 1974 DEDICATION To the inmates at Jackson State Prison, and all political prisoners ACKNOWLEDGMENTS Many peeple assisted me in many ways during my educational struggle which culminated with this dissertation. To all of them I am deeply grateful. I wish to extend particular appreciation to my understanding friend and chairman, Dr. Bob B. Winborn and to the members of my committee, Dr. Thomas S. Gunnings, Dr. Herbert M. Burks, Jr. and Dr. Fred Pesetsky all of whom assisted in the formulation and guidance of this research exploration. I am greatly thankful to Dr. Kowit Pravalpruk and Dr. Maryellen McSweeney for their input relative to the measurement and analysis concerns of this study. I appreciate the support of the several persons who enhanced the logistics of this data gathering effort. To Dr. Barry Mintzes and other officials at the State Prison of Southern Michigan I am thankful. This research document was made possible by an H.E.W. Grant No. ITOl MH18174-01, Masters and Ph.D. Training Program for Counselors of Urban Residents. iii I am thankful for the support from the National Institute of Mental Health. The assistance received from my dearest friends and colleagues: Michael Martin, Cliff Sanders, Calvin Matthews, Gayle King, Debra Walls, and Yvonne Carter, and Gloria Simmons; I love you all. To my family from which the root of my strength grows. To my mother, Mabel, family and relatives in New York; to my in-laws, The Honorable Ernest C. Moore and Mittie Moore, and family I am deeply grateful. I was fortunate to have Jacqueline Humphrey, who worked for hours typing the final manuscript, thanks to you. Finally, to my wonderful wife, Judy, and beautiful daughters, Malika and Kisha, who suffered during this rigorous experience. No words can express my feeling or thoughts for you. iv TABLE OF CONTENTS Page DEDICATION o o o o o o o o o o o o o o o o o o o o i 1 ACKNOWLEDGMENTS O 0 0 O O O O O O O O O O 0 O O O i i i LIST OF TABLES . . . . . . . . . . . . . . . . . . Viii LIST OF FIGURES . . . . . . . . . . . . . . . . . x LIST OF APPENDICES . . . . . . . . . . . . . . . . Vii Chapter I. STATEMENT OF THE PROBLEM . . . . . . . . . . 1 Purpose of Study . . . . . . . . . . . . . 6 Need For the Study 0 O O O O O O O O O O O 8 Delimitation of Study . . . . . . . . . . 10 Definition of Terms . . . . . . . . . . . ll Testable Hypotheses . . . . . . . . . . . 12 Overview . . . . . . . . . . . . . . . . . . 13 I I I RELATED LITERATURE O O O O O O O O O O O O O 1 5 ove rView O O . O O O O O O O O O O O O O O O l 5 Healthy Personality . . . . . . . . . . . 15 Self-Disclosure . . . . . . . . . . . . . 21 Subject Matter Differences . . . . . . . 24 Positive Mental Health . . . . . . . . . 27 Interpersonal Trust . . . . . . . . . . 32 Social Exchange Theory . . . . . . . . . 38 Social Penetration . . . . . . . . . . . 41 Chapter Page Early vs. Late . . . . . . . . . . . . . 43 Sensitivity Training and Cohesiveness . . . . . . . . . . 44 Manipulation . . . . . . . . . . . 47 Inmates . . . . . . . . . . . . . . . 57 Validity . . . . . . . . . . . . . . . 60 Proxemics . . . . . . . . . . . . . . . . 67 III. DESIGN AND METHODOLOGY . . . . . . . . . . . 76 overView O O O O O O O O O O O O O O O O 76 Population . . . . . . . . . . . . . . . . 76 subjects 0 O O O O O O O O O O O O O O O O 79 Procedure . . . . . . . . . . . . . . . . 81 Instruments . . . . . . . . . . . . . . . 84 8-D Inventory . . . . . . . . . . . . 84 PilOt Study 0 O O O O O O O O O O O O O 87 Proxemic Indicator . . . . . . . . . . . 88 MMPI O O O O O O O 0 O 0 O O O O O 0 O 88 Design . . . . . . . . . . . . . . . . . . 91 Analysis of Data and Decision Model . . . 91 IV. ANALYSIS OF RESULTS . . . . . . . . . . . . . 93 Results of the Test of Hypothesis I . . . 94 Results of the Test of Hypothesis II . . . 95 Results of the Test of Hypothesis III . . 96 Results of the Test of Hypothesis IV . . . 98 Results of the Test of Hypothesis V . . . 104 Results of the Test of Hypothesis VI . . . 104 Closest Inmate Friend Target . . . . . . 108 Correctional Staff Target . . . . . . . 108 Work Supervisor Target . . . . . . . . . 109 Family and Friends on Outside Target . . 109 Any Inmate Target . . . . . . . . . . . 110 Educational Staff Target . . . . . . . . 110 Results of the Test of Hypothesis VII . . 113 Results of the Test of Hypothesis VIII . . 114 Results of the Test of Hypothesis Ix & X . 115 Results of the Test of Hypothesis XI . . . 118 vi Chapter Page V. SUMMARY, DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS . . . . . . . . . . . 121 Summary . . . . . . . . . . . . . . . . . 121 Discussion . . . . . . . . . . . . . . . 125 Main Effects . . . . . . . . . . . . . 125 Sample . . . . . . . . . . . . . . . . 125 Instrumentation . . . . . . . . . . . . 126 Interaction Effects . . . . . . . . . . (127 Effects of Proxemics Distance . . . . . 131 Conclusions . . . . . . . . . . . . . . . 133 Recommendations . . . . . . . . . . . . . 134 Program Development . . . . . . . . . . 134 Systemic Intervention . . . . . . . . . ResearCh O O O O O O O O O O O O O O 0 LIST OF REFERENCES . . . . . . . . . . . . . . . 137 APPENDICES . . . . . . . . . . . . . . . . . . . 151 Appendix A. Breakdown of Personal Characteristics for Inmates Committed in 1972 (state-wide). . 152 B. Biographical Information for Sampled SUbjeCtS O O I O I O O O O O O O O O O O 158 C. Self-disclosure Inventory . . . . . . . . . 163 D. Job Descriptions for State Department of Civil Service - Related Target Persons (Department of Corrections) . . . . . . . 171 Vii Table 3.1. 3.2 4.6 4.7 4.8 4.9 LIST OF TABLES Probability Sampling Ratios for Jackson Prison Inmates . . . . . . . . . . . . . Internal Consistency Estimates (in Split- Half Coefficients) for MMPI . . . . . Univariate Tests for Differences between Group Means an Self-Disclosure Score . . Cell Means and Standard Deviations for Black, White and Chicano Race Groups by Recidivist, Non-Recidivist Groups on Total Self-Disclosure Scores . . . . Multivariate and Univariate Tests for Race Crossed with Target Interaction . . . . . . . . . . . . . Univariate Tests for Differences Among Race within Targets . . . . . . . Cell Means and Standard Deviations for Race Within Target on Self- DiSClose O O O O O O O O O O O O O O Multivariate Interaction Test for Non- Recidivist-Recidivist Crossed with Targets Multivariate Test for Three-Way Interaction of Race by Non—Recidivist- Recidivist by Targets . . . . . . . . . Univariate Two-Way Analysis of Variance for Self-Disclosure to Each Target Person (Race by Non-Recidivist- Recidivist). . . . . . . . . . . . . . . Cell Means and Standard Deviations for Black, White and Chicano Groups by Re- cidivist, Non-Recidivist Groups on Self-Disclosure to Targets . . . . . . . viii Page 80 90 95 98 99 100 103 104 105 106 111 Table Page 4.10 Correlation of Race Groups Total Self-Disclosure Score with Proxemic Area . . . . . . . . . . . . . . 114 4.11 Correlation of Subjects and Race Group Total Self-Disclosure Score with MMPI Paranoia Scale . . . . . . . . 115 4.12 Univariate Tests for Differences between Race, Recidivist, Non- Recidivist and Interaction on Total Proxemic Area . . . . . . . . . . . 116 4.13 Scheffe Post Hoc Analysis for Race Group Mean Differences on Total Proxemic Area . . . . . . . . . . . 117 4.14 Univariate Test of Differences for all Subjects and Interactions for Groups on Rear vs. Front Proxemic Distance................ 119 4.15 Cell Means and Standard Deviations for Black, White, and Chicano Groups by Recidivist, Non-Recidivist on Total Proxemic Area, Rear Proxemic Distance, and Front Proxemic Distance . . . . . . . . . . . . . . . . 120 ix LIST OF FIGURES Figure Page 3.1 Design Model . . . . . . . . . . . . . . . 91 4.1 Graph of Interaction Effects of Recidivist, Non-Recidivist by Race Groups on Total Self-Disclosure Score . . . . . . . . . 97 4.2 Graph of Cell Means for Race Within Targets on Self-Disclosure . . . . . . . 101 4.3 Graph of Cell Means on Self-Disclosure for Significant Interaction Effects (Race by Recidivist, Non-Recidivist) to Target Persons . . . . . . . . . . . 107 4.4 Graph of an Overhead View of Reported Proxemic Area for Black, White, and Chicano Groups . . . . . . . . . . . 118 CHAPTER I THE PROBLEM "If we cannot live as peOple we will at least try to die like men." Charles Horatio Crowley Crisis, confrontation and conflict between adminis- trative personnel of correctional insitutions and social offenders has characterized many penal institutions within the United States during the past five years. The quotation above was echoed by a Black inmate serving time in a New York correctional facility. The tragedy at the Attica Correctional Facility in New York State, in which 42 men died, has undoubtedly shocked the American public from coast to coast. Furthermore, these events generate questions of intense controversy. What factors precipitated the conflict? Are prison conditions as grim and inhumane as prisoners alleged? Illinois, California, Maryland, other states, and Washington, D.C. have witnessed turmoil and protest within their correctional facilities. Discrepancies in conditions of prison life and casual factors of conflict between inmates and correctional staffs have earmarked this decade as one of decay of the American correctional system. Acknowledg- ment of the prevailing concerns dictates the need for behavioral research of the prison world, coupled with an ideology of rehabilitation and objectives for change. The observation that inmates become more criminally sephisti- cated by virtue of their prison experiences must be con- fronted by the opportunity for results-oriented treatment programs. Communication and rapport between two or more persons or groups is an essential element in the development of an effective rehabilitation process. Given the nature and function of prisons, reported perceptual and causal dis- crepancies by inmates and correctional personnel may generate insights relative to the interpersonal trust and dynamics within the prison subculture. The need for sharing one's anxieties, conflicts, concerns, and emotions in counseling and psychotherapy is an essential aspect in the modification of behavior as it relates to clients and client systems. Information received by a therapist or consultant elevates the capacity of that person to enhance or restrict growth. Effective inter- personal communication, however, has been a difficult process for many individuals, groups and organizations. Suspicion and distrust which characterize the phenomeno- IOgical field of many individuals and groups, have aided and abetted breakdowns which stymied Open and authentic interpersonal communication. Jourard's (1964) work on the process of self— disclosure suggests that verbalizing one's self and 'one's problem to another individual is a way of experi- encing release from the tensions produced by the disciplines of the roles we play in everyday life. Freud viewed free association and catharsis as a process resulting in tension reduction. Horney's (1950) description of the neurotic personality postulates that often clients do not know themselves, and hence, the self-alienated become neurotic. Thus, a prerequisite for authentic self-disclosure appears to encompass a searching for the real self. Authenticity and self-disclosure can be healthy or unhealthy behavior depending on the situation in which it occurs. Hence, selectivity of disclosure is an area of concern. Shall we permit others to know us as we really are or shall we continue to camouflage and masquerade behind the mask that so many others wear? Jourard (1964) states that "disclosure begets disclosure," and there is a mutual building of interpersonal trust. The problem is that to trust, often involves taking a risk, and risk taking can have negative as well as positive payoffs especially within the confines of a prison. The manner in which justice personnel present them- selves and their agencies to the outside world is usually characterized by a quiet, personal secrecy. A tight control on personal feelings shields personnel in the criminal justice system and often provides themes for ideal behavior. The good professional law enforcement officer is "all business" with a client. The parole officer seeks to deal objectively with his parolees. The correctional officer soon learns not to talk about himself with prisoners (Brodsky, 1968b). The secretiveness is of'systems as well as people and serves to close communication. A first incarceration represents an abrupt change in the life of an individual. It is usually involuntary; it is accompanied by relocation and a physical barrier between the person and his previous life style. Others who are aware of the new situation for the incarcerate may sterotype and view him as an outcast. This can be stressful and requires supportive assistance from significant others. The new inmate is often stripped of his personal clothing, assigned a number, and relegated to a position which results in a loss of individuality and/or identity. Brodsky and Komaridis (1968)note that the first few days may produce a state of uprooted bewilderment and transformation of identity. Clemmer (1958) reports that the subsequent "prisonization" process of establishing and maintaining a new identity often promotes assimilation of the anti-social values of the prison culture. Brodsky and Komaridis (1968) aptly suggest that the degree of "prisonization" is influenced by (a) the type of relations maintained with persons external to the institution and, (b) the degree of affiliation with highly 4 "prisonized" groups. Positive relationships with persons external to the institution and minimum affiliation with highly prisonized groups are seen by Clemmer (1958) as positive signs contraindicating development of deep criminal attitudes. Hence, the self-disclosing behavior of the new inmate can be very important in the survival mode he chooses. Sommer (1959, 1961) has used observation of behavior in naturalistic settings, with occasional experimental interventions, to study the effects of spatial position on interpersonal relations and has demonstrated avoidance responses when a person approaches another too closely or with a "wrong" posture. Data on social distance, proxemics or territoriality may prove insightful to correctional planners and architects who are responsible for the physical development of institutional environments. A temporal measure of the magnitude of communication breakdowns, suspicion and distrust is again reflected by the prison riots during the past five years. In January of 1974, Perry Johnson, Director of the Michigan Department of Corrections, on a televised program entitled "Thirty Minutes," (WJIM-TV, Lansing, Michigan) responded to the question: "Does anyone know why Jackson Prison had a riot in 1952, and can it happen again?" His response was, WNo one knows why it happened. Yes, it can happen again and it can happen in any correctional facility in the country." To the question, "Do prison riots do anybody any good?" Johnson responded, "Riots have been good--they bring 5 attention to problems." Johnson further recognized that repressive tactics do more harm than good. The challenge for greater understanding and trust between inmates and correctional administrators is obviously a matter of great concern in prison reform. Massive efforts supported by substantial financing of research and data-gathering are needed if correctional systems are to assist inmates to become contributing members of our society. The time is long overdue for open communications coupled with planning for significant modification of deviant behavior. President Nixon has called prisons "universities for crime." Mitford's book, Kind and Usual Punishment (1973) has termed prisons useless. The effects of the women's liberation movement will certainly filter down to the courts, resulting in more female convictions. With more female convictions, more women will be sentenced, again magnifying the magnitude of the problem. Purposes of the Study The purpose of this research endeavor was to explore and determine to whom inmates of Jackson State Prison relate, trust, and disclose personal information. More specifically, this study investigated the self-disclosing behavior of incarcerated social offenders from different races and levels of time of incarceration. This study also explored group differences in terms of proxemic area. Proxemic area refers to the physical area required before 6 one begins to feel uncomfortable about the encroachment of another person. The specific objectives of the study were as follows: 1. To assess the degree of self-disclosure of recidivist and non-recidivist incarcerated social offenders. To determine if there are significant differences in self-disclosure among Black, Chicano and White incarcerated social offenders. To determine whether there are any significant differences in reported self-disclosure from Black, Chicano and White inmates to the following target person(s): closest inmate friend, work supervisor, counselor or psychologist, family and friends outside the prison, any inmate, correctional staff and education staff. To determine the similarities and differences in proxemic distance between and among levels of the independent variables, i.e., race and times of incarceration. To determine the extent of measures of association and relationship between self-disclosure and the Paranoia subscale of the Minnesota Multiphasic Personality Inventory (MMPI). To ascertain the correlation between total self- disclosure scores and the measure of total proxemic area. Need for the Study Research on self-disclosure has had a relatively brief history. Prior to 1958, the psychological literature did not include any empirical studies using self-disclosure as an independent or dependent variable. Self-disclosure research can now be identified in the literature under the rubric of interpersonal process. The need for the study is more descriptively noted by the following: 1. Empirical and descriptive research done in correctional settings has been almost nil. It is apparent to correctional planners and administrators that there is a need to know more about the behavior of social offenders, as well as the many facets of the interaction between the incarcerated offender and various treatment programs. Thompson and McAnaw (1972) noted that correctional institutions must develop a research data-gathering potential. Presently, these functions are minimized. Information is rarely collected in forms that may be properly and effectively analyzed so as to represent a force in policy making (Thompson & McAnaw, 1972). There is a need for correctional planners to ascertain and identify the kind of interpersonal trust which is transpiring in their institutions. Differential amounts of self—disclosure to 8 sclchcd institutional targets is the kind of data which will be useful in the planning of treatment programs. Correctional administrators need fixed data concerning the population which they serve. Hence, this inquiry utilized inmates as the sampled subjects. Thomas Gunnings, Assistant Dean of Health Programs, College of Human Medicine, at Michigan State University, has delivered numerous lectures on the need for applied research in the correctional setting. His participation in a counselor training program (1972) funded by the National Institute of Mental Health has generated concern and interest relative to systemic intervention in correctional institu- tions. His concern is reflected by some of the unpublished master's degree papers-of several students participating in this counselor training program. The following are titles of papers submitted in partial fulfillment for the master's degree: Therapy in an Anti-Therapeutic Environment: SystemicIntervention and the Prison System (Martin, 1973). AnAlternative Program to Institutional Incarceration for Youthful Offenders (Hayes, I973). Systemic Intervention for Some of the Pertinent Problems in the Correctional Institution (Ward, 1973). Although the field of correctional psychology is in its embryonic stage of development, the following schools are training graduate professionals to work in correctional settings: the University of Alabama, Southern Illinois University, Florida State University, Middle Tennessee State University (Speilberger, Magargee & Ingram, 1972). Michigan State University does not have a formal program. However, the faculty of the College of Urban Development has expressed the need for increased research in correctional settings. Delimitation of Study This research is concerned with a sample of incar- cerated social offenders located at the State Prison of Southern Michigan. Race was a primary independent variable of interest, encompassing three levels: Blacks (B), Whites (W), and Chicanos (C). It is a fixed and qualitative independent variable. The second independent variable for this study was~ Times of Incarceration (two levels). Inmates incarcerated for the first time (non-recidivist) constitute one level of the independent variable. The second level was characterized by inmates who have served more than six months on two or more independent occasions (recidivist) This independent variable was also a fixed and qualitative variable. 10 Most of the subjects are from large, urban areas, which may prove to be a delimiting factor affecting the external validity of the interpretations. Definition of Terms Self-Disclosure (SD) refers to the ability to share, express or characterize to others feelings of concern, fear, joy, anger, and the entire gamut of emotions which are experienced in interpersonal transactions. It . includes verbal behavior which represents authentic and honest self-references communicated with real and genuine intentions to another person. Openness and transparency will be used synonymously with (SD) in this study. Self-Concealment (SC) should be treated and conceptualized as an antonym of self-disclosure. Self-concealment is characterized by a masking of one's real and true feelings or emotions on a given issue. It represents a camouflaging which usually results in the deception of others. "Dyadic Effect" (DE) refers to the relationship in which two persons mutually exchange relevant and personal disclosures. It involves the interpersonal interactions which embrace a qgid 259 qgg dialogue between two or more persons. Target Person refers to a person to whom one may choose to self-disclose or self-conceal. Proxemic Area (PA) refers to the physical area required before a person begins to feel uncomfortable about the 11 encroachment of another person. This is the same as social distance, body buffer zones and territorial boundaries. "Rosenthal Effect" refers to the bias which enters a study as a result of experimenter input i.e., the experimenter knowingly or unknowingly influences responses of experimental subjects. Systemic Counselingis an approach to counseling which views the etiology of client problems as originating within social systems as opposed to the individual client. Testable Hypotheses The following hypotheses were tested during this investigation: Hypothesis 1 -— The Chicano, White and Black groups will differ significantly on total self— disclosure scores. Hypothesis 2 —- The non-recidivist and recidivist groups will differ significantly on total self- disclosure scores. Hypothesis 3 -- The total self-disclosure scores of the non-recidivist and recidivist groups will interact significantly with the total self-disclosure scores of the Chicano. White and Black groups Hypothesis 4 -- Self-disclosure to specific target persons will interact significantly with the Chicano, White, and Black groups. 12 Hypothesis 5 —- Self-disclosure to specific target persons will interact significantly with the recidivist and non-recidivist groups. Hypothesis 6 -- Self-disclosure to specific target persons will interact significantly with recidivist and non-recidivist by Black, White and Chicano groups. Hypothesis 7 —— Total self-disclosure scores will have significant negative correlation with Proxemic Area for the Chicano, Black, and White groups. Hypothesis 8 -- Total self-disclosure scores will have a negative correlation with the Paranoia Scale of the Minnesota Multiphasic Personality Inventory for Black, White, and Chicano groups. Hypothesis 9 —— The Chicano, Black, and White groups will differ significantly on Proxemic Area. Hypothesis 10-- The recidivist and non-recidivist groups will differ significantly on Proxemic Area. Hypothesis ll-- All subjects will require greater rear Proxemic Distance in contrast to frontal Proxemic Distance. Overview The need for the present research and the hypothesis of interest have been described in Chapter I. In Chapter II, the literature pertinent to the study has been reviewed. 