, ._._ 3-,,“ .--_ . fi ‘11- ' . .. . .I.I . . . 3%” . ..‘- |" ‘I‘ Ea a.-', «(I . . . I u: ‘. Mull.“ _. . . ..I. ..._\q . . . y. . "I. I \ .I’. ,l‘ ‘.I h. . 3 A“.~:. I" '91 /}l!'.. III". T‘s. '.~ is: ,‘. ‘ . V'l/t"? £72.“ $.3‘l\\ ?'§:'Ir “3" . -,:' ‘- . '1' II I“... I’I‘l' RN-‘mz: . \4 ’ II..‘ 3‘: .. . .-. - I121 (3:33 1.. II (:33: ,'.'-. ._‘I.‘ u '3' t I\' g. :I‘: . L. 9.155%} '0‘ .5... I I‘.'.'. .1 )."."| ,L'? 0. ‘. -. .._.Mo‘_ .- I.,-. "Jr? E I... IL. I - II. v V. r'“: I. '3. l‘..l.':‘ 'H ' 1" ".' '3‘ Ik-IIII I-r': I11» .-."'I'-'/"«’:- 'I n" "'II‘\'"."‘I|.'I:/"I: . IW‘.“H‘.V1‘I..II‘I‘. I I"): '5’": ’}\II..‘..|. v 'IIIIQ..‘.-I’I .I... . \I. .I . :~:--’.‘I... III-“II . . .’ ‘. ".I ' I'II. \I'. ”' ' 3* I'.'.":.'I'I.'.'I.’~:.'.'~'... ""3: ”‘5'" '4'. .InI"'r."'-M " { I‘I.'. II I f.- . H‘""|k‘1l';'lclIfi'H-‘tfl'.‘ I I' I 'th . . "..:‘..I. p I I "-1- .‘ .9. . I-,'II.II .' ." ,. ".1, I if}. 4.... .. .(I'. . ',.J' .. \" ft'l' Risi‘ "‘ I.I'I.I 'IIn" .:'-I'I I. I'.II..: . cw .. .. ‘ LZI'. ' . . .I \ 'j'.;-' I! .. I‘. I \ "f: I13“ 'Y'Nit . x ’ >89 . ..I . If? .. , .' . s...fi.rI' ‘h :I: .. y‘n"3-. I 3." 6. I 'IILT I“ I . .' .“.V \U‘ S \p \ I.:-:I .5 .I~ $3" ;;. . w...- _,._;~ I?“ . M I"? ‘I . 4.5.4:. I'f .Ig'I f!" I'M \"I ',-.".' 'I I. Y ' t -. -. . . . 'I‘. . -!..I: 3.... 32.22» '- f:'.'. {c‘IIt‘ {C‘x‘LI 3. "3- . .. .' '4‘. '.'H . ' A', I,.’."'. ‘4‘. ". ‘_I'\-. .-I .1... I }'I W53 9"“; I '.. ... .. .II .‘ '4'" .I. .'. “(.13. (31’3”... H3 . (l'k'fi $.le .. . ... . .Iv‘I .\ ‘ I.“ .IJ‘...I.1|.‘I‘ ‘rM ¢ RI”. 8‘. .1.§:.“'€(. 3’ ‘ ”59$ 0. ‘ ‘3‘ I- ' I . [IL ':...I 'IHLII. I ‘\‘,' II"I‘I HI 'I Kll‘i\1~l."(' ‘ .ufi' £15; “3.1:0“ ”I j:.’\\'.". Efié'z \'-' l . : I . . ‘... "I: '$ IL '4 LL .. ' \I U .;..iIH\1:O‘ “\I 3%“ '?"";.;h _ I‘_' '1' 2|. .II.I:I.lI'.I. I .3); .. ." .‘J ll.‘.l .W .6 .. . . .. .. . I {.15sz . . :(fu'. I...I.s.-..- . -.II».:.....I~ I1. . f.- -..-'4 . IN IJI.‘ . III .3. . II“. IIIII I.” ,.. |.I y...- ..I...I.I.... ”I... .. . I . ...I ..‘I .I' \O. '0. '93.”: .0 3. I' ":5" . 1v, .' . “I ‘. .. ._II.II. I"'...\‘I I." .waH 3...)... '5',” ' ”.x.‘ :. ' ' ' . fiz..««l‘.:;:;{4->~3; .-_ {fig/w": \ I ...I .u. Wu-flr-t .~ .' , . ‘ . _‘.I 1; 'I. 2 t ' .4 .1 .. ‘ (-3- LIE? A R Y ‘ Michigan Stat U . t)! C. This is to certify that the thesis entitled THE IMPACT OF INSTITUTIONAL MILIEU ON INMATE SELF-CONCEPT IN A CUSTODY AND A TREATMENT ORIENTED COUNTY JAIL presented by DAVID ALAN DAFFNER has been accepted towards fulfillment of the requirements for MASTER OF SCIENCE degree in CRIMINAL JUSTICE Major professor Date MARCH 3, 1978 0-7 639 ’ u g! mu; minimum in gm 1mm HIM Ill) ll ‘ JAN 2 1 195g (3,. MAR gag?” I {’c‘g; @ Copyright by DAVID ALAN DAFFNER 1978 ii THE IMPACT OF INSTITUTIONAL MILIEU ON INMATE SELF-CONCEPT IN A CUSTODY AND A TREATMENT ORIENTED COUNTY JAIL By David Alan Daffner .A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of 9 MASTER OF SCIENCE School of Criminal Justice 1978 I would like to dedicate this thesis to the Daffner clan. My Father My Mother My brother, Mitch And a very special new member, Nancy iii ACKNOWLEDGMENTS I would like to sincerely thank the chairman of my thesis committee, Dr. David Kalinich for his continuous support, patience and guidance in this project. Without his assistance this study would still be a theoretical concept on the blackboard at Olds Hall. I would also like to express my appreciation to Dr. John Hudzik for serving on my committee. His assistance with the statistical analysis was invaluable in the completion of this study. In addition, I would like to sincerely thank Dr. Robert Trojanowicz who also served on this committee. His comments were most helpful. I would also like to thank the sheriffs and admin- istrators at the two Jails in which this study was conducted. Specifically, Sheriff Charles Southworth and Undersheriff Donald Thelander at Jackson County Jail who both readily allowed this researcher access into their facility. I would also like to sincerely thank Sheriff Kenneth Pread- more and Mr. Jim Frank, the administrator at Ingham County Jail, for their complete cooperation in fully opening their facility to the researcher. Also, I