13 In Chapter III the population, sampled subjects, procedures, instruments, design and analysis of the decision model are described. A presentation of the results from a pilot study comprise the remainder of Chapter II. Chapter IV contains an analysis of the results. Finally, Chapter V includes a discussion of the results, conclusions, and a summary of the study with implications for further" investigation or contribution to the field of correctional psychology. 14 CHAPTER II RELATED LITERATURE Overview In this chapter the literature pertinent to self— disclosure and proxemics is reviewed. The general plan is to present an encapsulated View of the follow- ing tepics which have been shown to be directly related to self-disclosure: subject matter differ- ences, positive mental health, interpersonal trust, social exchange theory, social penetration, timing of disclosures, sensitivity training and cohesiveness, manipulation, validity studies, and a summary of the few studies which have utilized inmate subjects. The chapter begins with a section which discusses healthy personality, and the final section is devoted to issues related to proxemics. Healthy Personality Many traditional psychotherapeutic approaches have focused efforts on the reduction and/or elimination of mental disorders by encouraging authentic verbalization of the anxieties and stresses which confront the client. Clients will disclose their anxieties and conflicts 15 only to those whom they have reason to trust. With— out trust and good will, a client will often conceal or misrepresent his real self, with the intention of presenting a false image. Jourard (1968) views authenticity and honesty as characteristics surrounding healthy personalities. Less healthy personalities are people who function less than fully, who suffer recurrent breakdowns or chronic impasses, and who may usually be found to be prevaricators. Such persons say things that they do not mean. Their disclosures have been chosen more for cosmetic value than for truth (Jourard, 1968). The healthier personality is further viewed as one having common sense and good judgment. Issues related to defining healthy personality and normalcy are of concern to contemporary psychologists. Szasz (1961) challenges the myth of normality and mental illness which is popularized in our society. White (1973) has charged the medical model of diag- nosis and treatment as being value-laden by mental health professionals. He states: Health, of course is a value, but it is an impersonal, universalistic value which seems in no danger of contamina- tion by personal preferences on the question of how to live. Secure in the medical analogy, we could nudge our patients toward a better way of leading their lives without a qualm that the goal was anything but health. Nobody quarrels about the value of health (p. 4). 16 White (1973) strongly suggests that mental health professionals are carelessly off target in extending the metaphor of health to all personal difficulties of living. The metaphor has been extended to encompass more ground by substituting the term "positive mental health," which focuses on desirable and undesirable qualities of health (Jahoda, 1958). White (1973) accepts terms such as "emotional maturity" (Saul, 1947) and "self-actualization" (Maslow, 1954) as more appropriate labeling of mental health. PeOple are taught to believe that there is only one proper way for a person to be and to live. All other ways are deemed evil, illegal or insane (Laing, 1960). Those who protest or deviate are met by an onslaught that begins with parental criticism and often culminates with ostracism, imprisonment or incarceration in a "mental hospital." Substantial articles and references in the social science literature have recently drawn noticeable attention to the appeal for a more "pluralistic" society. What's good for the 'goose' may not be good for the 'gander' anymore. Evidence presented in research studies manifesting an interaction by method effect is supportive of pluralism. Government funding for social welfare and manpower programs is now focusing upon a planned variation thrust aimed at providing more adequate delivery of human services. The 17 intensified search by man to find better ways to live and preserve mankind, covertly recognizes the inbalance and stress which is normative in our culture. All social systems require their members to play certain roles. Unless the roles are adequately played, however, social systems will not produce the results for which they have been organized. Jourard (1958) views a person who plays his roles suitably as having a more or less normal personality. However, he differentiates normality of personality from healthy personality. Healthy personalities are reflected in those who receive a personal pay-off and satisfaction from role enactment. Here the nature of healthy personality is dependent. upon hedonistic values. Plainly, it is healthy to enjoy and be in agreement with one's work, play, self, and so on. Those who are diagnosed as normal by standard measures of psychological appraisals, and yet are living absolutely miserable lives are seen as unhealthy personalities by the Jourard paradigm. It is not an uncommon occurrence to perceive a person playing a role which does not do justice to part or all of his self. When one does not know his real self he may, in Horney's (1950) terms, be self-alienated. Hence, his real self is a stranger. Estrangement is a primary cause of neurotic personalities (Horney, 1936). Fromm (1955) also refers to the same phenomenon as a socially patterned defect. 18 A self—alienated person does not disclose himself truthfully and fully; hence, a blockage of authentic communication and love occurs. Effective loving calls for knowledge of the object. Hence, open and full disclosure of the self to at least one "significant other" appears to be one avenue by which one awakens not only to the breadth and depth of his needs and emotions, but also the nature of his own self-affirmed values. Full self-disclosure undoubtedly enhances ‘ the growth potential in the counseling or psychotherapy process (Rogers, 1951). However, self— disclosure of the real self is not synonomous with healthy personality. Self-disclosure of the real self is a necessary but not a sufficient condition for healthy personality (Jourard, 1958). Other variables, such as role models, educational opportunities, economic stability, environment, involvement with others, and even the weather have input for healthy personality development. Some people behave in more socially approved ways when they are not being themselves. This is characteristic of the defense mechanism known as reaction formation, which is again stressful. In general, Jourard (1958) states the relation— ship between self—disclosure and various values such as healthy personality, physical health, group effectiveness, successful marriage, effective 19 teaching, and so on, as follows: (A person's self is known to be the immediate determiner of his overt behavior. This is a paraphrase of the phenomenological point of View in psychology (Snygg and Combs, 1949). New if we want to understand anything, explain it, control it or predict it, it is helpful if we have available as much pertinent information as we possibly can. Self-disclosure provides a source of information which is relevant. This information has often been misinterpreted by observers and practitioners through such devices as projection or attribution. It seems to be difficult for people to accept the fact that they do not know the very person whom they are confronting at any given moment. We all seem to assume that we are expert psychologists and that we know the other person, when in fact we have only constructed a more or less autistic concept of him in our mind (p. 506). Jourard's presentation of self-disclosure in this context has stimulated research into self-disclosure as a construct related to healthy personality. The concept of "healthy personality" is confronted by the phenomenon of differential perception. Although there is generic agreement (Ellis, 1973; Shostrom, 1973; Ansbacher, 1973) on the definition of healthy personality, different schools of thought have a vested interest in sustaining their own perception. Mowrer's (1973) reaction to White's (1973) conception of healthy personality calls for substantially more interprofessional cooperation. This is consistent with the pluralist movement and will hopefully unfold more reality to those having monolithic mentalities. 20 In summary, if healthy personality is an objective for the American culture, and not just a chaotic verbalization of one's philosophy, posture or percept— ions, definitional consensus within the helping professions is a yet-to-be achieved goal. The disparity reflected in the discussion of health personality can only result in stereotyping or mis— labeling those who depart from accepted measures of normality. Self—Disclosure The major thrust for interest in the construct of SD has come from Sidney M. Jourard and his associates. In The Transparent Self (Jourard, 1964), he notes that psychologists have been negligent in their failure to confront traditional modus operandi of interpersonal relationships which encourage people to conceal their real self rather than reveal authenticity to others. Indeed, self-concealment is regarded as the most natural state for grown men. People who reveal themselves in simple honesty are sometimes seen as childish, crazy, or naive, as for example in Dostozevsky's novel, The Idiot, or Melville's Billy Budd (Jourard, 1964, p. iii). People will disclose their aims, feelings, and perceptions of the world only to those whom they have reason to trust. Today's society is plagued with stress and paranoia. This is especially characteristic of urban areas, penal institutions and other controlled environments. 21 A "Bus Rider Phenomenon" was Found to exist in laboratory self-disclosure research (Drag, 1972). Subjects disclosed significantly more to a Stranger than to a Roommate with whom they would have continued interactions. However, disclosure to a "therapeutic" Stranger enhanced disclosure to a Roommate. Subjects paired first with the Stranger disclosed significantly more to Roommates than subjects in the roommate—stranger condition. The latter group disclosed significantly less information to Roommates than to the Stranger, while subjects paired first with the Stranger disclosed at an equally high level to both targets. Many researchers may question the discrepancy between controlled laboratory studies such as that mentioned above and field studies. Hence, SD to a stranger in practice might prove to be inappropriate if not dangerous. O'Reilly's (1971) findings, however, showed that subjects with high SD scores accept more questions from a stranger and reject fewer questions, while subjects with low SD scores reject more questions from a stranger and accept fewer questions. These results tend to support the finding of Drag (1972). Other studies have investigated various dimensions of SD. Feldman (1971) found no significant difference between parents of stuttering children and 22 parents of non-stuttering children on SD. Persons (1969) found no significant difference between college student and inmate interviewers' ability to elicit intimate disclosures from inmate interviewees. Sinha (1972) reported that SD varies with the development of age in urban female adolescents of India. As the age of adolescents advanced from early to mid—adolescence, they became more conscience of the self and began to inhibit expression and SD. Baum (1972) found low inclusion-affection groups to be more self—disclosing than high inclusion—affection groups and mixed groups. Taylor and Oberlander (1969) substantiated that high self-disclosers have a greater ability for recognition and discrimination of incomplete faces. Jourard and Richman (1963) found that both men and women report significantly more disclosure input than output in relation to friends and same—sexed parents, another representation of the "Dyadic Effect." The research of MacDonald and Games (1972) supported the hypothesis that willingness to SD can be film-mediated. In still another study, Jourard (1961a) found that White females of different denominations (Baptist, Methodist, Catholic, Jewish) did not differ on SD. Jewish males, however self-disclosed more than Methodist, Baptist and Catholic males. Comparison of 100 Black and 100 White high school students on SD, to peers, counselors and teachers as 23 targets, revealed very interesting results. The subjects gave counselors the lowest rank as a possible confidant. Self-disclosure was greatest for female friend, male friend, and teachers. Secondary findings revealed that'Whites SD more than Blacks, females SD more than males, and socio-economic status was non- significant (Ramsey, 1972). Rodriguez (1971) investigated the effect of a group leader's cultural identity upon SD in two mixed groups consisting Of six Chicanos and six Anglo male student subjects. Age, time, sex, environment, and group activities were controlled. The treatment condition was the differential ethnic identity of the facilitators, one Chicano and one Anglo group leader. It was concluded that no significant difference in SD patterns surfaced in either of the bi-racial groups. Self-Disclosure and Subject Matter Differences In the mid 1950's Jourard and Lasakow embarked upon the initial phases of SD research. The Jourard SD - 60 questionnaire was developed to ascertain _' differential rates Of SD relative to taste, interest, attitudes, Opinions, work, money, personality, and body. Jourard and Lasakow (1958) found that Whites SD significantly more than Blacks and females SD more than males. There was a high cluster of SD about taste, interest, attitudes, Opinions, and work and a low cluster for money, personality, and body. 24 Several independent efforts to replicate significant sex differences in SD have shown that females SD more than males (Pedersen '&L Breglio, 1968; Bath, Daly, 1972; Rivenbark, 1971; Jourard, 1958; Jourard & Lasakow, 1958). Vondracek (1970), on the other hand, Observed that males SD more than females. However, the content of SD interacted with sex. Females appear to SD more feelings and emotions, while males SD more cognitive content. Jourard (1971) has elaborated on some lethal aspects of the male role. The male role requires man to appear tough, objective, achieving, unsentimental, and emotionally unexpressive. Jourard (1971) states: It follows that men, seeming to dread being known by others, must be more continually tense (neuromuscular tension) than women. It is as if being manly implies the necessity to wear the neuromuscular "armor" of which Reich wrote with such lucidity. Moreover, if a man has something to hide, it must follow that other people will be a threat to him; they might pry into his secrets, or he may, in an unguarded moment, reveal his true self in its nakedness, thereby exposing his areas of weakness and vulnerability. Naturally, when a person is in hostile territory, he must be continually alert, tense, opaque and restless. All this implies that trying to seem manly is a kind of work, and work imposes stress and consumes energy, manliness, then, seems to carry with it a chronic burden of stress and energy expenditure which could be a factor to a man's relatively shorter life-span (p. 35). Martin (1972) found support for the hypothesis that a son is influenced impersonally by his father and intimately (emotionally) by his mother. Dimond and 25 Hellkamp's (1969) research indicates that later-born siblings SD more than first born and that the mother is the favorite target for SD. They also Observed that females and males SD more to same sex friends. The occurrence of this phenomenon has also been documented by others (Jourard, 1958; Pedersen & Higbee,,1969;. Rivenbark, 1971). Rivenbark (1971) substantiated the longitudinal differences between males and females and showed that those differences increased with age, but only for disclosure to parents. No significant differences in overall disclosure to male and female peer targets were observed. Because of the variety of factors which may effect SD, the sample was chosen so that subjects were as homogeneous as possible in socio- economic background, and general intelligence and came from similar communities. An odd-even correlation of .94 was reported. Earlier findings by Jourard (1958) reveal that most people are at ease in talking to others about their attitudes, taste, and work, but are comparatively reticent about their financial affairs, personality, and bodily concerns. The reserve of Blacks on money matters is eSpecially marked. In a second study, Jourard (1958) supported his expectation that married men and women SD more to each other than they do to their parents. These data suggest that the stability of a marriage may depend in some 26 measure on the amount of mutual SD between couples. SD has been presented by Jourard (1971) as an essential aspect for the growth of healthy and rich marriages. In summary, the research indicates that people are willing to discuss and reveal some personal issues; however, they are secretive on other issues. Hence, openness and transparency are tOpic-specific, and other factors must be considered prior to self-disclosure. Self-Disclosure and Positive Mental Health Jourard (1964) and Mewrer (1964) have written extensively concerning the degree of association between SD and positive mental health. For Mowrer, successful outcome in therapy is predicated on the client's complete, honest and authentic disclosure of previous wrong—doings and omissions. Truax and Wittmer (1971) questioned whether the psychologically "disturbed" person self-discloses less to others in his everyday milieu than does the well—adjusted person. Contrary to the hypothesis, their results showed a general tendency for the least well—adjusted subject to show the least SD when the target person was the "closest friend." The same held true for intimate and semi-personal dis- closures as well as for total SD. The analysis showed no relationship between SD and personality adjustment when the family member served as the target person. MMPI scales were used to measure disturbance. 27 There was no"Dyadic Effect"evidenced by the study. Shinkunas (1972) tested the hypothesis that schizophrenics evidence bizarre and inappropriate language in order to escape intense interpersonal disclosure, hence avoiding interpersonal relationships. No significant difference was found with respect to age, IQ, or length of current hospitalization. Three groups were used in this study: paranoid-schizophrenics, non-paranoid-schizophrenics and non-psychotics. The number of years since the first hospitalization was significantly greater for both non-paranoid groups than any other group. The results provide strong support for the hypothesis that schizophrenics avoid intense interpersonal interactions by verbalizing peculiar beliefs and autistic concepts (Shinkunas, 1972; Jourard, 1959a, 1964). Contrary to Shinkunas (1972) and Jourard (1959a, 1964), Pedersen and Breglio's (1968) data suggest that more emotionally unstable males tend to SD more about their personality, health, and physical appearance than do stable males. The relationship between self-described personality and SD has been studied by Bath and Daly (1972). Dominance—submission and love—hate related signifi— cantly to reported SD. However, only love-hate was significantly associated with actual behavioral measures of SD. Bath and Daly (1972) also report that subjects 28 who described themselves as passive-aggressive reported having disclosed less than other types. Furthermore, self-described dominance and love (nurturance, responsibility, and interpersonal close- ness) were related to higher levels of reported dis— closure than were self—described submissiveness and hate (hostility, competitiveness, and suspicion). Lind's (1971) data focused attention on the relationship of selected psychological variables, i.e., anxiety, maladjustment, self-esteem and self-concept discrepancies. Total actual SD correlated significant- ly and negatively with anxiety, self-concept, and maladjustment. Actual SD means for high and low anxiety and self-esteem groups were significant. High anxiety was associated with low actual SD and high actual self-esteem was associated with high actual SD. Lind (1971) recommended that the data be analyzed separately for each target, rather than summing into a composite SD score. The results also reflect the discrepancy problems associated with actual and reported SD. Numerous other studies have attempted to determine the relationship between SD and positive mental health using self-concept and self-esteem indices. 'VOSen (1967) explored the relationship between self-disclosure and emotional health as reflected in self-esteem. He 29 reported that the role of SD as a contributor to changes in esteem is inferential since no reported studies have dealt directly with the problem. This study used the Tennessee Self-Concept Scale and a SD scale to test the following hypotheses: (a) the sensitivity group, in contrast to the control group will have a greater increase in self-esteem; (b) SD and self-esteem will be linearly related; (3) there will be no interaction between the amount Of SD and the group situation. The first hypothesis was not supported. This finding appeared to be due to the reduction in esteem of the low disclosers. The second hypothesis was partially supported. SD and self—esteem were linearly related. However, the high disclosers had no change in esteem, and the linearity was due to a reduction in esteem of the low disclosers. The third hypotheSis was confirmed. The relationship between SD and esteem was consistent and was held not to be fully dependent upon the type of situation in which one discloses. Shapiro (1968) used the Tennessee Self Concept Scale and the Swensen SD questionnaire as dependent variables and found that subjects who were high in self-concept were high in extraversion and SD while being low in neuroticism. Low self-concept subjects were reported to be lower in extraversion and higher in neuroticism. 3O Doyne (1972) tested the hypothesis that SD in encounter groups would be linearly related to self- esteem, using the Tennessee Self-Concept Scale along with two ranking scales of SD as dependent variables. The hypothesis was not supported. Low and moderate disclosing subjects were higher in esteem than those in the high disclosure group. In reference to change in self-esteem, the low disclosers were the only disclosure group to change before, during, or after the encounter group experience. Contrary to other studies (Jourard, 1958; Jourard & Lasakow, 1958; Pedersen & Breglio, 1968; Jones, 1972; Bath & Daly, 1972), females were lower in disclosure than males on both SD, scales. Jones (1972) investigated sex differences and levels of positive mental health (PMH). He reported that males of high PMH self-disclose less than females at the same level. Males Of high PMH self-disclose less than males and females at medium and low levels of PMH. Males and females at medium and low PMH self-disclose the same. The "Dyadic Effect" was present. No relationship was found between socio-economic status and SD. The Tennessee Self-Concept Scale in conjunction with the Jourard SD Questionnaire (1958) and a SD rating form served as dependent measures to determine if a relationship exists between the disclosure of feelings about self to significant others and changes in self- 31 concept (Petersen, 1972). The subjects were academic probation students who voluntarily enrolled in group counseling. The results indicate that group counseling produced positive changes in both the global and social self-concepts of students. Self-disclosure was signifi— cantly related to changes in self-concept. The strongest relationship appeared between other persons' ratings of an individual's self—disclosing behavior and changes in self-concept. A weaker relationship appeared between an individual's own appraisal of his self—disclosing behavior and changes in self—concept. The relationship between self-disclosure, changes in self-disclosure and changes in self-concept was found to be a linear relationship with no significant quadratic effects. It was concluded that SD and self-concept are positively correlated. Although there are still definitional problems related to the mental health status of people, the traditional measures of psychological assessment tend to reflect a direct and positive relationship between SD and indica— tors of psychological well—being. Self-Disclosure and Interpersonal Trust The literature on SD does not provide evidence of consistent relationships between SD and any therapeutic or personality variables (Himelstein & Lubin, 1965; Lubin, 1965; Stanley & Bownes, 1966; Vondracek, 1970). Deutsch's 32 (1962) research on trust indicates that the client's risk-taking in the counseling relationship will be related directly to his trust in the counselor. Friedlander (1970) substantiated that the trust level between the counselor and the client is a key predictor of the success of counseling. The study focused on how trust is developed. The two variables systematically studied in the experiment were the expression of acceptance or rejection of another person and his statements. These were followed by the reciprocation or non—reciprocation of self-disclosures. Confederates were trained to present memorized scripts of either a personal or impersonal nature and to express acceptance or rejection via reciprocation or non-reciprocation of §§_self-disclosures. It was hypothesized that a person would trust another person more when the other person responds to the person's self-disclosures with acceptance than when he responds with rejection. The results indicate that the hypothesis was supported. A number Of recent publications have suggested that the therapeutic nonprofessional has a positive therapeutic effect on those with whom he interacts, similar to that which occurs in therapy (Bergin, 1966; Carkhuff, 1968; Poser, 1966). The notion that personal rather than professional variables differentiate the helpful from the unhelpful person suggests that therapeutic 33 transactions can occur outside of therapy and in situations in which neither the helper nor helpee are aware that they are engaged in therapeutic transactions. Research has suggested that counselors high in genuineness, empathy, and warmth elicit greater self— exploration from their clients. To test the generality of this finding outside of therapy, Shapiro, Krauss, and Truax (1969) investigated 36 undergraduates, 39 police applicants, and 20 day hospital patients. Each group rated levels Of therapeutic conditions they perceived themselves receiving from each of their parents and two closest friends. They also completed a SD scale which measured their SD to the targets. The data suggest that gs disclosed themselves in depth to those targets Offering the most genuineness, empathy, and warmth. The inference that similar interpersonal variables are effective in leading to more open, honest and full relationships in and out of therapy was substantiated. The hypothesis that SD will be positively correlated with trust appears in several studies of SD (MacDonald, Kessel & Fuller, 1972; Granoff, 1971). Granoff (1971) Observed that a significant and positive relationship occurred between SD and interpersonal relationships. Degree of satisfaction was also positively correlated with SD and self-esteem. 34 Jourard (1964) and Truax and Carkhuff (1965) support the idea that an individual's ability to disclose requires some degree of trust in the target person and some feeling that the discloser will not be penalized for his c00perative disclosures. Hence, SD can be perceived on part of the discloser to render himself more vulnerable without a pre-existing trust in the target person. In this context, SD would be an act of foolishness. MacDonald, Kessel & Fuller (1972) used Rotter's Interpersonal Trust Scale along with a SD Scale, a modified form of Deutsch's Prisoner's Dilemma Matrix (Deutsch, 1960). Sixty college students (37 males and 26 females) enrolled at West Virginia University were the §§ for this study. The data supported the hypothesized relationship Of positive linearity between SD and the Prisoner's Dilemma Game. NO significant relationship existed between Rotter's Interpersonal Trust Scale and the SD score or the sub-scores for the target persons. Furthermore, the findings suggest that Rotter's Interpersonal Trust Scale and the Prisoner's Dilemma Game were measuring two independent constructs. Deutsch (1960) focused attention on trusting behavior when he developed a two-person non-zero sum game (see Figure 2.1) in which the gains or losses incurred by each person are a function of the choices made between 35 two alternatives by one's partner as well as the choices made by oneself. Person I has to choose between Rows X and Y; Person II has to choose between Columns A and B. The amount of (imaginary) money each Berson wins or loses is determined by the box they get into as a result of their respective choices. For example, if Person I chooses Row X and Person II chooses Column A, they get into the AX box and they each win $9. Figure 2.1 Person II Person I The Options for Person I indicate that he can win most and lose least by choosing Y. Similarly, Person II can win most and lose least by choosing B. However, if Person I chooses Y and II chooses B, they both lose $9. Both can win only if they end up in the AX box. If Person I is reasonably sure that Person II is going to choose A, he can win more by choosing Y. Analogously, if Person II is confident that Person I is going to choose X, he can win more by choosing B rather than A. In the first position, the S made his choice first and his choice was announced to the other person before 36 I the other made his choice. The other person was "fictional" and, hence, the S was not informed what the "other person" chose after the first S made his choice. In the first position, the S was faced with the decision of trusting the other person or not. In the second position, the S chose second after he knew the choice of the other person. Here, too, the other person was fictional and the actual S was always informed the."other person" had chosen Row X (i.e., had trusted). Hence, in the second position, the S was faced with the decision of being trustworthy or not. The essential psychological feature of the game is that there is no possibility for "rational" individual behavior unless the conditions for mutual trust exist. If each player chooses to obtain either maximum gain or minimum loss for himself, each will lose. But it makes no sense to choose the other alternative, which could result in maximum loss, unless one can trust the other player. Unlike some of Deutsch's previous studies (in which SS were induced to assume a COOperative, individualistic, or competitive orientation), the SS were given no motivational orientation. The S did not know the identity of the other person, and he was instructed to assume whatever orientation they wanted (Deutsch, 1960). 37 The results of Deutsch's (1960) investigation indicate that SS who were "trusting" when they chose first, tended to be trustworthy when they chose last. On the other hand, the SS_who were "suspicious" when they chose first, tended to be untrustworthy when they chose last. In summary, it is important to acknowledge that interpersonal trust is not a given, but follows a relationship of successive peak experiences. Deutsch's (1960) development of the Prisoner's Dilemma Game represents a very simple design aimed at identification of trust discrepancies. The next section will focus more specifically on the social exchange of personal data. Se1f~Disclosure and Social Exchange Theory Pe0ple are highly selective in the data they provide to others about themselves. The work supervisor, the close associate, and the parent or sibling may all receive a different picture of the same individual, a picture that largely depends on what Goffman (1959) has termed the "art of impression management." Similar phenomena have been termed "fair exchange," by Homans (1950, 1961), and "The norm of reciprocity" by Gouldner (1960). The general notion is that relationships develop on a quid‘pro quo basis, resulting in a "Dyadic Effect" which is consistent within the context 38 Of social exchange theory. A study by Jourard & Landsman (1960) revealed that subjects showed a strong tendency to disclose most about themselves to fellow students whom they knew best, and least to those whom they knew less well. SD was a reciprocal kind of behavior which preceded tola.1ével of intimacy agreeable to both parties in each possible dyad and then stopped. Subjects' liking for each of their fellows was not a strong determiner of their disclosure to them. Hence, likeability for fellow students was not systematically related to disclosure output to students or disclosure intake from others. This is supportive evidence that SD proceeds on a quid pro qgg basis. Exchange of SD in a laboratory setting was found to function in a manner consistent with social exchange theory (Worthy, Gary, & Kahn, 1969). The intimacy of disclosure exchange tended to follow the norm of reciprocity. More intimate initial disclosures were made to those who were initially better liked. In turn, final liking was higher for those who made more intimate disclosures. The authoritarianism of the subject had no significant effect on self-disclosure. Likewise, no significant differences in disclosure were attributable to whether or not the situation permitted eye contact. Very strong support was found for the reciprocity hypothesis. Over time the SS tended to disclose more 39 intimate information. The"Dyadic Effect"was present. However, no evidence was found to support the expectation that authoritarian subjects, being more suspicious, would disclose less than non-authoritarian subjects. Conditionally, it may be suggested that once initial contact has been made between two persons, they proceed to "uncover" themselves one to the other at a mutually regulated pace. Gary and Hammond (1970) have found this relationship among alcoholics as well as drug addicts. If it is generally true that intimate self- disclosure begets intimate self-disclosure while impersonality begets impersonality, then certain implications follow for a number of areas of interpersonal endeavor. For example, it might prove to be true that therapists, in order to maximize disclosure in their patients, will be obliged to go beyond impersonal "technique" and "be themselves", i.e., disclose what they are experiencing during the therapy hour as freely to their patients as they expect the latter to disclose to them. Rogers (1958) has already advanced a similar technique for counselors. Possibly similar importance can be applied to teaching, nursing, and to lawyer- client relationships. Jourard (1959b) vividly expressed the same relationship when he stated: 40 Self-disclosure produces consequences, influencing the behavior of others toward oneself for weal or woe. Possibly, then, persons disclose or fail to disclose themselves in accordance with the consequences that they expect to follow. Frequently, the expectancies may not be warranted; in such cases, one would likely find both inaccurate "other- concepts" (cf. Jourard, 1958, p. 182-191) and either more or less self-disclosure than is apprOpriate to the situation (p. 430). Interpretation of this statement lends itself to social as well as political consideration relative to environments conducive for SD. The parameters for SD must be well-defined, and one would expect variation as the context and situations vary. Hence, it would appear important for inmate groups to be selective in their disclosures. Self-Disclosure and Social Penetration Related studies on social penetration have demonstrat- ed that interpersonal exchange processes are modified by situational variables i.e., compatibility, socially restricted mileu, and reward-cost characteristics (Taylor, 1968). In this context, social penetration refers to the reciprocal behaviors which transpire between persons in the development Of an interpersonal relationship. Within the framework of Heider's balance theory, Leuchtmann (1969) hypothesized that, all other conditions being constant, subjects would resist association with a person who upon first meeting the subject discloses 41 personal data directly to him. He expected this to occur despite the subject's positive attitude toward the content of the disclosure. He also expected the central peripheral variable to have some effect in relation to the other variables so that the negative impact of open and direct SD would be stronger on subjects for whom the issue was central as opposed to those for whom it was peripheral. The results confirmed the hypothesis that the recipent's feelings toward the content of disclosure and manner of disclosure have a strong influence in determining interpersonal choice. Although the manner of disclosure affected choice more strongly than attitude toward content, the direct manner of SD had a positive rather than negative effect on persons who are "high receivers." Leuchtmann (1969) therefore acknowledged that Heider's theory offers only a limited theoretical base for the results. Taylor (1968) employed a repeated measures design to explore differential rates of intimate and non- intimate levels of mutual activities and information exchange between college freshmen roommates over an extended period of time. As hypothesized, greater SD by both high-and low-revealer dyads occurred in later weeks of their relationship than earlier weeks. The prediction that high-revealer dyads would exhibit 42 greater breadth of penetration (more SD) than low— revealer dyads over repeated measure of intimacy was only partially substantiated. Hence, both low and high revealer groups demonstrated a monotonic rate of development. The research data on social penetration and social exchange theory support the notion that the breadth and depth of SD is not a random behavior. Self-disclosure is, furthermore, a process variable which develops and grows over time. Thibaut and Kelly (1959) viewed the growth and gratification achieved in dyads as a trading or bargaining relationship. In this context, an individual voluntarily enters and stays in any relationship as long as it is adequately satisfactory in terms of rewards and cost. Self-Disclosure - Early vs. Late Goffman (1963) maintains that the more an individual's past has deviated in an undesirable direction, the more he is obligated to reveal his past. Otherwise, the individual may feel guilty that he made friends or Obtained a position under false pretenses (Jones & Gordon, 1972). Jones and Gordon (1972) requested subjects to evaluate a target person from a taped interview in which the target talked about past events in his life. During the interview, the stimulus person mentioned an important 43 experience that was either pleasant or traumatic and was an event for which the target either was or was not responsible. The presentation of the critical information occurred either early in the interview or in response to a probing question at the end. The findings generally supported the expectation'that it is unattractive to disclose a good fortune early in a relationship. When disclosure of a negative experience is involved, the consequences of early versus late disclosure depend on the target person's responsibility for the event. The responsible person who discloses bad fortune early is more attractive than the late discloser. On the other hand, when the person is not responsible for an experience of bad fortune, he is better liked if the event is disclosed late in the interview. The results provide a clear demonstration that the manner Of describing a past experience may have at least as much impact on attraction as the nature of the experience described. Self-Disclosure - Sensitivity Training and Cohesiveness Efforts to forecast actual SD via self-report questionnaires have yielded mixed results (Hurley & Hurley, 1969). Ribner (1972) investigated the effect of an explicit group SD contract on self-disclosing behavior and group cohesiveness. He predicted SS SD by use of a self-report instrument. Self-disclosure and group 44 cohesiveness have generally been regarded as highly valued properties of groups; hence, a major thrust of this study was to determine if these variables could be influenced through the introduction of a contract specifically calling for self-disclosure as normative behavior. The conclusions of the study were as follows: (E) the contract to SD significantly increased both the frequency and depth of self-disclosing behavior in the groups, but did not affect the level of intimacy of tOpics discussed;'(§) a contract facilitated group cohesiveness, but mutual liking was decreased;(g) the contract significantly increased the physical distance among the group members; (S) the SD questionnaire did not predict actual disclosure in the groups. However, it did predict the direction of SS attraction in their group especially for groups under contract. Robbins (1966) hypothesized that S§_wou1d SD more to a high cohesive group than to a low cohesive group. It was also hypothesized that under threatening conditions, the effect of cohesiveness on increasing SD is less for high anxious SS in comparison to low anxious §§° The first hypothesis was clearly supported, and the second hypothesis was not supported. The researcher concluded that cohesiveness can be used to increase SD; however, an increase is dependent upon the formation of cohesive groups. 45 Fuller (1971) found high self-disclosing groups to be more cohesive, to prefer expression of negative affect, and to rate their group experiences more positively than low SD groups. Significant correlations were found between targets and three dimensions of attitude toward affect expression (positive, negative and total). The effects of trust and cohesiveness upon self-disclosing behavior were non-significant. There was no significant correlation between trust and SD or between SD and locus of control. There was no evidence of the "Dyadic Effect." However, sensitivity training did increase SD of the group members. These finding were similar to Scherz's (1972) who found that feed- back during sensitivity training increases SD. Piper (1972) compared high trust vs. low trust groups as measured by Rotter's Interpersonal Trust Scale. §§ who participated in a high-trust sensitivity group increased their ability to roleplay understanding significantly more that §§ who were members of a low- trust sensitivity group. The results indicated that sensitivity training was not an effective agent in modifying the behaviors measured by the dependent variable. In another study (Levy & Atkins, 1971), SS were 29 middle class, white, female undergraduates, all of whom were single and who had had no prior encounter with 46 group experience. High, low and mixed SD groups participated in verbal encounter groups. The groups focused on attitude skill training over several weeks. The dependent variable was a measure of self-image. It was found that individual subjects were rating them- selves more positively than they were being rated by their fellow group members. The high and low groups differed significantly in their relative group rating, with the high group being reported as more competitive, less cooperative, colder, more talkative, less friendly, and less trustworthy than the low group. In summary, the implementation of sensitivity group experiences has aided the growth and understanding between and among peeple. These groups lend them- selves to experimental learning and exploration of other variables effecting SD, i.e., the use of contracts, threatening stimuli, proxemics, trust, risk taking, cohesion and sociometric variables. Self-Disclosure and Manipulation During the past decade the psychological literature has evidenced a rapid growth in reports of self-dis- closure studies involving manipulation. Greenspoon (1955) demonstrated the efficiency of a well—placed verbal reinforcer to increase the rate at which selected verbal Operants are emitted. Jourard (1958) indicates that deliberate manipulation of man by man is not a healthy 47 interpersonal transaction. Hence, manipulation begets manipulation. Chittick and Himelstein (1967) studied manipulation of self-disclsoing behavior in a group setting. This study attempted to ascertain if submissive subjects were more susceptible than ascendant subjects to self- disclosure manipulation. It was hypothesized that naive subjects would acquiesce to norms of behavior presented by confederates and would self-disclose personal data when confederates did so. Subjects would be reticent to self-disclose when fellow confederates exhibited reticence. A measure of association between resistance to background influences and the degree of submissiveness was tested. Both ascendant and submissive subjects tended to conform to confederate self-disclosing behavior, giving support to the "Dyadic Effect." There was no significant difference between the two groups on the basis of ascendant- submissive scores. It is qualitatively desirable to support modeling and imitative behavior which has received social approval from the cultural milieu. Several behavioral scientists have researched self-disclosure in relationship to a "follow the leader" phenomenon (Jourard & Resnick, 1970; Jourard & Jaffe, 1970; Culbert, 1966; Drag, 1969). 48 Jourard and Jaffe (1970) investigated whether subjects would "follow the leader" and emulate the disclosing behavior of an experimenter, in content as well as duration of self-revealing utterances. The subjects in all the conditions tended to match their speaking time with that of the interviewer. When the experie- menter spoke briefly on each tOpic, the subjects also spoke briefly. When the experimenter spoke at length, the subjects spoke significantly longer. When the experimenter changed from long to short utterances, the subjects did likewise. -When the experimenter switched from short to long utterances, the subjects again followed suit. A correlation of .75 was found between the experimenter's and subjects' mean disclosure time. But more generally, this experiment supports Rosenthal's (1966) well-documented thesis that one of the most powefful "determiners" of a subject's performance in an experimental setting is the experimenter himself (Jourard & Jaffe, 1970). Culbert's (1966) examination of still another dimension of manipulation of self-disclosure lends support to the "follow the leader" notion in training groups. His findings showed that the SS in a group with less self-disclosing trainers more often entered "mutually perceived therapeutic relationships" with 49 "critical" group members (two trainers and their dyadic partner). The subjects with the more self-disclosing trainers had more "mutually perceived therapeutic relationships" with non-critical members. The results suggest a behavioral training technique where trainers enhance the likelihood of optimal group participation by manifesting an initial risk of high self-disclosure. In this context, the trainer is a model for other group members. Initial high self-disclosure is followed by sequential reduction Of self-disclosing so that group members might be freer to focus on their involvement with one another (Culbert, 1966). In the same study (Culbert, 1966), the question of whether subjects known to be characteristically high or low in self-disclosure would modify their disclosure patterns under the influence of the peer was subjected to a test. Low-disclosing subjects disclosed more to high disclosing subjects than to low disclosers. Jourard and Resnick (1970) descript- ively express the same technique in stating: If Openness before another (when safe) is more rewarding than concealment, and if a person of higher status has some behavioral influence upon these lower in status, then additional force may be assigned the argument that psychotherapists and other kinds of interviewers will be most effective at eliciting Openness in clients when they risk the first step and embody the "Open way" in their transactions with clients (p. 91). 