would like to extend iv a note of thanks to Mr. Fred Lowe, the Ingham County Jail intake coordinator, for his assistance in the selection of subjects for this project. I would also like to thank Douglas Hathaway for his enduring assistance during the analysis of the data. A special note of thanks is also given to Vera Kean for her typing assistance under very demanding time constraints. Finally, I would like to thank all of the inmates in both institutions for their cooperation in this study. Their patience was invaluable in the completion of this thesis. Hopefully, the results of this investigation will benefit the inmates in both of these facilities. TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES. Chapter I. INTRODUCTION RELATED LITERATURE II. METHODOLOGY. A Description of the Institutions. The Method Research Questions The Research Procedure Limitations of the Study III. DATA PRESENTATION AND ANALYSIS Analysis of Inmate Data. . . . Treatment vs. Non-Treatment in Ingham. Jackson vs. Non-Treatment at Ingham. The Effect of Criminal Background on Inmate Self— Concept. One-Way Analysis of Test for Extraneous Variables. Analysis of Jail Officer Data. IV. CONCLUSIONS AND IMPLICATIONS V. PROBLEMS OF DATA COLLECTION AND IMPLICATIONS FOR FUTURE RESEARCH . APPENDIX . BIBLIOGRAPHY vi Page vii ix 3“ 3A 43 All us 65 69 69 76 78 85 93 ion 115 130 131 LIST OF TABLES Tables l. 10. ll. T—Test for Self-Concept Differences at Pre-Test, Post-Test and Differences Between Ingham and Jackson . T- Test for Self- -Concept Differences at Pre-Test and Post- Test Within Ingham and Within Jackson . . . . . . . . . Analysis of Variance for Treatment vs. Non-Treatment at Ingham . T-Test for Self-Concept Differences at Pre—Test, Post-Test and Difference Between Treatment and Non-Treatment Groups at Ingham . . . . T-Test for Self— -Concept Differences at . Pre-Test and Post- Test Within Treatment and Non-Treatment at Ingham . . . Analysis of Variance for Self-Concept Differences for Jackson and Non-Treated Inmates at Ingham . . . . . . . . . . T-Test for Self-Concept Differences for Jackson and Non-Treated at Ingham . One-Way Analysis of Variance for Ingham and Jackson at Pre-Test Controlling for Criminal Background . . . . . . . . . . . . Mean Scores for Ingham and Jackson at Post-Test Controlling for Criminal Background . One-Way Analysis of Variance for Jackson at Pre-Test Controlling for Criminal Background One-Way Analysis of Variance for Jackson at Post-Test Controlling for Criminal Background vii Page 71 72 73 74 76 77 77 79 80 81 82 Table Page 12. Mean Scores for the Difference at Jackson Controlling for Criminal Background . . . . . . 82 13. Mean Scores for the Pre-Test at Ingham Controlling for Criminal Background . . . . . . 83 1A. Mean Scores for the Post-Test at Ingham Controlling for Criminal Background . . . . . . 84 15. Mean Scores for the Difference at Ingham Controlling for Criminal Background . . . . . . 84 16. One-Way Analysis of Variance at Ingham for Pre- and Post-Test Controlling for Parents' Marital Status . . . . . . . . . . 86 17. T-Test at Ingham for Pre- and Post-Test Controlling for Parents' Marital Status . . . . 87 18. One-Way Analysis of Variance at Ingham for Pos-Test Controlling for Current Charge . . . . 88 19. T- Test at Ingham for Post-Test Controlling for Current Charge . . . . . . . . . . . . . . 89 20. T- Test at Ingham for Post- Test Within Groups Controlling for Current Charge . . . . . . 90 21. One-Way Analysis of Variance at Ingham and Jackson for Pre-Test and Post-Test Controlling for Parents' Marital Status . . . . 91 22. T-Test at Ingham and Jackson for Pre- and Post-Test Controlling for Parents' Marital Status . . . . . . . . . . . . . . . . 92 viii LIST OF FIGURES Figure Page 1. Research Design . . . . . . . . . . . . . . . . 47 2. Research Design for the Treatment and the Custody Jail . . . . . . . . . . .,. . . 47 3. Research Design for Treatment and Non-Treatment Within Ingham . . . . . . . . . 48 4. Research Alternative . . . . . . . . . . . . . . 121 5. Research Design . . . . . . . . . . . . . . . . 