50 Manipulative SD was also studied by Drag (1969), who used experimenter behavior and group size as vari- ables influencing self-disclosure. Two situational variables of experimenter'(§) behavior and group size interacted to produce significantly different subject self-disclosing behavior in 4-person groups in which the S_participated in the self—disclosing process. Disclosure in the 4-person groups in which the S did not participate occurred at significantly lower intimacy levels as compared to disclosure in groups of 4 subjects in which the S took part. The findings indicate a significant replications effect in groups of 8 persons and 4 persons, and a near-significant effect in the 2-person groups in which the S did not participate in self- disclosure. Apparently, subjects utilize the experi- mental situation as a type of "projective task" when the S fails to provide a model for them as Opposed to that situation in which subjects can model the S. In the latter situation, replication mean cell values were not significantly different among groups. Query (1964) found in group psychotherapy that the more an individual self-discloses in a group, the more he will be attracted to the group. Research by Kangas (1967) revealed that the greater the number of self- disclosures by the group leader, the greater the number of self-disclosers in his group. This study was designed 51 to determine more precisely the degree to which self~ disclosure on the part of group leaders and on the part of group members will elicit self—disclosure on the part of other group members. Three different group settings were used: (3) an adolescent group-therapy group, (2) a traditional group-therapy group, and (g) and a marathon group-therapy group. Kangas (1967) concluded that self-disclosure begets self-disclosure in small group settings. This held true when a group member or the group leader took the initial risk to self-disclose. Of particular interest is the higher magnitude of paired disclosures in the marathon groups. Upon replication, findings of this sort might support the utility and longevity of marathon groups as an effective way of increasing interpersonal transactions. The above findings may shed light on the current controversy in counseling and psychotherapy relative to therapists sharing real emotions and involvement with their clients. Many traditional therapists advocate an uninvolved approach to counseling. Other therapists (Mullan, 1953, 1955a, 1955b; Mowrer, 1964) advocate an increasing involvement by the therapist. Since one of the goals of therapy is to get accurate personal infor- mation or disclosures from the client, the research of Kangas (1971) would lend support to those who believe therapist involvement is an important part Of effective 52 therapeutic interchange. Edelman and Snead (1972) studied a tangential issue of therapist-client involvement by investigating confidentiality and coerced psychiatric interviews in two separate but related experiments. The results of the study showed that: (2) subjects avowedly revealed more personal information to mental health professionals than they would in a control employment interview situation with no differences between mental health professionals, (S) informing subjects that the interview was not confidential produced significant information loss from female but not male subjects, and (3) subjects who received no information regarding confidentiality behaved like subjects receiving a communication that the information was confidential. Edelman and Snead (1972) attribute the main effect for confidentiality to the fact that subjects believed they would reveal less when told that the information was not confidential than when they were told either that the information was confidential or were given no instruction concerning the confidentiality of the information. The main effect for sex was due to the fact that females were less revealing than males. This is a departure from previous findings concerning sex differences in this area. 53 A second experiment (Edelman & Snead, 1972) explored the effects of avowed self-disclosure occurring when peeple are coerced into psychiatric interviews. Covertly, coerced psychiatric interviews are normally a part of pre-sentence investigations of social offenders. Numerous investigators (Szasz, 1963; Ullmann & Krasner, 1969) have suggested that psychiatric consultations serve primarily as a socially sanctioned means of deal— ing with cultural deviance. This viewpoint has emphasized that persons exposed to psychiatric interviews are often coerced into this situation rather than voluntarily seeking help. Subjects who were forced to receive a psychiatric interview in Eldeman and Snead's study self-disclosed significantly less than subjects who were not forced to receive a psychiatric interview. These findings do not agree with Fritchey's (1971) study which showed that, under threat, subjects in a high-anxious group were more consistent in high intimacy self—disclosure. Low moderate anxious subjects were less consistent in intimate self—disclosure. The findings suggest that high anxious subjects are more amenable to self—disclosure manipulation than are less anxious subjects. The American Psychological Association, report of the Stuart Cook Committee (1973) on the ethical principles in the conduct of research with human 54 participants has taken a firm position on the issue of informed consent. Ethical practice requires an investi- gator to inform participants of all features of the research that reasonably might be expected to influence willingness to participate, and to explain all other aspects of the research about which the participant inquires. Failure to make full disclosure increases the investigator's responsibility to maintain confident- iality and to protect the welfare and dignity of the research participant. Jourard (1968) has stressed that Openness and honesty are essential characteristics of the relationship between investigators and research subjects. Rosenthal and Rosnow (1969) suggested three less established characteristics of §§ who volunteer to participate in experiments. Such subjects are more sociable, more arousal-seeking, and more unconventional. This would appear to indicate an interaction style favoring catharsis. Four studies especially suggest that individuals who volunteer may be seeking cathartic situations (Bell, 1964; Lasagna & von Felsinger, 1954; Martin & Marcuse, 1958; Riggs & Kaess, 1955). Hood & Back (1971) examined the nature of bias introduced by using volunteer subjects in laboratory experiments. 39 male and 55 female undergraduate students 55 were given an Opportunity to volunteer for experi— mental participation. Statistically significant differences in the mean amounts Of past self- disclosures appeared. Male volunteers were higher in self-disclosure than male non-volunteers. An interpretation of these findings suggests that the potential subject views an experiment as a self-revealing situation. The researchers recognized that individuals may vary in their attraction to self-revealing situations. This variation can introduce an uncontrolled bias into experimental studies. Mann (1972) investigated the reflection of feelings among Black and White subjects who received verbal reinforcement during treatment groups. Both groups increased their level of self-disclosure when expressed feelings were reflected. Both groups were exposed to a period of extinction, and the White groups were more resistant to extinction than the Black experimental subjects. In this study, no "Rosenthal effect" was present. The resistance to extinction of self- disclosure by the White group indirectly supports other studies which indicate that Whites self—disclose significantly more than Blacks (Dimond & Hellkamp, 1969). In summary, the studies mentioned in this section all tend to support the assumption that SD can be and“ 56 is manipulated. It is also important to recognize that the behavioralist school of thought has researched and pioneered the evaluation and quantification of manipulative behavior. Self-Disclosure and Inmates The prevailing norm that men are tough and masculine is magnified for men who are incarcerated social offenders. Jourard (1971) has discussed certain lethal aspects of the male role, stating: If a man has something to hide, it must follow that other people will be a threat to him; they might pry into his true secrets, or he may, in an unguarded moment, reveal his true self in its nakedness, thereby exposing his areas of weakness vulnerability. Naturally, when a person is in hostile territory, he must be continually alert, tense, opaque and restless. All this implies that trying to seem manly is a kind of work, and work imposes stress and consumes energy. Manliness, then, seems to carry with it a chronic burden of stress and energy expenditure which could be a factor related to man's relatively shorter life- span (p. 35). The above passage has significant implications for the role one plays while incarcerated. The struggle to achieve the status of a man in prison embraces the "I am tough" syndrome which is merely a carryover from the larger society. Men are generally characterized as cold, impersonal, instrumental, and calculating. The traditional identity relative to the nature of man negates 57 Open, authentic, and loving relationships between and among men. The restricted parameters placed on the male role foster a growth which incapacitates males from accepting significant portions Of their own feelings. Weakness, fears, doubts, and weeping all are part of the unacceptable area as defined by our culture. Jourard (1964) has further suggested that men self-conceal so much that they become alienated from and unaware of their own feelings. Then they are not only unable to deal with others well, but also cannot work and live comfortably with themselves and their own problems. Jourard's theory raises several issues which might invite counselors, psychologists, and others to investigate the concept and theory of self-disclosure, particularly in reference to prison pOpulations. Very little research has been done with self-disclosure and prison populations. Brodsky and Komaridis (1966), however, investigated self-disclosure of military prisoners; 67 military prisoners were followed over a 13 week-period after their initial incarceration at the Disciplinary Barracks, Fort Leavenworth, Kansas. The researchers Observed that prisoners maintained closest relationships with individuals outside the institution, while less disclosing relationships with fellow prisoners were reported. Even smaller amounts of 58 confiding to the institutional staff were reported. It was concluded that these individuals did not represent a highly prisonized group ("prisonization" refers to the development of the identity of oneself as a prisoner) and that the patterns of confiding tend- ed tO be consistent for individual prisoners: In another study, Brodsky (1968a) conducted a preliminary inquiry into institutional figures. The subjects were 50 military prisoners who were tested approximately 10 days after confinement as part of an admissions battery. The target persons (possible confidants) were chosen to represent major institut— ional figures. Tabulation of disclosure indicated that very little disclosure occurred to either the intermediate or administrator targets. Voluntary disclosure to fellow prisoners was significantly higher than disclosure to other institutional targets. The use of this subject group probably biased the findings toward minimal disclosure because of the limited time of incarceration. The pattern of disclosure Observed was the same as noted in other studies by Brodsky and Komaridis (1966). Braithwaite (1973) reported a study involving self-disclosure differences between Black and White inmates incarcerated as Marion Federal Penitentiary in Marion, Illinois. A modified Jourard self-disclosure 59 scale, develOped by Brodsky and Twomey (1967), was administered to 30 Black and 30 White inmates. Friedman (1969) administered the same questionnaire to two groups (passive and aggressive homosexuals) confined at the same facility and reported test-retest reliability coefficients in the low nineties. Contrary to Brodsky and Komaridis' (1966) study of military prisoners, in which the closest confidant was found to be someone external to the institution, Braithwaite (1973) reported that both Black and White inmates maintained greater self-disclosure to their "closest inmate friend." The researcher recognized the possible spurious effect resulting from the difference in long-term sentences of the Marion pOpulation in contrast to the shorter sentences at the disciplinary barracks. Braithwaite (1973) advocates a systemic counseling interpretation of his data, and states: Implications of this study from a Black and minority perspective are profound and significant to the rehabilitation process of Black and minority inmates. The most significant indentifiable variable stifling the psychological well-being of minority inmates is undoubtedly the lack of Black and minority personnel at all levels. Affirmative action efforts aimed at recruitment of minority personnel are mandatory if minority inmates are to servive and return to society with rejuvenated role models (p. 94). Self-Disclosure and Validity Several investigators have studied the predictive validity of Jourard's & Lasakow's (1958) Self—Disclosure 60 Questionnaire (Burhenme & Mirels, 1970; Himelstein and IKimbrough, 1963; Himelstein & Lubin, 1965). However, the construct validity of the instrument has only been researched by a few (Jourard, 1961b; Persons & Marks, 1970; Benner, 1968). Jourard (1961b) explored whether subjects who have disclosed little to their mothers, fathers, or closest female friends give fewer responses to Rorschach plates than subjects who are higher self-disclosers. Based on findings of a low but positive correlation between productivity on the Rorschach test and a measure of the extent of self-disclosure to selected significant others, Jourard (1961b) concluded that the data support the construct validity of his self-disclosure theory. Low productivity was viewed as defensiveness and an unwillingness to be known. Persons and Marks (1970) studied the interaction of personal attributes of interviewer and interviewee that make it possible for the severely antisocial interviewee to be self-disclosing. The subjects were 81 inmates from Ohio State Reformatory for Men in Mansfield and three male undergraduates from Ohio State University. All subjects were White, of normal intelligence, and between 18 and 22 years of age. The inmate subjects each had been incarcerated at least once before. The subjects were selected by examining the files of approximately 1,800 inmates to determine MMPI scale 61 scores. The investigators reported that the subjects were equally self-disclosing prior to administration of the experimental conditions. Analysis of the data showed no significant difference between the college interviewers' and inmate interviewers' ability to elicit intimate interviewee disclosures. The subjects' intimacy level was significantly greater on all dependent variables when the interviewer and interviewee had the same MMPI code type. Jourard's (1958, 1961c) finding that self—disclosing peeple are adaptive, psychologically healthy, integrated, well-liked and flexible, is not substantiated by the above study, which found that the subjects with the most psychopathology were the most self-disclosing of the personality types. Furthermore, the interviewee MMPI sub-scales Of Validity, Psychopathic deviate, Paranoia and SchiZOphrenia depicted a positive correlation with self-disclosure. Persons & Marks (1970.) concluded: The present findings are only in partial agreement with Jourard's earlier work, but add to the nomological net for the construct Of self-disclosure. The above results suggest that it is not always adaptive and healthy to be self-disclosing, and in some circumstances could be self- defeating and maladaptive. Perhaps it was not judicious or adaptive for an inmate to be self-disclosing with a college student in the first encounter, and it appeared that it was the person who was experienc- ing the most intra-psychic pressure who was self-disclosing. It could be that this person was self—disclosing because the Opportunity presented itself to be cathartic and not because it was adaptive 62 or mature to be self-disclosing. Self- disclosure or interpersonal intimacy certainly does not seem to be a uni- dimensional construct, but instead is a complex interaction of social class, psychopathology, situational variables, and personality attributes of two people in an interpersonal encounter (p. 391). Jourard (1959a) has alluded to the same complexity of self-disclosure relative to selectivity of self- disclosure. Hence, multivariate analysis may be the route for greater comprehension and interpretation of variables interacting with SD. Based on theory and current research, Moldowski (1966) hypothesized that self-disclosure was related to the discloser's perception of the listener and himself in predictable ways. More specifically, it was pre- dicted that the "value" of disclosures about self ratings would not be affected by value of disclosure. No support was found for the hypothesis that judging the listener would be differentially affected by subjects' revealing different kinds of self-information. Jourard (1961c) studied whether a measure Of one aspect of nursing students' interpersonal behavior (self-disclosure to parents and peers) would predict grade point average at the end of their nursing program. High measures of association and correlation were found between scores for disclosure to mother and grades in nursing courses. Jourard (1961c), discusses the implications of his findings: 63 Experience at communicating Openly with one's mother is good preparatory practice for communication with other female authority figures, viz., the faculty of the nursing college. It is likely that those students who were the most "open" impressed the faculty most favorably, and hence, earned the higher grades. The present findings provide further evidence for the predictive validity of SD questionnaires, and suggest that they may have promise of providing a measure of one of the important non- intellectual factors which might predict success in programs of training for these vocations (p. 246). In another validation study, West (1971) used target derived data to study the validity of the SD Inventory for Adolescents (SDIA) with mother, father, same sex friend, and Opposite sex friend as the target persons. Analogous data regarding SD Of subjects were orthongonally collected from each target. Correlation coefficients between the subject derived data and target derived data were computed. A checks and balance design was developed where scores collected from the targets were used to set the criterion scores for evaluating the validity of the Self-Disclosure Inventory for Adolescents. Influences were to be based upon the extent to which SDIA data coincided with influences based on criterion data, hence indirectly providing evidence for the validity Of the SDIA. The study confirmed that girls SD more than boys. For both sexes, parents and friends of the Opposite sex appear to overestimate the relative extent of disclosures they receive, whereas 64 friends of the same sex appear to underestimate the relative extent of disclosures they receive. West recognized that although the study employed target derived data as the criteria for investigating the validity of the SDIA, by no means were the data free from subjective influence. Recognition was also given to the possible occurence of response set. However, Hathaway (1965) suggests that, although response set may determine portions Of the scores on such scales, the same response set may also determine criterion behavior. In this context, response set is a valid contributor to the measurement of the construct of interest. West (1968) postulated that the tendency to say "yes" (acquiescence) to items of a self-report instrument parallels a tendency to answer in the affirmative the self posed questions related to real- life disclosure decisions. Panyard (1973) studied the validity of SD between friends using a modified SD questionnaire (subjects had six Options for each question). The amount of personal data shared between friends served as the dependent measure. The result indicated a "Dyadic Effect". Panyard concluded that the consensual validation of the quantity of personal data exchanged between pairs of friends suggests that the SD questionnaire is a valid measure Of SD to a specific target person. 65 A strong contributor to the knowledge concerning SD as a construct was presented by Benner's (1968) dissertation. His study has heuristic value and his discussion about his data is summarized by the following: (3) SD is an interpersonal construct which involved consideration of the demographic characteristics of the disclosure, the topic Of communication, the target of the disclosure, and the relationship between the sender and the receiver; (2) the comprehensive nomological network of SD constructs consisted of more hypothetical constructs than intervening variables; (2) theoretically, the content of the SD construct consisted of cognitive, affective and behavioral dimensions. Support was found for the cognitive and behavioral domains, but the affective domain was not adequately delineated. In addition, Jourard's ideas relative to the content of SD topic-targets were generally found to be theoretically and empirically sound. The Jourard instrument was highly correlated with three other self-report instruments of self-disclosure. In summary there is only a small number of studies that provide information relative validation of the self-disclosure construct. Jourard (1968) has recommended the replication of self-disclosure studies as an essential component to theory building and acceptance of his SD theory. 66 Proxemics Proxemic distance (Hall, 1963), body buffer zones (Horowitz, Duff & Stratton, 1964) or personal space (Sommer, 1967) are all terms which have been used interchangeabley in the literature. Proxemic distance has been found to function as a catalyst, hence increasing distance under stress (Dorsey, Meisels, 1969; Leipold, 1963). Newman and Pollack (1973) measured the interpersonal physical space of 30 deviant and 30 normal high school students. It was hypothesized that proxemic area is greater in deviant than in normal adolescents and also that proxemic distance increases from front to rear. The analysis of the data indicated support of both hypotheses. Investigation of proxemics has been extensively researched in the natural sciences. More specifically, the effect of territorial domain in the animal kingdom has contributed to the knowledge of ecology and habits of lower primates. The number of seconds that four dolphins (Lagenochgnchus obliquidems) spent within five feet of one another was estimated under four conditions: (3) immediately after capture, (2) after three weeks in captivity, (g) with a person in the pool, and (S) when the subjects were alone, (Maier, 1968). Comparison of conditions (3) vs. (2): 67 and (g) vs.'(S) confirmed that subjects spend more time in tight schools (close together) under stressful conditions in contrast to nonstressful conditions. Bates (1970) notes that "Primate field studies Of the last decade have reported much information of relevance for understanding private territoriality (p. 271)." He reviewed several field studies which suggest that there are territory-related spacing : mechanisms which are operative in some primate species. However, he recognized that territorial parameters between primate groups are far from universal, even among those species which do establish territories. He further noted that the significance of territor- iality should be interpreted as one of complexity of ecological adaptations. Hall (1963) developed a system for the notation of proxemic behavior. He defined proxemics as: The study Of how man unconciously structures microspace - the distance between men in the conduct of daily transactions, the organiz— ation of space in his houses and buildings, and ultimately the layout of his towns (p. 1003). Hall's (1963) system of Proxemic notation has the following dimensions for quick and ordered recording: (g) postural-sex identifiers, (g) sociofugal- sociopetal orientation, (2) kinesthetic factors, (9) touch code, (3) retinal combinations, (S) thermal code, (g) olfaction code, and (g) voice loudness 68 scale. The eight factors are equal in complexity, and they do not all necessarily function simultaneously. The thermal and olfaction factors function only at close distances. Vision is even more complex, and is usually screened out at close distances. This record- ing scheme enables notation in 30 to 60 seconds, and it is a systematic Observational system which allows for repeated measures over time. Interestingly, Hall refers to displacement as the capacity of language to deal with displacements in time and space. In the animal kingdom, territorial markers (particularly the Olfactory markers) represent displacements. Boundary markers, e.g. Open doors, fences, arrangement of chairs in a group setting, and the layout in a courtroom, all enable someone who knows the system to interpret past events or intended communication. McBride, James and Shoffner (1963) found that aggregates of domestic fowls exhibit nonrandom spacing patterns. They report that subordinate birds avoid frontal aspects of their dominant neighbors moreso than other aspects. King's (1965) interpretation of frontal avoidance focused attention on the pecking order of fowls and the number of packs received from dominant members. In another study, King (1964) found consistent, nonrandom spacing patterns in a group of college sophomores. 69 Social distance as a measure of status differentials in interviewer-interviewee relationships were studied by Hyman and Cobb (1954). They found data which supported the "Rosenthal effect," i.e., the tendency of interviewers wanting to enter deeply into a respondent's affairs tended to bias the respondent's answers. The data also indicated that White interviewers obtained more biased answers from low- status Black reapondents than from high-status (economically) Black respondents. Social distance was defined (Bogardus, 1933) as the degree of sympathetic understanding that exists between persons. In this context, social distance represents a feeling or emotion in contrast to an actual physical distance. Sommer (1967) investigated the systematic arrange— ment of individuals in small groups. The results indicated that physical arrangement affects communication, friendship, and status differentation between individuals. In another study, Sommer (1959) studied leadership behavior and fixed physical arrangements. Leaders of small discussion groups consistently gravitated to the head position at rectangular tables. Horowitz, Duff, and Stratton (1964) investigated physical space differentials between schizophrenic and nonschizophrenic mental patients. Subjects were 70 directed to approach either a person or a hatrack. When the approaching subject stopped, the resulting distance was measured. It was found that both groups approached the inanimate object closer than the person. Williams (1963) found that introverts situated themselves further from others than did extroverts. Campbell, Kruskal and Wallace (1966) used seating arrangements Of Blacks and Whites as an attitudinal dependent variable in a study involving three Chicago colleges. The resulting clustering of both groups with same race peers was interpreted as differences in ethnic attitudes. Sommer (1967) views the knowledge Of how groups arrange themselves as significant input for those fostering or discouraging group interaction. A library which has special quiet sections would be interested in arrangement techniques aimed at minimizing unwanted disturbance and social contact. In support Of manipulation of physical environments, Sommer (1967) states: It is mindless to design mental hospitals without taking cognizance of the connection between physical environment and social behavior. The study of small group ecology is important not only from the standpoint of developing an adequate theory of relation- ships that takes into account the context of social relationships, but also from the practical standpoint of designing and maintaining functional contexts in which human relationships can develop (p. 150). 71 The same rationale is applicable to the growth and development of social Offenders who experience an even more restricted environment. Cheyne and Efran (1972) investigated invasion of group controlled territories. They observed the number of individuals who walked between a conversing couple standing at different distances and compared these frequencies with those generated by a chance model. It was hypothesized that fewer people would walk between two conversing dyads than would walk between either dyads not interacting or between inanimate objects. When dyads were not interacting there was no significant difference between the preportion of people who walked between them and the preportion who walked between two inanimate Objects. Horowitz (1968) studied the interpersonal approach distance in psychiatric residential patients. 30 females diagnosed as schizophrenic or depressive were requested to approach an object person by leading with various aspects of their bodies and stopping when getting closer made them uncomfortable. The schizophrenic required more initial space. However, over time and with improved health they required less distance. Likewise, members of the depressive group, who were closer on initial measurements than the schizophrenic group, established greater physical closeness later in 72 hospitalization when coupled with improved health. Albert and Dabbs (1970) investigated physical distance and persuasion. A friendly or hostile speaker delivered two persuasive addresses to a subject seated at 2 feet, 5 feet and 15 feet away from the speaker. Measures of attitude change, recall of content, and selective attention were greatest at the middle distance. The data suggested that there is an optimal point at which resisting and accepting forces meet. Hall (1966) has proposed a series Of culturally bound spatial zones an analysis that implies that acceptance of influence is curvilinearly related to distance. Effective communication is dependent upon the appropriateness of the communication for the parti— cular audience and the Optimal spatial zones Of the speaker. Violations of Optimal spatial zones will distract from the content of the message itself and arouse defensive reactions from the listeners (Albert & Dabbs, 1970). Argyle and Dean (1965) theorized that as an experimenter reduced distance between himself and a subject, the subject would decrease his self-disclosure to the experimenter, thereby restoring some eXperience Of distance equilibrium. In the first of two experi- ments, (Jourard & Friedman, 1970), distance between experimenter and subject was varied from a maximum 73 where the experimenter left the interviewing room, to a minimum where the experimenter offered continuous eye contact to the subject. The dependent measure was duration of self-disclosure on a number of personal topics varying in intimacy level. As distance decreased, the female subjects reduced their self- disclosure while the males showed no significant increase or decrease. The findings provided partial support for Argyle and Dean's "distance-equilibrium" hypothesis. In a second experiment, the same experimenters interviewed male and female subjects at further decreases of "distance". These ranged from being present but silent, to making physical contact with the subject, and disclosing to the subject. There was a linear increase in subjects' disclosure as distance was reduced. This finding was contrary to an expectation based upon the Argyle-Dean hypothesis. Self-disclosure from an experimenter, in combination with minimal physical contact, facilitated self-disclosure from subjects rather than inhibited it. The reductions in distance were accompanied by increases in positive experience reported by subjects and by increased change in subjects' impression of the experiment. In summary, the literature on proxemics has explored flight distance, critical distance, and personal and social distance of birds and mammals (Hediger, 1955). 74 However, the study Of proxemic behavior in man is relatively new and, according to Hall (1966), very complex. Inquiry relative to spatial needs Of people is of concern to architects and urban planners. The space requirements among lower order animals vary_ from species to species. The same sort of variability manifested within the animal kingdom is more or less operative in the human species. Hence, mental health hospitals, prisons and other human environments have Optimal space needs which are conducive for growth and development of people. It is therefore a challenge for environmental engineers (including psychologists) to identify optimal space needs and to support efforts aimed at more enhanced and healthier living environments. 75 CHAPTER III DESIGN AND METHODOLOGY Overview The major thrust of this study was to investigate the differences in self-disclosure and proxemics of incarcerated social offenders. Information was ascertained from Black, White,and Chicano inmates relative to their willingness to relate, reveal, or discuss personal issues with six institutional affiliates and one external target group. Each ethnic group consisted of half recidivist and half non-recidivist inmates. Measures of proxemic distance and calculation Of proxemic area were taken for each subject to determine territorial need parameters. The remainder of this chapter will describe the procedures used in collecting the data Of the investigation. Population The population came from the Michigan Department of. Corrections--State Prison Of Southern Michigan (Jackson State Prison). The inmates are adult offenders, serving time for either felonies or misdemeanors. Four different 76 classifications are used to categorize inmates within the system. Group A includes offenders with no previous record, or perhaps a record of juvenile probation, and/or one jail term. It is this group that will constitute the non-recidivist group in the study. Group B includes Offenders with previous records of juvenile commitment, multiple jail terms, or one previous probation term. Group C includes offenders with previous records of two or more probation terms, violators of probation with a new sentence, or more previous prison terms. Group D includes escapees, parolees with a new sentence, and those committing serious crimes for which probation is prohibited or unlikely. The grand total (system-wide) of individuals committed to Michigan State Prisons in 1972 was 4,734 inmates, down 712 from 1971. Fifty-five percent of all male commitments were under the age of 25 years, with 17 percent under the age of 20 years. Thirty—seven percent of all commitments were born outside of the State of Michigan. Most commitments reflected a need and capacity for more education. Of 4,547 commitments on whom test results were available, 2,814 or 61 percent had an intelligence quotient of 90 or above, adequate to complete high school. Of the total tested, 911 were at the fourth year school level or belOw. 77 An examination Of personal characteristics Of persons committed to prison during 1972 revealed that many had significant personality disorders and were in need of specialized treatment. Over 29 percent had a history of referral, examination, or diagnosis for emotional or mental disorders. The work records showed that 44 percent had an unstable employment history. Seventy percent had a history of prior confinement in a correctional institution. A breakdown of personal characteristics can be seen in Appendix A. The mean length of sentences of all inmates is approximately 13 months, with a median of 12 months time until the next parole board hearing. In 1969, the median I.Q. for Jackson inmates was 95, with an overall median grade rating of 5. The median age was 32. Jackson has approximately 3,839 inmates. Approx— imately 1,501 are White, 2,263 are Black, 19 are American Indian, 36 are MexicaneAmericans, 2 are Orientals, 3 are in an "others" classification, and no information was available for 15 inmates. Prison assignments are in the following areas: 48% - institutional operation, 19% - industries, 12% - academic school, 3% - vocational school, and 18% are unassigned. Additional personal characteristics can be seen in Appendix A. 78 All pOpulation statistics were ascertained through the assistance of the Department's Program Director, Mr. William Kime. Subjects Two computer printouts, one for Whites and one for Non-Whites, were generated for this study.l These listed all inmates at the State Prison of Southern Michigan who had an average grade reading level of 4.9 or above. The printouts listed 1633 Non-White inmates and 1087 White inmates at this reading level or above. The two lists served as the sampling frame from which the sample was derived. An additional list consisting of all Chicano inmates was procured from Mr. Peter Rios, Coordinator of the Chicano inmate organization. The Chicano list was then matched with the two printouts to determine which Chicano inmates scored above 4.9 on the Average Grade Reading Test. In recognition Of the comparatively small percentage of Chicano inmates, their selection entailed a cluster sample, where 10 recidivist (B,C, and D prefix numbers) and 10 non-recidivist (A prefixed numbers) were sampled from a total list of 35 inmates. Two criteria were specified to recruit subject participation. The first criterion focused on a minimum reading level of 4.9, which controlled for confounding by instrumentation. Hence, all §§ were believed capable 79 of accurate interpretation and response to a self- disclosure questionnaire. The second criterion strat- ified the recidivist (R) from the non—recidivist (NR) inmates. Table I represents the systematic sampling fractions used to insure equal chance proportional to size of inclusion: Table 3.1 Probability Sampling Ratios for Jackson Prison Inmates Recidivist Sample Fraction Black N = 144 1/7 White N = 135 1/7 Non-Recidivist Black N White N 793 1/40 480 1/24 The initial sample fraction was aimed at selecting 20 inmates for each level for the time incarcerated in- dependent variable. The initial sample also represented an oversample,with anticipation of initial mortality. A random deletion process was used after the desired number of categorical S§_was obtained. The initial mortality involved eight potential §§ who declined to participate. Specific biographical data on SS who participated in the study can be seen in Appendix B. However, the following represents summary data for the 80 sampled SS: the mean age for the White NR and R, Chicano NR and R, and Black NR and R was 36.4, 37.6, 33.9, 28.8, 26.5, and 34.4 respectively. Marital status for the three racial groups showed that within the White SS 8 were divorced, 11 single, 9 married and 2 widowed; -within the Chicano SS 7 were divorced, 8 single, and 5 were married; within the Black §§ 4 were divorced, 20 single, and 6 were married. Procedure Response set and faking are big problems in non- cognitive measurement. Hence, the importance of establishing a receptive and mutual atmosphere Of good will and rapport between the investigator and the subject was emphasized. All SS selected for the sample received a "call- out notice" (appointment) one day prior to testing. This notice requested that they report to the education department of Jackson Prison. The SS_were scheduled one half hour apart. When the §§ arrived they were directed to a private testing room where the investigator awaited them. The investigator, who is Black, introduced himself as aw Michigan State University graduate student and presented a brief explanation of how the subject was chosen and the nature of the study. The investigator attempted to initiate and facilitate a verbal exchange aimed at establishing a feeling of good will. Approx- imately 80 percent of the §§ were receptive to initial 81 overtures by the investigator to engage in conversation. In fact, many of the inmates appeared very eager to talk with someone from what is called the "free world." The investigator ended each exchange by reiterating the purpose of the study and reinforcing the importance of frank and honest responses on the self-report instru- ment (see Appendix C) and the proxemic distance measure. Issues related to confidentiality and anonymity of the data analysis were discussed. The SS_were also informed that an abstract of the study would be available for their perusal. The SS were then requested to participate by re- sponding to the two parts of the study. Part one involved completing a self-disclosure questionnaire (testing time = 35 minutes). If the subject declined, the researcher probed for a reason. The potential sub— ject was informed, however, that it was his prerogative to decline participation. The reasons why subjects declined were recorded. It was emphasized that no reprisals would result to any declining subject. Conn senting §§ were given a pencil and a cepy of the instru— ment, which had instructions and three example questions on the cover page. Each Chicano S had the Option of responding to a questionnaire written in Spanish or English. After the S read the instructions and responded to the three example questions, the investigator probed the S toihsure that the S had accurately interpreted the 82 instructions. The researcher then told the S_tO take his time and to answer every question. After the S completed the questionnaire he was given a slip to complete that requested his age, length of sentence, and hometown. The S was then directed to go to another room where the proxemic distance measure was obtained. The S was met by a Black graduate assistant who explained the procedure for the final part of the study. The dimensions of the empty experimental room were 16 by 17 feet. The graduate student instructed the subject to place his feet in a pair of silhouette shoe prints taped to a centrally located floor tile. The subject was then informed by the graduate assist— ant that he (the graduate assistant) would be approaching him from four sides: front, back, left, and right (proxemic distance test). Each subject was instructed to tell the approaching graduate assist- ant to stop at the instant when he (the subject) was just beginning to feel uncomfortable about closeness. The graduate assistant always began his approach eight feet from the subject and approached twelve inches at each step. After each step, the graduate assistant paused for three seconds. The graduate assistant would stop himself whenever he was within six-nine inches and the subject had not called for a "stOp." The graduate assistant recorded the linear distance each time the subject said "stop." 83 Actual recording of distances was done on a sheet Of l/2-inch graph paper. This procedure is a replication of the procedure used by Newman and Pollack (1973) in their study of proxemics in deviant adolescents. After completion of the second part of data- gathering, the answer sheets were coded by race and by recidivist vs. non-recidivist categories. The data were collected over a two-week period during June, 1974. All the MMPI scores were obtained from the reception and diagnostic center, which is the intake division for the prison. Instruments S-D Inventory Scores on a 21-item modified Jourard Self- Disclosure Inventory (SDI) served as the primary dependent variable in this study. (See Appendix C). Jourard (1971) has demonstrated that his question— naires (of lengths that include 15, 25, 35, 45, and 60 items) have satisfactory reliability. Odd— even reliability coefficients in the .80's and .90's have been reported (Jourard, 1964). Taylor and Altman (1966) use of 35 and 20 item SD instruments yielded split half and alternate reliability coefficients of .82 and .86. 84 Items 1 through 8 Of the SDI have been used in prior research (Jourard & Jaffe, 1970) and have been shown to have high intimacy values. The numerical ratings were done by 80 college female students. The median (2.60) was used as the dividing point between high and low intimacy ratings. Items 1 through 8, which were used in this study, have the following intimacy values: 3.43, 4.01, 3.50, 3.40, 3.20, 3.73, 3.17, and 3.34, respectively. Items nine through thirteen come from Jourard's 60-item questionnaire, which yielded an Odd-even corrected correlation coefficient of .94. The remaining eight items were extracted from several studies (Brodsky, 1968a; Jourard & Resnick, 1970; Jourard & Friedman, 1970) investigating self—disclosure. The Jourard Self-Disclosure Questionnaire has been used in at least 17 published studies, including four studies which were conducted with groups (Benner, 1968). Query (1964) used a modified and shortened form of Jourard's scale in a study of SD as an independent variable in group psychotherapy. He reported a correlation of .59 between ratings made of each subject after each session by group leaders and the scores made on the Jourard SD scale. The 21—item inventory was administered to elicit self-report response of self-disclosure topics. The inventory permitted the respondent to 85 choose one of four possible conditions depicting his feelings on self-disclosure of the item to a target person. Condition A (numerical value equals 0) Condition B (numerical value equals 0) Condition C (numerical value equals 1) Condition D (numerical value equals 2) 86 The four conditions are: —Have lied, or would lie or misrepresent myself to the target person(s) so that he has a false picture of me. -Have not discussed and would not discuss anything about this aspect of me with the target(s). -Have discussed or would discuss in general terms only about this item. The target person(s) has only a general idea about this aspect of me. -Have discussed or would discuss in full and complete detail about this item to the target person(s) so that he could or would be able to describe me accurately. The numerical entries from conditions C and D were summed, yielding totals which constituted a self— disclosure score. The subject reaponded to each item separately for the following target persons: 1. Closest Inmate Friend 2. WOrk Supervisor 3. Psychologist and Counselors 4. Correctional Staff 5. Family and Friends on the Outside 6. Any Inmate 7. Education Staff Job descriptions for work supervisor, psychologist or counselor, correctional staff, and education staff targets can be seen in Appendix D. Pilot Study A test-retest S D reliability study was con— ducted by the investigator at Ionia State Reformatory, which is another correctional facility under the auspices of the State Department of Corrections. 39 inmate SS participated in a pretest; however, only 25 §§ of the initial 39 were available for the post— test. Split half reliability (odd-even) coefficients for the 39 pretest §§ were .92, .94, .94, .97, .95, .93, .95 (with Spearman—Brown correction) for closest inmate friend, work supervisor, counselor or psycho- logist, correctional staff, family and friends on 87 outside, any inmate, and education staff, respectively. A .85 correlation coefficient was found between pre- and posttest measures (with a two week time lapse) for total self-disclosure scores for the 25 §§' Proxemic Indicator The procedure for obtaining the proxemic area and distance was described in a previous section of this chapter. The following description indicates the procedure for arriving at the proxemic area. In calculating proxemic area, a transparent plastic overlay with l/8-inch graph print was used. A best fit curved line was drawn between the fixed points on the recording sheets by using a clear illuminated window glass. An overlay was placed over the recording sheet and the inscribed l/8-inch squares (16 squares = 1 square foot) were counted. One square foot occupied by the subject was subtracted, yielding the proxemic area. Measures for the front, rear, right, and left proxemic distances were calculated. MMPI The Paranoia subscale on the MMPI (number 6), served as the third dependent measure. The qualities evaluated by this scale are suspiciousness, feelings of being picked on or persecuted, and oversensitivity. The criterion group 88 for this scale was composed of subjects who were suspicious, prone to delusions of persecution, and to a grandiose sort of egotism (Hathaway and McKinley, 1951). MMPI scores were on file with the following ratios: Black (R), 5 out of 15; White (R), 9 out of 15; Black (NR), 11 out of 15; White (NR), 11 out Of 15; Chicano (R), 5 out of 10; Chicano (NR), 5 out of 15. All MMPI scores were retrieved from the Reception and Diagnostic Center Of Jackson Prison. .Estimates of internal consistency (split-half coeffi— cients) were reported in the MMPI Handbook (Dahlstrom and Welsh, 1960, p. 474) and are depicted in Table 3.2. 89 .wHOOm whom mafia 800m mason BB one 8.28m =musm= Panama so mmHOOm c853 one moofldflgooo .OH mdmom HE gmflmmnnoamfioz 05 E0 magma 880m gouge. Scum ammofi mums Henge .Hmouo emocmamnumugoo u 693: u w\H Snow uoaxoon 90mm ".0 Show Emu Agog? u Hm » c 38.3% mm mm H 8 S 2 S 3 mm 8 .. H HES 2 SH 0333 6.83 S H 383% az ex: 8233 affine 98 SEE H mm H om 8 8 mm Q. H mm oh Edema mi .8830 o§§ z 2 3 mm 8 S cm A m. mm. em 9 ob mind: mi .3830 ong 2 Ne me 3 am e. s. A S 3 mm a. uh STSH ma .wucmfio 09:3 2 mm 873 mi S 8 S 8 8 8 me me mm 8 on" mmucmfio o§§ 2 3133335 . H.505 "Hog. . . E H ASHE H«Hams ex: 8223: 886mm e8 Fangs: mm 2. 8 mo- ms 3. 2 mm 2. .. H no 3895 mmmfloo ex: 313311.380 Em Ends mam—595 .3502 :9 «S 89. 38 Be o5 8e Se as We a o m m in.) m mp v m m a Show mg 96.5 wow 2 8.30m momma 30¢th . 8.30m N.m OHDOH. Hes you Aflfiaoflmmoo flaming no message Ocfimamfio 1.585 90 Design A two-by—three fixed effects model was employed in this investigation. Schematically, this model is shown below in Figure 3.1 Figure 3.1 Design Model Fixed I.V. - Race Fixed I.V. - Times W B C Incarcerated i i NR 15 : 15 : 10 I l i i R 15 : 15 g 10 u l Independent Variable (I.V.) (1) Race = three levels B=B1acks = 30 W=Whites = 30 C=Chicanos = 20 (2) Times Incarcerated NR Non—recidivist R Analysis of Data and Decision Model Recidivist = having served more than two independent sentences The data analysis employed an inferential model to identify differences between and among groups. More specifically, a multivariate analysis of vari- ance (MANOVA) was used to simultaneously correlate 91 the relationship among the three dependent variables. MANOVA analysis also yielded univariate tests. A repeated measures analysis was done to identify differences between targets. When appropriate, Scheffe post hoc procedures were used to determine the direction and magnitude of the differences. An a priori alpha was set at .05 to represent the level of significance acceptable for this investigation. 92 CHAPTER IV ANALYSIS OF RESULTS The statistical analyses for this investigation were conducted at the Michigan State University Computer Center on the Control Data 6500 Computer (CDC). The summary data and statistical test reported in this chapter were computed by a multivariate analysis of variance and covariance programs developed by Finn (1967), which were modified and adapted for use on the CDC 6500 by Scheifley and Schmidt (1973). Scheffe (1959) post hoc analyses and graphs are presented to depict and specify the significance of group differences. The major purpose Of this study was to test the differences in self—disclosure as reported by Black (B), White (W), and Chicano (C) and by recidivist (R) and non-recidivist (NR) social offenders to specified target persons. To test those differences six major hypotheses were formulated. The first three were con— cerned with main effects and interactions of total SD scores, and the second three focused on differences among the repeated measures. 93 A secondary purpose of this investigation was to identify the correlational relationship between total SD scores and Proxemic Area Scores; and the correlational relationship between total SD scores and the Paranoia Sub-Scale Score of the Minnesota Multiphasic Personality Inventory (MMPI) for Black (B), White (W) and Chicano (C) groups. Finally, this study sought to test the differences in Proxemic Area Distance as reported by Black (B), White (W) and Chicano (C) groups. A 2 x 3 data matrix was constructed for the test of main effects and interactions. An alpha level of .05 was selected to determine statistical significance in testing the omnibus hypotheses of this study. Results of the Test of Hypothesis 1 It was hypothesized that there would be a signifi- cant difference between Black, White and Chicano groups on reported total SD scores. A univariate analysis of variance procedure was applied to test Hypothesis 1. Results of the univariate test of Hypothesis 1 are reported in Table 4.1. 94 Table 4.1 Univariate Tests for Differences between Group Means on Total Self-Disclosure Score Source Mean Square SS F-Ratio P less than Decision Black, White 556.2372 2 1.14 .3255 N.S. & Chicano Groups R X N R 379.5018 1 .78 .3808 N.S. R—NR.X 3196.8991 2 6.55 .0025 *Sign. Race A Interaction Error 488.08 74 Term *Significant - Alpha Level = .05 Since a significant univariate F-ratio was not obtained (F=l.l4, P < 0.3255), the hypothesis of race group differences was not supported. This indicates that no statistical difference existed between Black, White and Chicano groups as measured by total SD scores. Hypothesis 1, thus, was not supported. Results of the Test of Hypothesis 2 It was hypothesized that there would be a signifi- cant difference between R and NR groups on reported total SD scores. A univariate analysis of variance procedure was applied to test Hypothesis 2. Results of the univariate test of Hypothesis 2 are reported in Table 4.1. 95 Since a significant univariate F—ratio was not obtained (F=.78, P < 0.3808), the hypothesis of times of incarceration differences was not supported. This indicates that no statistical difference existed between R and NR groups as measured by total SD scores. Results of the Test of Hypothesis 3 It was hypothesized that there would be a signifi- cant interaction effect between Black, White, and Chicano groups with R and NR groups on reported total SD scores. A univariate analysis of variance procedure was applied to test Hypothesis 3. The test supported Hypothesis 3. Results of the univariate test of Hypothesis 3 are reported in Table 4.1. A statistically significant F-ratio (F=6.55, with 2 and 74 degrees of freedom, P < 0.0025) did result in testing the interaction hypothesis on total SD scores, as can be seen graphically in Figure 4.1 96 Figure 4.1 Interaction Effects of Recidivist — Non-Recidivist by Race Groups on Total Self—Disclosure Score 185 175 165 NR = IIIIIIIIII 155 R=— 145 Q .~' 9 Q Q 1 35 "'Illllllflu" 125 115 105 95 85 75 ,’ Blacks Whites Chicanos . The graph shows mean values ranging from 89.80 to 178.10 on total SD scores. The Black groups were relatively close in mean values (NR=130.26, R=129.06), followed by the White groups (NR=133.60, R=166.06) and the Chicano groups differed the most in total mean SD values (NR=178S10, R=89.80). Within the recidivist groups, the Chicano (NR) group was the least disclosing and the White (NR) group was the most disclosing. Within the non-recidivist groups the Black (NR) was the least disclosing, while the Chicano (NR) group was the 97 most disclosing. The standard deviations for the Black groups were relatively similar, however, the standard deviations for the White and Chicano groups was considerably more variable, as can be seen along with the cell means in Table 4.2. TABLE 4.2 Cell Means and Standard Deviations for Black, White and Chicano Race Groups by Recidivist — Non—Recidivist Groups on Total Self-Disclosure Scores BLACK WHITE CHICANO NR R ' NR R NR . R Mean 130.26 129.06 133.60 166.06 178.10 89.80 Standard 55.30 59.51 42.79 75.66 70.25 37.33 DiviationL Results of the Test of Hypothesis 4 The construction of the Self-Disclosure Inventory (SDI) made it possible to quantify and observe the reported self—disclosing behavior of §§ to specific target persons. It was hypothesized that there would be a significant interaction between Black, White and Chicano groups on reported SD scores to specific target persons. A multivariate and univariate analysis of variance for a repeated measures procedure was applied to test Hypothesis 4. The tests supported Hypothesis 4. 98 Results of the multivariate and univariate tests of Hypothesis 4 are reported in Table 4.3. TABLE 4.3 Multivariate and Univariate Tests for Race Crossed with Target Interaction Multivariate D.F. 12 and 138 F—Ratio 1.85 P < 0.0463 Univariate Variable Between Mean Univariate F P less Squares Ratio than Diffl 178.5256 2.9291 0.0597 Diff2 65.1726 1.7745 0.1768 Diff3 43.2776 .6317 0.5346 Diff4 110.4863 1.9316 0.1522 Diffs 180.2934 2.2594 0.1116 Diff6 102.0635 2.4001 0.0978 D.F. for Hypothesis = 2 D.F. for Error A significant multivariate F-ratio (F=l.85, with 12 and 138 degrees of freedom, P < 0.0463) did result in testing the race by target person(s) hypothesis of interaction. However, the significant differences were not on the marginal distribution which is observed 99 by the univariate F-test. A series of one way analyses of variance generated tests for differences among race within specific target persons. The results of the multiple univariate F's yielded no significant differences among race within target persons. These results are reported in Table 4.4. TABLE 4.4 Univariate Tests for Differences among Race within Targets Univariate Variable Between Mean Error Univariate F— P less Squares Terms Ratio than Closest Inmate 26.1083 74.527 .3503 .7057 Friend Work Superr 310.0833 129.639 2.3919 .0985 visor Counselor or 179.3333 126.602 1.4165 .2491 Psychologist Correctional 328.3854 138.574 2.3697 .0986 Staff Family & Friends 31.2771 105.787 .2957 .7450 on the Outside Any Inmate 196.2083 118.203 1.6599 .1972 Education 164.6604 140.459 1.1723 .3154 Staff D.F. for Hypothesis = 2 D.F. for Error = 74 100 101 Inspection of the cell means (see Figure 4.2) shows that the patterns of SD are not the same. Figure 4.2 Cell Means for Race Within Targets on Self-Disclosure Closest Inmate Friend Family and Friends on the Outside Counselor or Psycho— logist Educational Staff Work Supervisor Correctional Staff HHHHHHN hmmqmoo Any Inmate OHMWDU‘O‘Q Black White Chicano Figure 4.2 shows a similar interaction for the work supervisor and correctional staff targets. The White group was higher on total SD to the work supervisor and correctional staff targets when compared to either the Chicano or Black groups. The variability was similar for SD to the work supervisor target, however, the Black and White groups' variability for SD to correctional staff targets was 4.03 and 2.71 standard deviations higher than the Chicano group. This indicates closer agreement within the Chicano group on SD to the correct- ional staff targets. Self-disclosure to the closest inmate friend and family and friends on the outside targets was highest in magnitude and the standard deviations were greater within the Black group and White group. These data can be seen in Table 4.5. The mean score values on SD to counselor or psychologist target showed greater SD for the White and Chicano groups. However, the variability was greatest within the Chicano group, followed by the White group (standard deviations, Chicano = 12.47, White = 10.92, Black = 10.72). The graph also indicates that the White group was higher in SD magnitude to the any inmate target, and the standard deviations for the race groups were close (standard deviations, White = 11.22, Chicano = 11.60, Black = 9.98). Finally, SD to the educational staff was highest for the White group and lowest for the Black group. The cell means for the White and Chicano groups are very close (White = 18.79, Chicano = 18.10, Black = 14.36). The variability was greatest for the Chicano group and least for the White group (standard deviations, Chicano = 13.18, Black = 11.40, White = 11.36). Observation of all cell means and standard deviations can be seen in Table 4.5. 102 ma.ma oa.ma em.HH mh.mH ov.HH mm.va. om.aa om.m NN.HH mm.ma mm.m mmWOH mv.m om.m~ om.oa ob.mm mm.oa mm.e~ Hm.m mm.ma mm.HH mw.mH «m.ma mN.NH hv.NH ov.mm mm.0H mm.o~ Nb.oa mo.NN vm.HH mm.vH mm.oa mv.o~ ov.HH mm.¢H mm.m on.bm mm.m mw.mm vo.m mm.mm eemum Hmaoaumonem muoEcH and oofimuso may no camwum w hHHEmm mmmuw HMCOfiuowHHOU umwmoaonommm HO HOHomcsOU Homfl>umdsm xuoz UGOflHh OfimpfiGH “$.mean 103 coaumfl>mo new: onmocmuw mmmmmmm coflumfl>mo cooz eumecmum WBHmZ coflumw>mo new: oumocmum MUdAm HmomonumHQImqmm ZO Bmwmdfi ZHEBHZ m04m mom mZOHB¢H>mD QMdQZflBm QZd mZ¢m2 AQMO m.v Manda Results of the Test of Hypothesis 5 It was hypothesized that there would be a signifi- cant interaction between NR and R groups on reported SD scores to specific target persons. A multivariate analysis of variance for a repeated measures procedure was applied to test Hypothesis 5. Results of the multivariate test of Hypothesis 5 are reported in Table 4.6. TABLE 4.6 Multivariate Interaction Test for NR—R Crossed with Targets Multivariate D.F. = 6 and 69 F-Ratio = 1.0093 P < 0.4266 Since no significant multivariate F-ratio was ob- tained (F=1.0009, P < 0.4266), the hypothesis of times of incarceration differences was not supported. This indicates that no statistical difference existed between NR and R groups on reported SD scores to specific target persons. Results of the Test of Hypothesisy6 It was hypothesized that there would be a signifi- cant three-way interaction effect, i.e., race by times 104 of incarceration by specific target person(s) on SD. A multivariate analysis of variance for repeated measures procedure was applied to test Hypothesis 6. The results of the test are reported in Table 4.7. TABLE 4 . 7 Multivariate Test for Three Way Interaction of Race by NR-R by Targets Multivariate D.F. = 12 and 138 F—Ratio = 2.59 p < 0.0040 Hypothesis 6 was supported. A statistically significant multivariate F-ratio (F=2.59, with 12 and 138 degrees of freedom, p < 0.0040) did result in testing the three-way interaction effect. To further identify the patterns of SD, multiple two-way analysis of variance were computed and indicated statistical significant interactions for all target persons except for SD to counselor or psychologist target (p < 0i2813). The corresponding univariate F-ratios are reported in Table 4.8 and cell means are graphically presented in Figure 4.3. 105 Houum How .h.Q m u mfimwnnomhm How .m.o mmoo.o Hmom.m mme.oea ahhm.mam mwmum coflumosom mmso.o H6s~.m mo~.mHH mmoo.smm mumscH age mooo.o meom.m www.moa Hero.maoa ooflmuso.mnu no opossum one wawfimm mmoo.o emso.m esm.mma Hmom.mme , emmum HBGOHDOOHHOU mHm~.o som~.H Noe.mma seem.mea umamoHoBOSmm HO Hoammosoo omqo.o smsa.m mmm.mma Reas.sos Homa>umasm Rhos omao.o emsv.s smm.6s sasm.amm cameos OHMEGH fimmmOHU can» mmmH m m museum>flno mange cwm3umm manowum> Houum mumsqm new: mumwum>wcb “mlmz up mommy cOmuwm Homans zoom on swamoaomwalmawm How mochauo> mo mammamcm hmzuosa ounwum>flcb m.¢ mqmde 106 Figure 4.3 Graph of Cell Means on Self—Disclosure for Significant Interaction Effects (Race by Recidivist - Non-Recidivist) to Target Persons 40 35' 30 25 20 15 10‘ 5 0 Closest Inmate Friend .4. “ —Black ‘Wfiite Chicano Correctional Staff 0“““ ||“ ‘ Black White I Chicano Any Inmate ac ite C cano NR It _ quuuuuu R =I----l 107 40 35 30 25 20 15 10 5 0 Work Supervisor Black LTWhite Chicano Family and Friends on the outside '—IBlack White Chicano Educational Staff ‘O 6““ “ “‘|"‘ C‘|““ Black White Chicano Observation of the six graphs depicting the three-way interaction effects reveals the following: Closest Inmate Friend Targgp Self-disclosure to this target was greatest for the Chicano (NR) group and smallest for the White (NR) group. Both Black (NR) and Chicano (NR) groups were higher disclosures when compared to the Black (R) and Chicano (R) groups. In general SD to this target was relatively high in magnitude. Correctional Staff Target Self-disclosure to this target was greatest for the Chicano (NR) group and smallest for the Chicano (R) group. This reflects a substantial amount of variability within the Chicano race group. These differences are especially marked when compared with the Black and White race groups where very close agreement resulted. Within the NR groups the order of SD magnitude was lowest for the Black group, higher for the White group and highest for the Chicano group. Within the (R) groups the magnitude from lowest to highest was for Chicano, Black and White groups respectively. In general SD to this target was relatively low. 108 Work Supervisor Target Self-disclosure to this target was greatest for the White (R) group and smallest for the Chicano (R) group. Both Black (R) and Chicano (R) groups disclosed less when compared to the Black (NR) and Chicano (NR) groups. However, the White (R) group was higher in SD than the White (NR) group. The magnitude between the Black groups (R and NR) is small on SD. Greater differences are seen between the White groups (R and NR), and even greater differences are observed between the Chicano groups (R and NR). In general the differences among the (NR) groups (Black, White and Chicano) are slight and can be seen by the almost parallel sloPe. Familyand Friends on the Outside Target Self-disclosure to this target was highest for the Chicano (NR) group and lowest for the Chicano (R) group. Mean values are closer for the Black race group, followed by the White race group and the Chicano race group mean values are most distinct. Within the (R) groups, the Black and White group mean values are close, and the Chicano (R) group was the smallest. Within the (NR) groups, the White group was the smallest in SD magnitude, followed by the Black group and the Chicano (NR) group was highest. In general SD to this target person was relatively high. 109 Apy Inmate Target Self-disclosure to this target was highest for the White (R) group and lowest for the Chicano (R) group. Mean values are lowest for the Black race group, followed by the White race and highest for the Chicano race group. Within (R) groups SD was less for the Chicano group, followed by the Black group and the White group was highest. Within the (NR) groups, the three race groups were very similar on SD. In general SD to this target was the smallest in magnitude when compared to the other targets manifesting significant interaction effects. Educational Staff Target Self-disclosure to this target was highest for the Chicano (NR) group and smallest for the Chicano (R) group. Within the (R) groups SD was highest for the White group, followed by the Black group, and the Chicano group was least in magnitude. Within the (NR) groups SD was highest for the Chicano group, followed by the White group, and was lowest for the Black group. The graduated differences of the Black and White groups are very similar as can be seen by the near parallel line. The magnitude for the (R) (Black and White) groups are consistently higher than the (NR) (Black and White) groups. 110 The cell means and standard deviation for SD to each target person for Black, White and Chicano groups by (NR) and (R) groups are reported in Table 4.9 TABLE 4.9 Cell Means and Standard Deviations for Black, White and Chicano Groups by Recidivist, Non-Recidivist Groups on Self—Disclosure to Targets. Black (NR) White (NR) Chicano (NR) Stand. Stand. Stand. Mean Dev. Mean Dev. Mean De’v .. Closest Inmate - Friend 29.33 7.28 22.60 5.85 30.30 8.49 Work Supervisor 16.26 11.45 17.66 8.12 19.90 14.67 Counselor or Psychologist 20.53 12.93 25.26 9.70 28.50 10.70 Correctional Staff 12.86 13.56 18.00 8.47 23.10 14.23 Family and Friends on the Outside 26.80 11.77 22.93 9.91 35.80 5.69 Any Inmate 11.66 9.72 11.26 8.43 13.80 16.12 Educational Staff 12.80 12.08 15.86 8.50 26.80 15.88 111 TABLE 4.9 (Cont.) Black (R) White (R) Chicano (R) Stand. Stand. Stand. Mean Dev. Mean Dev. Mean Dev. Closest Inmate Friend 23.13 8.74 28.66 11.30 25.10 9.32 WOrk Supervisor 13.40 11.35 23.26 13.25 9.20 8.37 Counselor or Psychologist 23.60 7.93 28.46 12.02 22.30 14.03 Correctional Staff 11.73 12.09 19.26 13.92 4.00 3.05 Family and Friends on the Outside 28.66 9.47 28.60 11.82 16.60 10.56 Any Inmate 9.00 10.38 16.60 13.45 3.20 3.08 Educational Staff 15.93 10.68 21.73 13.64 9.40 9.77 112 Results of the Test of Hypothesis 7 The study sought to identify measures of association and relationship among and between variables dealing with sensitivity to openness and closeness as reported by the SDI and by use of a behavioral measurement of proxemic area. It was hypothesized that total SD scores would have a significant negative correlation with proxemic area for the Chicano, Black and White group. Table 4.10 reveals a non-significant correlation between total SD scores and proxemic area for the Black and Chicano groups, and total §§ when treated as a collective. However, the hypothesis was supported for the White group, which had a significant negative correlation of -.427. All groups were tested with a two-tailed test and an alpha level set at .05. 113 TABLE 4.10 Correlation of Race Groups Total Self-Disclosure score with Proxemic Area Variable Ss '57 = 80) Black Group (N = 30) White Group (N = 30 Chicano Group (N = 20) Correlation -.128595 -.101981 -.427890 .507047 Critical Value .183 .306 .306 .378 Decision Sign Alpha Level = .05 (two tailed test) Results of the Test of Hypothesis 3 A secondary aspect of this investigation was to obtain a measure of association between reported total SD scores and a score on the Paranoia scale of the MMPI. It was hypothesized that the correlation between these two variables would be negative for each race group and hence support Jourard's (1959a) notion that $0 is an important index of healthyvpersonality. Although the results showed no significant correlations, all correlations were low in magnitude, with the Black 114 group displaying a low negative correlation. The possibility of insufficient power must be adknow- ledged, in that the MMPI data was not available for all §§3 The actual correlational results can be seen in Table 4.11. TABLE 4 . ll Correlation of Subjects and Race Group- Total Self-Disclosure Score with MMPI Paranoia Scale ‘Source Correlation Critical Value Decision 9.9. (N = 46) .1837 .243 N.S. Chicano (N = 10) .3069 .549 N.S. White (N = 20) .1153 .378 N.S. Black (N = 16) —.0328 .426 N.S. Alpha Level = .05 tho tailed test) Results of Tests of Hypotheses 9 & 10 A replication of the procedure used by Newman and Pollack (1973) yielded a behavioral measurement of proxemic area. It was hypothesized that levels of the independent variables would yield main effects as well as interaction effects. A univariate analysis 115 of variance procedure was applied to test Hypotheses 9 & 10. Table 4.12 shows a statistically signifi— cant difference (F=4.44, with 2 and 74 degrees of freedom, P < 0.0152) among the race levels of the independent variable. Thus, Hypothesis 9 was supported. TABLE 4.12 Univariate Tests for Differences between Race, Recidivist — Non-Recidivist and Interaction on Total Proxemic Area Source Mean Squagg 9:. F-Ratio P less than Decision Chicano, White and 4186.3104 2 4.44 .0152 Sign Black Groups NR X R 1702.0125 1 1.80 .1833 N.S. Race X NR-R 1066.5104 2 1.13 .3283 N.S. Error Term 942.99 74 Significant of Alpha = .05 No differences were observed between the NR and R groups, and no interaction effects were present. A Scheffe post hoc analysis was computed to identify the direction and magnitude of the race group differences. The post hoc analysis revealed that the magnitude 116 and direction of the differences came from the con- trast between the White and Chicano groups, where significant mean differences were at least —3.358 and at most -47.84l. All post hoc simple contrasts can be seen in Table 4.13. TABLE 4.13 Scheffe Post Hoc Analysis for Race Group Mean Differences on Total Proxemic Area \p i S‘Variance (‘1’) Where: . (l—a) S = (J‘l), F N—J J~1, W - Contrast of interest 21: 44.53 (Blacks) 22: 39.5 (Whites) i3: 65.1 (Chicanos) Results VI = 21 ~ 22 = -14.858 < 21 - 82 < 24.924 12 = g2- 23 = -47.841 < 22 — 23 < ~3.358* W3 = i3" 21 = -42.303 < 21 _ 23 < 1.67447 *Significant statistical differences where i i 1 0 2 ” 3 ‘ Figure 4.4 represents an overhead view of reported proxemic area for Black, White and Chicano groups. 117 Figure 4.4 .4. I “‘0‘ -"l ‘\ 1‘ 49 4 ‘\ T 0 ,0 3 E 2 1 I - 2 s 7 I 514 2 1 s 1 4 s 5 7 8 \ 1 .9 \‘ 2 2 _> - .1 1...” it - <~'4;15/ 5 6 Eran! 8 (Each square equals 1 square foot) Black — White munuuuunluuunuununuu' Chicano ‘I—Illl_llll_ Results of the Test of Hypothesis 11 It was hypothesized that all §§_wou1d require greater rear proxemic distance than frontal proxemic distance. A univariate analysis of variance proced— ures was applied to test Hypothesis 11. The hypothesis was strongly suppprted (F=54.07, with l and 74 degree of freedom, P < 0.0001). 118 Thisindicates that statistical significance exists between rear and frontal proxemic distance when all §§ were treated as one group. The results of non~ hypothesized interaction effects resulted in no significant differences. The corresponding test- Hypothesis 11 can be seen in Table 4.14. TABLE 4.14 Univariate Test of Differences for all Subjects and Interactions for Groups on Rear vs. Front Proxemic Groups Source Mean Square D: F-Ratio P less than Decision 3 A11 83 Rear v§_Front 88.5062 1 54.07 .0001 Sign. Interaction of Rear vs Front 3.9062 1 .7237 .3977 N.S. by NR by R Groups Interaction of Rear VS Front 15.8177 2 2.9303 .0597 N.S. by Race Groups Interaction of Rear vs Front 2.1260 2 .393 .6759 N.S. for Race Groups by NR-R Groups Error Term 5.397973 74 Significant at Alpha LeveI—= .05 Table 4.15 presents the cell means and standard deviations for Black, White and Chicano groups by R and NR groups on total proxemic area, rear proxemic area and front proxemic area. 119 pH.H om.s mm.a os.v mm.~ 08.8 mm.H oo.m mm.a mm.m mH.H ma.m H os.a os.m mo.m oo.m m¢.~ mm.e eH.N om.v ma.a oo.m mo.~ mm.m __ W 1 ms.mm ow.~e mm.Hv ov.nm vs.mm om.om mm.sm om.mm om.m~ em.se mm.mm o¢.H¢ H . M .>oo cams .>mo cmmz .>mo ammz .>mo new: .>mo new: .>mo cmmz v .Ucnum .ocmum .vsmum .psmum .pcmum pcmsm 2 . _ mz mz mz osmoflro when; somam wosmumflo mocmumfln owmeoum usoum mocmumfla anmxonm “mom mend erwxoum Hmuoa OHmeoum ucoum paw .mocmpmfio UHmeoum umwm .mwnm owmeoum Hmuos so umfl>flpflommlsoz ma.¢ mdmdfi .umfl>fivflomm ma mmsouw osmowno paw mufinz .xomam How msoflumw>mo pumpcmum one name: damo 120 CHAPTER V SUMMARY, DISCUSSION,'CONCLUSIONS, AND RECOMMENDATIONS Summary The basic purpose of this study was to investigate self-reported differences in self-disclosure (SD), and proxemic distance parameters of Black (B), White (W) and Chicano (C), inmates; and differences between recidivist (R), and non-recidivist (NR), inmates from each of these ethnic groups. It is generally accepted in the literature of counseling and interpersonal processes that self—disclosure (SD) is one of the most important and desirable process variables. Self- disclosure apparently functions to enhance client growth, insight, and behavioral change. Self- disclosure has been treated as a process variable having many conditional contingencies which may all serve as factors affecting the timing, amount, kind, motive, and selectivity of personal disclosures. Jourard and his associates (1958, 1964) have postulated and supported the general notion that openness and transparency via self-disclosure to significant others are a necessary condition for the actualization of 121 healthy personality. As correctional administrators begin their search for answers associated with the problems and adequacy of treatment programs for the meeting of inmate needs, they must acknowledge the anxiety and stress produced by imprisonment. The occurrence of healthy verbal and nonverbal transactions between inmates and prison officials would possibly reduce this anxiety and stress. The nature and complexity of human behavior decrease the chance of accurate interpretation of deviant, or what some consider obnoxious behavior, if accurate self-disclosure is not present. Self-disclosure, which is an integral part of any rehabilitation effort, has been studied among college students, psychiatric patients, and other independent populations. However, only a very few studies have researched the interpersonal dynamics via self-disclosure where social offenders served as the target population. It is important to note that although inmates are in captivity their basic social and psychological needs are no different from the general population outside our prisons. The basic design of this study was a survey and quasi-experimental "posttest only" design recommended for educational research by Campbell and Stanley (1963). All subjects were male social offenders incarcerated at 122 the State Prison of Southern Michigan and were selected from a computer-generated printout which indicated a grade reading level of 4.9 or above for the subjects. Each subject in the study was randomly selected from his race group, with a probability of inclusion proportional to total race group size. The sample consisted of 15 Black non-recidivists (NR's), 15 Black recidivists (R's), 15 White NR's, 15 White R's, 10 Chicano NR's and 10 Chicano R's. The three dependent measures used in this evaluation were: (a) self-reported scores on a modified version of the Jourard Self-Disclosure Questionnaire, (SDI); (p) behavioral measures of proxemic distance and area need parameters; and (g) the Paranoia sub-scale score on the Minnesota Multiphasic Personality Inventory (MMPI). The data were collected over a four—week period. All §§ were administered the SDI and the proxemic test within a one-hour time frame. The investigator was cognizant of the general climate of suspicion and distrust among inmates; hence, an atmosphere of good will and rapport was initiated by the investigator in an effort to decrease faking and response set. The SDI sought to identify differentials among and within levels of race and NR-R variables for total SD scores, as well as sub-scores for the following seven disclosure targets (target persons): 123 closest inmate friend, work supervisor, counselor or psychologist, correctional staff, family and friends on the outside, an‘ inmate and education staff. The SDI requested the inmates to respond to 21 personal statements concerning how comfortable they felt or would feel about discussing or revealing the tOpic of the statement with the seven target persons. Four response options ranging from "would lie," "would not discuss," "would discuss in general terms only," to "would discuss in full and complete detail," were presented for each statement. The basis hypotheses tested in this study focused on main effects and two—and three-way interactions. The hypotheses predicted that significant differences would be observed among the three race groups and between NR-R groups. The latter hypotheses were not supported. Repeated measures analysis of SD for targets crossed with race, however, indicated significant differences. The differences were graphically displayed to further depict the patterns of SD. A hypothesized significant negative correlation between total SD and the Paranoia sub-scale score of the MMPI was not supported by the results for any of.the race groups. The investigation also resulted in statistically significant differences among the races in proxemic area. Scheffe/post hoc tests indicated that the differences 124 were attributable to differences between the Chicano and White inmate groups. A hypothesized significant negative correlation between total SD score and proxemic area for race groups was supported only by the data derived from the White group. The data strongly supported, for all gs, the anticipated greater need for rear proxemic distance as compared to frontal distance. Discussion Main Effects Multivariate analysis of variance of the data obtained in this study failed to reveal statistically significant main effects for the Black, White and Chicano groups on total SD scores. Furthermore, the data analysis failed to show statistically significant main effects for any of the hypotheses which tested differences between the NR and R groups. Hence, these null hypotheses were false. However, a number of factors may be responsible, at least in part, for the failure to reject the latter null hypotheses. These factors are discussed under the follow- ing headings: Sample and Instrumentation. Sample The definitional distinction between recidivist and non-recidivist may have functioned as an intervening variable in this study. Although the NR was defined as a person serving a prison sentence for the first time, no control was placed on the length of time served. 125 An examination of the biographical characteristics indicates that six NR subjects were serving life sentences. Also, it can be assumed that a first offender who has served 10 years of a 20 year sentence has to a large extent learned a fixed perspective and posture relative to his survival mode in prison. Clemmer's (1958) reference to the process of "prisonization" is a cogent reflection of the same phenomenon, where establishing and maintaining a new identity is linearly related to assimilation of antisocial values of the dominant prison culture. It is possible, then, that the hypothesis for rows (R vs. NR differences) was not adequately tested. However, pre-data gathering analysis of statistical power revealed a phi coefficient of .99 for the R-NR variable (Alpha = .05 with 45 §§ in each level). This indicates sufficient statistical power to reject any times of incarceration main effect hypothesis, although this study only involved 40 SE per level. Instrumentation As constructed, the SDI consists of one sub-score for each of the seven targets. The seven sub-scores are combined to yield a total SD score. However, it is important to note that the interpretation of the total SD score does not reflect one's trust or feelings about SD towards any single target, but instead toward a composite of the seven targets. 126 There are, therefore, some problems associated with the interpretation of the SDI. The value of the total SD score may not be as valuable as the seven subscores. However, the results of a pilot study, as noted in Chapter III, indicated a test re-test reliability coefficient of .85 for total SD scores. Internal consistency (odd-even) measures of reliability for the seven sub-scores were in the high 80's to low 90's. To overcome the problem relative to inter- pretation of total SD scores, Lind (1971), has suggested a follow—through procedure which focuses on analyzing the data separately for each target. Subsequent repeated measures (target) differences in this study were tested in an effort to capture race within target differences and times of incarceration within target differences. Significant interactions resulted for six of the seven targets. Interaction Effects Interpretation of the significant interaction effects (see Figure 4.1) revealed that the patterns of SD within race groups were more similar for the Black groups, followed by the White groups and Chicano groups, respectively. As anticipated, the data indicated that all race groups were willing to reveal and SD more information to their closest inmate friend, family and friends on the outside, and next, to their counselor or psychologist 127 targets. The aforementioned target persons represent significant others who are generally seen as non- threatening and safe confidante. Furthermore, the psychologist and counselor may be viewed as an alien to the correctional system, and therefore inmates are more receptive to interactions from this target person. Family and friends would undoubtedly elicit empathic understanding and greater trust in contrast with ad- ministrative personnel targets. However, the counselor or psychologist may also be viewed as a safe confidant by virtue of role and training. The latter target persons are also professionally bound to maintain the confid- entiality of personal disclosures. Hence, an inmate's risk taking and interpersonal trust for a counselor or psychologist target is of peripheral importance when compared to other administrative targets, where risk taking may be viewed as a central issue. This observa- tion is supportive of earlier research of Braithwaite (1973) and Brodsky and Komaridis (1968), where inmates were relatively reticent about self-disclosure to administrative targets. Ramsey (1972) however, found that Black and White high school subjects revealed less to trained counselors than to teachers and peers. The observed three-way interaction of race and NR-R by targets, for six of the seven targets, supports the pluralistic assumption that "what's good for the goose, may not be gOod for the gander." It has been generally accepted that individuals confronted by an 128 initiator of an interpersonal disclosure would seek to make that situation as stable or balanced as they could. This balance theory, introduced by Heider (1958), enables the individual to perform many varied transactions with others and at the same time perceive consistency in his behavior. The interaction effects evidenced by this study did not indicate any consistent pattern of SD across targets, and supports the notion of disclosure selectivity. The patterns of disclosure to work supervisor and correctional staff targets were quite similar within race groups. The patterns depict higher SD by the White groups to work supervisor and correctional staff in contrast to the Black and Chicano groups, which displayed less SD to these targets. These findings are important, especially when one considers the fact that it is the correctional staff and the work supervisors who have the most direct contact with inmates. This supports the concept that racial minority groups (Black and Chicano groups) View authoritative officials with suspicion and distrust. In Jourard's (1958) paradigm, the White groups are seen as having an advantage relative to reception of therapeutic interactions from work supervisors and correctional staff. Therefore, the foundation for "Dyadic Effect"is enhanced. Furthermore, therapeutic 129 transactions have been shown to have measurable potency regardless of the helper's formal education and training (Bergin, 1966; Levitt, 1963; Carkhuff, 1968; Poser, 1966; Lewis, 1965; Eysenck, 1965). It would seem of value to the total rehabilitation effort to improve the existing interpersonal interactions among work supervisors, correctional staff, and the inmate pOpulation. All race groups revealed less personal data to the "any inmate" target. This again reflects a suspicion and self-concealment where the target is an unknown person. These results appear to indicate a general paranoid posture that one inmate would have for any other whom he does not know. The reticence of inmates' SD to "any inmate" targets was anticipated in light of the stigma associated with inmate informers. In the prison culture, informers are seen as gross violators of an unwritten code of ethics. To summarize, it was generally found that the Chicano (NR) group was consistently higher in SD to the six targets, where interactions were present, than the Chicano (R) group. Conversely, the White (NR) group was consistently lower when compared to the White (R) group on SD to the same six targets. The Black (NR) group was consistently higher in SD to the closest inmate friend, correctional staff, work supervisor, and any inmate targets. In comparison, the Black (R) group'was 130 higher in SD to family and friends on the outside and educational staff targets. These interaction effects may serve as a basis and rationale for the initiation of specialized therapeutic community treatment programs with- in the institution that utilize specific targets of SD. Effects of Proxemic Distance This study also sought to test the hypothesis that if an inmate is a high self—discloser, in general, his concern about the closeness of physical distance of another will be inconsequential. However, it is important to note that the race of experimenter (Black) may have influenced the responses of some inmates. The analysis of the results for the White group supports the hypothesized relationship of SD having a negative cor- relation with proxemic area. Jourard's (1958) theory of the healthy personality is therefore partially supported. The results from Black and Chicano groups, however, failed to affirm the expected relationship. Failure to obtain the anticipated negative correla- tion between SD and the Paranoia sub-scale scores of the MMPI also negates support for the healthy personality theory. These findings are consistent with Persons and Marks (1970), who concluded that it is not always adaptive and healthy for inmates to be self-disclosing. It is important to note, however, the errors of measurement associated with the interpretation of the total SD scores. 131 Proxemic area may be conceived as a construct indicating fear and paranoia. Inmates who resist personal closeness of others may be avoiding inter- personal contact, and therefore neglecting real and normative psychological and social needs. This study found significant differences in proxemic area need parameters among the race groups. More Specifically, the analysis showed that the differences were most divergent between the Chicano and White groups. The Chicano group required significantly more proxemic area than the White group. Perhaps the Chicano group members feel more threatened by closeness due to the apparent language barrier experienced by some Chicano inmates. The Chicano group was highest in proxemic area need, followed by the Black group, with the White group last. This is again a reflection of caution and incredulity among racial minorities where closeness of others may be threatening and/or result in personal harm. The surreptitious possibility of a surprise attack may increase the realistic paranoia and place inmates continuously on their guard. These data may be interpreted as the Chicano group's being more defensive and on guard, followed by the Black and White groups, respectively. However, all §§ did require greater rear proxemic distance as opposed to frontal proxemic distance. This observation indicates an element of 132 fear and discomfort when approached from the rear. Newman and Pollack (1973) observed an identical relationship within normal and delinquent adolescents. Conclusions This study dealt with two design factors (Race and NR-R), which are both fixed and qualitative independent variables. The external validity of the findings is therefore applicable to similar confined populations of inmates. Caution is in order as it relates to the nuisance variable, which may have contaminated the R groups which therefore restricts the generalizability. This study failed to show any main class effects as measured by total SD scores. This failure suggests that race may not be a distinct or important variable when SD is conceptualized as a general trait. The failure to find main effects for the R-NR independent variable may be partially due to a confounding in the sample selection procedure. However, interaction effects revealed and supported the general assumption that inmates are selective in the personal data shared with others. This is evidence to substantiate that reticence by racial minority groups to self-disclosure to authoritative officials is very much a reality. Trusting and taking interpersonal risks appeared to be an easier task when the possible confidant was a peer, family member, or a trained counselor. Although 133 trusting and SD are viewed as healthy dynamics of an individual's personality, this study did not evidence the expected relationship between SD and paranoia. Therefore, support was not found for the healthy personality theory espoused by Jourard and others. Proxemic area did appear to be a major issue for the Chicano inmate group. These inmates required a mean of 65 square feet in proxemic area before arousal and discomfort. This reflects that significantly more space is required by the Chicano group than White peers. All race groups became uncomfortable when approached from the rear. This finding may indirectly reflect the intensity and fear which are prevalent in the sub— culture of a prison. Recommendations Program Development Self—disclosure did not appear to be a unilateral variable, but instead a complex variable having varied dimensions and contingent upon who was the receiver of personal disclosures. Based on the findings of this study, the following actions are recommended: 1. Initiation of a peer counseling program, involving inmates who are known high and low revealers. This program can focus on one-to— one interpersonal relations or group sessions. 2. Development of a training program for families of inmates to focus on interpersonal skills and supportive reinforcement. 134 3. Development of differential program options for Black, White and Chicano inmates. Initiation of new treatment programs should accentuate the unique needs of the three ethnic groups. Systemic Intervention l. N w o Initiation of an in-service training program for correctional employees, to focus on system analysis; with the objective of identification of system-oriented barriers which impede client growth and rehabilitation. Restructuring of the selection criteria for employees whose job calls for substantial inmate contact to include: v (a Ability to respond to inmates with acceptance and authentic feelings. (b) Ability to self-disclose personal information to inmates; and (0) Ability to confront openly, giving honest feedback. Greater efforts by correctional planners and architects to increase the size of inmate cells or perhaps provide community group living arrangements for certain inmates. Research Future studies involving SD among social offenders should approach the problem from an applied research perspective. Emphasis must be placed on the implement- ation of experimental and innovative designs which function to enhance rehabilitation outcomes so as to decrease the client population of prisons. More specifically, future research should: 1. 2. Attempt to isolate and control differences between NR and R groups. Attempt to identify subject matter differences on SD of social offenders. 135 Attempt to correlate other psychological measures of mental health status with SD. Attempt to investigate the differential effects of living quarters as they relate to the dimensions of space and territorial need. Focus on behavioral measures of self-disclosure and proxemics which are not based on self-report. Attempt to isolate differences of reported SD between counselor and psychologist target. 136 REFERENCES Albert, S., & Dabbs, J.M, Jr. Physical distance and persuasion. Journal of Personality and Social Psychology, 1970, 15, 265-270. American Psychological Association, Stuart Cook Committee. American Psychologist, 1973, 28 (l), 79. Ansbacher, R.R. Comments on White's paper. The Counseling Psychologist, 1973, 4 (2), 23. Argyle, M., & Dean, J. Eye contact, distance, and affiliation. Sociometry, 1965, 28, 289-304 Bates, B.C. Territorial behavior in primates: A review of recent field studies. Primates, 1970, 11, 271—284. Bath, K., & Daly, D. Self-disclosure relationship to self- described personality and sex differences. Psychological Reports, 1972, 31, 623-628. Baum, R.C. Self-discolsure in small groups as a function of group composition. Dissertation Abstract Interna- tional, 1972, 32 (7-B), 4200—4201. Bell, C.R. Personality characteristics of volunteers for psychological studies. British Journal of Social and Clinical Psychology, 1964, 1, 81-95. Benner, H.J. Self-disclosure as a construct. Unpublished doctoral dissertation, Michigan State University, 1968. Bergin, A.E. Some implications of psychotherapy research for therapeutic practice. Journal of Abnormal Psychology, 1966, 11, 235—246. Bogardus, E.S. Social distance scale. Sociology and Social Research, 1933, 11, 265-271. Braithwaite, R.L. A paired study of self-disclosure: Black and white inmates. Journal of Non-White Concerns 1973, l (2), 86-94. Brodsky, S.L. An inventory for measuring prisoner disclos- ure of institutionally related events. Correctional Psychologist, 1968, 3 (3), 18—20. (a) 137 Brodsky, S.L. Secrecy and self—disclosure in Criminal Justice Agencies. Unpublished paper, Southern Illinois University, 1968. (b) Brodsky, S.L. & Komaridis, G.V. Military prisonization. Military Police Journal, July 1966, 8-9. Brodsky, S.L. & Komaridis, G.V. Self—disclosure in prisoners. Psychological Reports, 1968, 22, 403-407. Brodsky, S.L. & Twomey, J. The self—disclosure scale. Unpublished manuscript, Southern Illinois University, Center for the Study of Crime, Corrections, and Delinquency, 1967. Burhenne, D. & Mirels, H. Self—disclosure in self-descrip- tive essays. Journal of Consulting and Clinical Psychology, 1970, 22, 409-413. Campbell, D.T., Kruskal, W.H. & Wallace, W.P. Seating aggregation as an index of attitude, Sociometry, 1966, 22, 1-15. Campbell, D.T. & Stanley, J.C. Experimental and quasi- experimental designs for research. Chicago: Rand McNally, 1963. Carkhuff, R.R. Differential functioning of lay and pro- fessional helpers. Journal of Counseling Psychology, 1968, 22, 117-126. Cheyne, J.A. & Efran, M.G. The effect of Spatial and inter- personal variables on the invasion of group controlled territories. Sociometry, 1972, 22 (3), 477-489. Chittick, E. & Himelstein, P. The manipulation of self- disclosure. Journal of Psychology, 1967, 65 (1), 117-121. Clemmer, D. The prison community. New York: Rinehart, 1958. Culbert, S.A. Trainer self— disclosure and member growth in a T-group. Unpublished doctoral dissertation. University of California, 1966. Dahlstrom, W.G. & Welsh, 6.8. An MMPI handbook. Minneapolis: University of Minnesota Press, 1960. 138 Deutsch, M. Critique and notes: Trust, trustworthiness, and the F scale. Journal of Abnormal and Social Psychology, 1960, 22 (1), 138-140. Deutsch, M. C00peration and trust: Some theoretical notes, Nebraska, symposium on Motivation. Lincoln: University of Nebraska Press, 1962. Dimond, R. & Hellkamp, D. Race, Sex, ordinal position of birth and self—disclosure in high school students. Psychological Reports, 1969, 22 (1), 235-238. Dorsey, M.A. & Meisels, M. Personal space and self—protec- tion. Journal of Personality and Social Psychology, 1969, 22 (2), 93-97. Doyne, S.E. The relationship between self-disclosure and self-esteem in encounter groups. Dissertation Abstract International, 1972, 22 (4-B), 1786. Drag, L. The bus rider phenomenon and its generalizability: A study of self-disclosure in student-stranger versus college roommate dyads. Dissertation Abstract International, 1972, 22 (ll-B), 6616. Drag, R.M. Experimenter behavior and group size as variables influencing self—disclosuring. Dissertation Abstract International, 1969, 22 (5-B), 2416. Edelman, R. & Snead, R. Self—disclosure in a simulated psychiatric interview. Journal of Consulting and Clinical Psychology, 1972, 22 (3), 354-358. Ellis, A. Can there be a rational concept of healthy person— ality? The Counseling Psychologist, 1973, g (2), 45-47. Eysenck, H.J. The effects of psychotherapy. International Journal of Psychiatry, 1965, 2, 99-178. Feldman, R.L. Self-disclosure patterns in the parents of stuttering children. Dissertation Abstract Inter— national, 1971, 22 (6-B), 3688. Finn, J. Multivariance: Fortran_program for univariate and multiVariate analysis of variance and co-variance. State University of New York at Buffalo, 1967. 139 Friedlander, F. The primacy of trust as a facilitator of further group accomplishment. Journal of Applied Behavioral Science, 1970, 2, 387—400. Friedman, F. Study of self-disclosure between passive and aggressive homosexuals. Unpublished master's thesis, Southern Illinois University, 1969. Fritchey, K.H. The effects of anxiety and threat on self— disclosure. Dissertation Abstract International, 1971, 22 (7—B), 4336. Fromm, E. The sane society. New York: Rinehart, 1955. Fuller, J.B. An investigation of self—disclosing behavior and the affective response within a T—group setting. Dissertation Abstract International, 1971, 22 (4—A), 1852. Gary, A.L. & Hammond, R. Self-disclosure of alcoholics and drug addicts. Psychotherapy: Theopy, Research and Practice, 1970, Z (3), 142-143. Goffman, E. Asylums, essays on the social situation of mental patients and other inmates. New York: Doubleday, 1959. Goffman, E. Stigma notes on the management of spoiled identity. Englewood Cliffs, N.J.: Prentice-Hall, 1963. Gouldner, A.W. The norm of reciprocity: A preliminary Statement. American Sociological Review, 1960, 22, 161-178. Granoff, M. An analysis of meaning and consequences of self-disclosuring behavior. Dissertation Abstract International, 1971, 22 (ll-A), 5844. it'cnSpoon, J. The reinforcing effects of two spoken sounds on the frequency of two responses. American Journal of PsychologY. 1955, 22, 409-416. Hall, E.T. A system for the notation or proxemic behavior. American Anthropologist, 1963, 22, 1003—1026. Hall, E.T. The hidden dimension. Garden City, N.Y.: Doubleday, 1966. Hathaway, S.R. Personality inventories. In B.B. welman (Ed.) Handbook of Clinical Psychology: New~Yorkz McGraw- Hill, 1965. 140 Hathaway, S.R. & McKinley, J.C. Minnesota multiphasic personality inventory manual. New York: Psycho- logical Corporation, 1951. Hayes, D.M. An alternative pregram to institutional incar- ceration for youthful offenders. Unpublished master's thesis, Michigan State University, 1973. Hediger, H. Studies of the psychology and behavior of captive animals in zoos and circuses. London: Butterworth, 1955. Heider, F. The psycholoqy of interpersonal relations. New York: Wiley, 1958. Himelstein, P. & Kimbrough W. Jr. A study of self-disclosure in the classroom. Journal of Psychology, 1963, 22, 437—440. Himelstein, P. & Lubin B. Attempted validation of the self-disclosure inventory by the peer-nomination technique. Journal of Psychology, 1965, 22, 13—16. Homans, G.C. The human group. New York: Harcourt, Brace, 1950. Homans, G.C. Social behavior: It's elementary forms. New York: Harcourt, Brace & World, 1961. Hood, T.C. & Back, K.W. Self-disclosure and the volunteer: A source of bias in laboratory experiments. Journal of Personality and Social Psychology, 1971, 21 (2), 130-136. Horney, K. The neurotic personality of our time. New York: Norton, 1936. Horney, K. Neurosis and human growth. New York: Norton, 1950. Horowitz, M.J. Spatial behavior and psychoPathology. The Journal of Nervous and Mental Disease, 1968, 146 (1), 24-350 Horowitz, M.J., Duff, D.F. & Stratton, L.O. Body-buffer zone. Archieves of General Psychiatry, 1964, 11, 651-656. —- Hurley, J. & Hurley , S. Toward authenticity in measuring self-disclosure. Journal of Counseling Psychology, 1969, 22, 271-274. - 141 Hyman, H.H. & Cobb, W.J. Interviewing in social research. Chicago: University Press, 1954. Jahoda, M. Current concepts of positive mental health. New York: Basic Books, 1958. ' Johnson, P. Jackson prison riot, SympOsium presented on the televised program Thirty Minutes, WJIM, Lansing, Michigan, January, 1974. Jones, E.E. & Gordon, E.M. Timing of self—disclosure and its effects on personal attraction. Journal of Personality and Social Psychology, 1972, 23 (3 , 358-365. Jones, L.K. Relationship between self-disclosure and posi- tive mental health, modeled self-disclosure, and socio—economic status. Dissertation Abstract Inter— national, 1972, 22 (9-A), 4953-4954. Jourard, S.M. Personal adjustment, An approach through the study of healthy personality. New York: Macmillan, 1958. Jourard, S.M. Healthy personality and self-disclosure. Mental Hygiene, 1959, 22, 499-507. (a) Jourard, S.M. Self-disclosure and other-cathexis. Journal of Abnormal Social Psychology. 1959, 22, 428-431. (13) Jourard, S.M. Religious denominations and self-disclosure. Psychological Reports, 1961, 2, 446. (a) Jourard, S.M. Self—disclosure and Rorschach productivity. Perceptual and Motor Skills, 1961, 22, 232. (b) Jourard, S.M. Self-disclosure scores and grades in nursing college. Journal of Applied Psycholong 1961, 22, 244-247. (c) Jourard, S.M. The tranSparent self. New York: Van Nostrand Reinhold, 1964. Jourard, S.M. Disclosing man to himself. Princeton: D. Van Nostrand, 1968. Jourard, S.M. The transparent self. (Rev. ed.) New York: Van Nostrand Reinhold, 1971. 142 Chaurard, S.M. & Friedman, R. Experimenter—subject "distance" and self-disclosure. Journal of Personality and Social Psychology, 1970, 22 (3), 278-282. JIDurard, S.M. & Jaffe, P.E. Influence of an interviewer's disclosure on the self—disclosing behavior of interviewers. Journal of Counseling Psychology, 1970, 22 (3), 252—257. Jraurard, S.M. & Landsman, M.J. Cognition, cathexis, and the dyadic effect in men's self—disclosing behavior. Merrill-Palmer Quarterly, 1960, 2, 178—185. Jkaurard, S.M. & Lasakow, P. Some factors in self-disclosure, Journal of Abnormal and Social PsycholOng 1958, 22, 91-98. .Jourard, S.M. & Resnick, J. Some effects of self—disclosure among college women. Journal of Humanistic Psy— chology, 1970, 22 (1), 84-93. .Jourard, S.M. & Richman, P. Factors in the self—disclosure inputs of college students. Merrill-Palmer Quarterly, IQangas, J.A. Self—disclosure in small group as a function of structure, leadership, and sex. Unpublished master's thesis, Washington State University, 1967. Kkingas, J.A. Group members' self-disclosure a function of preceding self-disclosure by leader or other group member. Comparative Group Studies, 1971, 2 (1), 65—700 ICing, M.G. Structural balance, tension and segregation in a university group. Human Relations, 1964, 21, 221—225. I(ing, M.G. Peck frequency and minimal approach distance in domestic fowl. Journal of Genetic Psychology, 1965, 106, 35-38. Ifiiing, R.D. The divided self: A study of sanity and madness. London: Tavistock, 1960. Jl-éisagna, L. & von Felsinger, J.M. The volunteer subject in research. Science, 1954, 120, 359—361. Ifiaipold, W.E. Psychological distance in a dyadic interview. Unpublished doctoral dissertation, University of North Dakota, 1963. 143 ¥ Leuchtmann, H. Self-disclosure, attitudes and sociometric choice. Dissertation Abstract International, 1969, 22 (l-B), 371-372. Levitt, E.E. Psychotherapy with children: A further evaluation. Behavior Research and Therapy, 1963, 1' 45—51. Levy, S.J. & Atkins, A.L. An empirical investigation of disclosing behavior in a verbal encounter group. Proceedings of the 79th Annual Convention of the American Fsychological Association, 1971, 2, 297- 298. (Summary) Lewis, W.W. Continuity and intervention in emotional dis- turbance: A review. Exceptional Children, 1965, 2.1;, 465—475. Lind, D.R. A study of the self—disclosure and self-presenta- tion variables. Dissertation Abstract International, 1971, 22 (2—B), 1217—1218. Lubin, B.A. A modified version of self—disclosure inventory. PsycholOgical Reports, 1965, 22, 498. Mac Donald, A.P., Jr. & Games, R.C. Film- mediated facilita- tion of self-disclosure and attraction of sensitivity training. Psychological Reports, 1972, 22 (3), 847- 857. Mac Donald, A.P., Kessel, V.S. & Fuller, J.B. Self-dis- closure and two kinds of trust. Psychological Reports, 1972, 22 (1), 143—148. Maier, R.A. Effects of stress upon social distance in dolphins. Perceptual and Motor Skills, 1968, 22, 862. (Abstract) Mann, J. The effects of reflection and race on verbal con- ditioning of affective self-disclosure in black and white males. Dissertation Abstract International, 1972, 22 (6-A), 2717-2718. Martin, M. Therapy in an anti-therapeutic environment: Systemic intervention and the prison system. Unpub— lished master's thesis, Michigan State University, 1973. 144 Martin, Martin, Maslow, R.M. & Marcuse, F.L. CharacLerisLics of volunteers and non—volunteers in psychological experimentation. Journal of ConSulting Psychology, 1958, 22, 475—479. W. Parental and situational determinants of trust. Dissertation Abstract International, 1972, 33 (6-B), 2816. __ A.H. Motivation and personality. New York: Harper and Brothers, 1954. McBride, G., James, J.W. & Shoffner, R.N. Social forces determining spacing and head orientation in a flock of domestic hens. Nature, 1963, 194, 102. Mitford, J. Kind and unusual punishment. Westminister, Maryland: Knop, 1973. Moldowski, E.W. Some judgmental effects of self-disclosure. Mowrer, MOWI' er , Mullan, Mullan, Mullan, Newman, Dissertation Abstract International, 1966, 21 (5-B), 1626. O.H. The new group therapy. Princeton: D. Van Nostrand, 1964. O.H. Commendation and a few questions. The Counsel— ing Psychologist, 1973, 2 (2), 21—22. H. Counter-transference in groups. American Journal of Psychotherapy, 1953, 1, 680—688. H. The group analyst's creative function. American Journal of Psychotherapy, 1955, 2, 320—334. (a) H. Transference and Countertransference: New Horizons. International Journal of Group Psycho— therapy, 1955, 2, 169-180. (b) R.C. & Pollack D. Proxemics in deviant adolescents. Journal of Consulting and Clinical Psychology, 1973, _4__0_ (fit 6‘8- O'Reilly, E.F., Jr. An investigation of the construct vali- dity of self—disclosure questionnaire responses using a social judgment technique. Dissertation Abstract International, 1971, 22 (4—B), 2424-2425. Panyard, C.M. Self-disclosure between friends: A validity study. Journal of Counseling Psychology, 1973, 22 (1), 66-68. 145 Pederson, D.M. & Breglio, V. Personality correlates of actual self-disclosure. Psychological Reports, 1968, 22 (2), 495-501. Pederson, D.M. & Higbee, K.L. Self-disclosure and relation- ship to the target person. Merrill—Palmer Quarterl , 1969, 22 (2), 213-220. Persons, R.W. Interpersonal intimacy with recidivists. Dissertation Abstract International, 1969, 30 (4-B), 1905. ‘— Persons, R.W. & Marks, P.A. Self-disclosure with recidivists. Optimum interviewer-interviewee matching. Journal of Abnormal Psychology, 1970, 12 (3), 387-391. Petersen, D.J. The relationship between self-concept and self—disclosure of underachieving college students in group counseling. Dissertation Abstract Inter- ggtional, 1972, 22 (5-B), 2354. Piper, W.E. Evaluation of the effects of sensitivity train- ing and the effects of varying group composition according to interpersonal trust. Dissertation Abstract International, 1972, 22 (6—B), 2819. Poser, E.G. The effect of therapists' training on group therapeutic outcome. Journal of Consulting Psycho— logy, 1966, 22, 283-289. Query, W.T. Self-disclosure as a variable in group psy— chotherapy. The International Journal of Psycho- therapy, 1964, 22, 107-116. Ramsey, G.A. Self—disclosure patterns among selected black and white high school students. Dissertation Abstract International, 1972, 22 (3-A), 975-976. Ribner, N. The effects of an explicit group contract on self-disclosure and group cohesiveness. Disserta— tion Abstract International, 1972, 32 (7-B), 4226- 4227. '— Riggs, M.M. & Kaess, W. Personality differencesbetween volunteers and non—volunteers. Journal of Psy— ghology, 1955, 22, 229-245. Rivenbark, W.H. Self-disclosure patterns among adoles- cents. Psychological Reports, 1971, 22, 35-42. 146 _¥— L Robbins, R.B. The effects of cohesiveness and anxiety on self-disclosure under threatening conditions. Dissertation Abstract International, 1966, 22 (6-B), 2145. Rodriguez, R.M. The effects of a group leader's cultural identity upon self-disclosure. Dissertation Abstract International, 1971, 22 (6-A), 3039. Rogers, C.R. Client-centered therapy. Boston: Houghton- Mifflin, 1951. Rogers, C.R. The characteristics of a helping relationship. Personnel and Guidance Journal, 1958, 22, 6-16. Rosenthal, R. Experimenter effects in behavioral research. New York: Appleton-Century-Crofts, 1966. Rosenthal, R. & Rosnow, R. The volunteers subject. In R. Rosenthal & R. Rosnow (Eds.), Artifacts in behavioral research. New York: Academic Press, 1969. Saul, L.J. Emotional maturity. (2nd. ed.) Philadelphia: J.B. Lippincott, 1947. Scheffe, H. The analysis of variance. New York: John Wiley, 1959. Scheifley, V.M. & Schmidt, W.H. Jeremy D. Finn's multi- variance- Univariate and multivariance analysis of variance, covariance, and regression modified and adapted for use on the CDC 6500. Occasional Paper No. 22, College of Education, Michigan State University, 1973. Scherz, M. Changes in self-esteem following experimental manipulation of self-disclosure and feedback con- ditions in a sensitivity laboratory. Dissertation Abstract International, 1972, 22 (4-B), 1805-1806. Shapiro, A. The relationship between self-concept and self- disclosure. Dissertation Abstract International, 1968, 22 (3-B), 1180-1181. Shapiro, J.G., Krauss, H.H. & Truax, C.B. Therapeutic conditions and disclosure beyond the therapeutic encounter. Journal of Counseling Psychology, 1969, 22 (4), 290—294. 147 IIIIIIIIIIIIIIIIIIIIIIIIIIIE:;_____________________ L ___ Shinkunas, A. Demand for intimate self-disclosure and pathological verbalization in schizophrenic. Journal of Abnormal Psychology, 1972, 22 (2), 197—205. Shostrom, E.L. Self-actualization: A scientific ethnic. The Counseling Psychologist, 1973, 2 (2), 29—34. Sinha, V. Age differences in self-disclosure. Developmental Psychology, 1972, Z (3), 257-258. Snygg, D. & Combs, A.W. Individual Behavior. New York: Harper and Brothers, 1949. Sommer, R. Studies in personal space. Sociometry, 1959, 122, 247-260. Sommer, R. Leadership and group geography. Sociometry, 1961, 22, 99—110. Sommer, R. Small group ecology. Psychological Bulletin, 1967, 21 (2), 145-152. Speilberger, C.D., Margargee, E.I. & Ingram, G.L. Graduate education. Psychologists in the Criminal Justice S stem, American Association of Correctional Psy— chologists, 1972. Stanley, G. & Bowens, A.F. Self-disclosure and neuroticism. Psychological Reports, 1966, 22, 350. Szasz, T.S. The myth of mental illness, New York: Dell, 1961. Szasz, T.S. Law, liberty, and psychiatry: An inquiry into the social uses of mental health practices. New York: Macmillan, 1963. Taylor, D.A. The development of interpersonal relationship: Social penetration processes. Journal of Social Psychology, 1968, 12, 79-90. Taylor, D.A. & Altman, I. Intimacy scaled stimuli for use in studies of interpersonal relations. Psycho- logical Reports, 1966, 22, 729—730. Taylor, D.A. & Oberlander, L. Person-perception and self- disclosure: Motivational mechanism in interpersonal processes. Journal of Experimental Research in Per- sonality, 1969, 2 (1), 14-28. 148 Thibaut, J.W. & Kelley, H.H. The social psychology of . groups. New York: Wiley, 1959. Thompson, W.N. & McAnaw, R. Politics and prison reform. Corrections Quarterly, 1972, 22 (2), 31-45. Truax, C.B. & Carkhuff, R. Client and therapist transParency in the psychotherapeutic encounter. Journal of Counseling Psychology, 1965, 22, 3-9. Truax, C.B. & Wittmer, J. Self-disclosure and personality adjustment. Journal of Clinical Psychology, 1971, 22 (4), 535-537. Ullmann,Iu & Krasner L. A_psych010gical approach to abnormal behavior. Englewood Cliffs, N.J.: Prentice-Hall, 1969. Vondracek, 8.1. The measurement and correlates of self- disclosure in preadolescents. Dissertation Abstract International, 1970, 22 (ll-B), 5230. Vosen, L.M. The relationship between self-disclosure and self-esteem. Dissertation Abstract International, 1967, 21 (8-B), 2882. Ward, W. Systemic intervention for some of the pertinent problems in the correctional institution. Unpub- lished master's thesis, Michigan State University, 1973. West, L.W. Patterns of self-disclosure for a sample of adolescents and the relationship of disclosure style to anxiety and psychological differentiation. Unpublished doctoral dissertation, University of Alberta, 1968. West, L.W. A study of the validity of the self-disclosing inventory for adolescents. Perceptual and Motor Skills, 1971, 22 (1), 91-100. White, R.W. The concept of healthy personality: What do we really mean? The Counseling Psychologist, 1973, 2 (2), 3-12. Williams, J.L. Personal space and its relation to extro- version-introversion. Unpublished master's theses, University of Alberta, 1963. 149 Worthy, M., Gary, A.L. & Kahn, C.M. Self-disclosure as an exchange process. Journal of Personality and Social Psychology, 1969, 22 (1), 59-63. 150 APPENDICES APPENDIX A Breakdown of Personal Characteristics for Inmates Committed in 1972 (State—wide) amm amuuommm uoz was mummy ca uw>o can mummy as on m mmn mamas m on n mmm. mumu>.m cu H mmma Ham» a as mead 0:02 «new Hangs monsssHemzH q<20Heommmoo 2H usua gases hon emuuommm uoz com muoaHHmnmm sacwmo has uaumacommuaa coma ucouwuuaocn omaa ousuomaoo vmhfi Hmuoa szmxemsha¢_uoomom Hma. pmunomum uoz MhH UHHOflOUHfl OHCOHSU mmv noxcauo sadness mum moamumaoe 30a sues.mumumuoz wNNN wussmuoz muss unmoamaamam uoz vase Haves AOSQUA‘ m0 mmD and nouuomum uoz NmNH coswsheuecca moon 0.3.395 Nmod mumsvant mm canmvcumaa seamen «mum deuce ezmzsmshamjsmos, mac umuuomum uoz Hm ucmw>mo servo mm ”maxemwm on Hmsxmmofio: came Hmsxmmoueuum «nee . . Haves monaiuHazH.Qom nmn musuo no em: Hmsowmmooo hmm mmaua mcwuownud mo emu voswuumam nmw mmsuo raw: cowumucuawuemxm Nmmd 0:02 the .Amuoa moama WC mm: and nuuuomum uoz pm smmnmsa unmauumuao hmm ucafiummua now runwamsowusuwumsn omm mwmosmmwa Ho sowumsam>m Hmoqmonosuhmm mmm mwmosmmwo Ho newumsam>m owuumwfivamm mean ucufiummfla uo sowumcwamxm .Hmuumuma mo huoumwmwoz anew Adana NNOFmHm.UHMHEH:UNmm Egan—"Huang "had I EMA—ENC ho 8HHWHNNHU§ 4489““ 153 mm m.N one _. o_ o. s N.P NNl m. N.mm NPmN 0.14. ¢N_m NNNNP 1wmsaocw ”62.. ON NN NN N u u u u u u u : NN Na .. ONN NN NN NN ON NN ON N ., N .:- - a .a NNN NNN u ONN N NN NON ONN ONN NN NN N N H. N u . NNN NNN u ONN N N NN NNN NON NNN NNN NNN NON NN NN u NNNN NON : ON . n N N NN NNN NNNN NNN NON ONN NN N NNNN NN . ON . a u N N NN ON . NNN NNN NNN ON ON NNN ON 1 ON . 1 a u n a N. NN NN NN NN v NN NN n O ON NNN NNN NON NNN NNN NNN NNN NNN NNN NN NN NNNN. mNNuoa NN NN ON N. N N N ,N .v . N N N mNNuoa mosouu u u u u u n s n n 1 u u u m m o N.m o n u n n n u u u n u u u n u u .o.N 8.7.832on NNNN n 9598 .aN we oauemm momma Noofim 84mg... 155 o .fium w .500 ...m.9.m mum .noum w .5800 .>sb mum .QOHm .>bb w .m.B.m mmm coaumaoum wawsm>5b mm .5500 museccmmmn mmm .EEOO .>Dh Hmnuo NNN ..m.e.m ovom msoz vmnv Hmuoa hmoumflm .HHOU mawcm>sh o .v o h m N w ON N N NN NN q ms hm m mma mma N mum mmm H gem mmmm wcoz mHHm «mnv Nance vmev mMHH Hmm mhb mam bvm awe Hod enhv N: I NN ON u NN NN . NN NN n NN 8N : NN NN u ON .ON amass NNNoe mmfluwuonusm mo NONuamuum umuNm.uN mom HNMQ‘WWI‘QDO‘ m S 0 Z Nauos mEHmB cowumnoum .uasnd mfiwozmmho NO QMOUMN MDOH>mmm menus somwum mv eyes Ho 5 an m mvH m BAN v Nmm m 000 N mOHH H mmHN mcoz vmnfi HMUOB MEMO“. H flab mBZmZBHZSOO thH I mmeZMKWO NO WUHBmHmMBUQMQmU Aézommmm .. 156 xm.mm Nm.¢ Nm.m Nm.mN No.NN No.0N &¢.mp No.mN No.0 NN.N fim.~ zomHma wqum zomxuqn mod zzoo x4 166 14. 15. l6. 17. 18. 19. 20. 21. Su Presidiario Mas Intimo Las veces cuando me he sentido solo. A A lo que le tengo mas miedo. A Lo que me disgusta mas en otros. A Las razones por lo que no 1e caigo bien a otros. A Su punto de vista en la manera en que esposo y esposa deben vivir su matrimonio. A I O / Detalles sobre su conv1cc1on D A Sentimientos sabre otros que viven aqui. A O I O I Su primer impre51on a1 entrar a la prision A 167 Instructions: For each of the statements below indicate on the answer space how comfortable you feel or would feel about relating, revealing or discussing this statement with the target or targets listed at the top of the page. Answer how comfortable_you feel in discussing these statements by checking one of the following four conditiOns: Have lied, or would lie or misrepresent myself to the target person, so that he has a false picture of me. Condition A Have not discussed and would not discuss anything about this aspect of me with the target. Condition B Condition C Have discussed or would discuss in general terms only about this item. The'target person has only a general idea about this aspect of me. Condition D Have discussed or would discuss in full and complete detail about this item to the target person, so that he could or would be able to describe me accurately. Remember You are to answer for each of the 21 statements, your feelings about discussing each statement with the target person at the top of the page. In the example below, the target person is your "closest inmate friend." Target Person "Closest Inmate Friend" Condition A Condition B Condition C Condition D Would lie or Have not and Have discussed Have or would misrepresent would not or would discuss discuss fully discuss in general only_ and completely Examplelguestions 1. How much I care about what others think of me. ' A B C D 2. Whether I like doing things alone or 1n a group. A B C D 3. The best friendship I ever had. A B C D 168 ‘¥—_ :IIIIIIIII-l Target Ferson(s) Closest Inmate Friend Condition A Condition B Condition C Condition D lV()ll III I i (. ()l‘ Il;1\/¢2 111)!; 111111 1111\/(> (i i :;(:ll:if;l‘(i ll;n\/(‘ c) l‘ Vl()ll [<1 l‘nisrepresent would not or would di::eu:;:; discuss fully (I l:€(:ll!$:; ill ([(!l|(}l‘-IJ (>111 5’ ;:I|