123 ix CHAPTER I INTRODUCTION The intention of this thesis is to investigate what impact institutional milieu has on inmate self-concept. The evaluation of institutional effects upon inmates is commonly conducted using the recidivism variable as the measurement instrument. In this study I will measure the effects of both a treatment oriented and a custody orient- ed jail on the inmateHsconcept of self. Empirical research studies on county jails have been sparse in comparison with studies conducted on penitentiaries. Therefore, there is a great need to perform a study on county jails in order to learn more about the effects of these facilities on their residents. RELATED LITERATURE The first section of this chapter will examine the history of jails beginning in England in the 12th century and continuing to the present time in the United States. This historical perspective will enable the reader to better understand current problems which, in many instances, are caused by events deeply rooted in the past. Next, self-concept itself will be defined and ex— amined from a psychological-social perspective. By measur- ing inmate self-concept, practitioners and professionals may be better able to determine program effectiveness, realizing that everything that happens to an individual in A jail is 'program' because the total environment will have some type of social and psychological effect on the inmate. Traditionally, recidivism rates have been used to determine program effectiveness but recidivism, as a measuring device, has come under increasingly critical attack due to severe, definitional and methodological deficiencies. Con- sequently, if self-concept is used as a measure in deter- mining program effectiveness, the term.must be adequately defined and understood in a psychological-social perspective. In the final section of this chapter self-concept studies which have been previously conducted in institu- tions will be examined, analyzed and two of these studies will be critiqued for methodological deficiencies. Examin- ing previous studies will enable the reader to appreciate the validity of utilizing self-concept as one measure for determining program effectiveness. Although there are approximately 3,000,000 people incarcerated in jails every year in this country there is a tremendous dearth of literature on the subject.1 There- fore, most of the studies cited in this paper will come from research conducted in prisons. There are many simi- larities between jails and prisons but there are also a great number of differences. Due to the scarcity of liter- ature in the field the author has no choice but to general— ize from and use, primarily, prison studies in reviewing the literature. This is done with a great deal of caution and trepidation. Before previous self-concept studies can be examined it will be necessary to look at the history of jails. The first known jails were constructed in 1166 under an order by Henry the II in England. The purpose of jails at that time was to detain prisoners awaiting trials in court. Jails were not used to punish inmates since punishment in this era was either corporal or capital and not retributive (by incarceration) in nature. The con- struction of jails was rapid and by the 18th century there were two hundred county and municipal jails in England.2 lRonald Goldfarb, Jails: The Ultimate Ghetto of the Criminal Justice System. (Garden City, New York: Anchor Books, 1976), p. 14. 2Henry J. Burns, Origins and Development of Jails in America. (1974). p. 2. During this period (12th to 18th century) counties were the geographic entities which controlled the communi- ties of England. Although states were developing at this time, transportation and communication were such that the country could only function at a local community level in these formative years. Consequently, jails were built in every county rather than being constructed and maintained at the state level. Thus, the development of county jails in America is a historical rather than a geographical accident. The jails (or gaols as they were called in England) did not have cells but usually contained one or two rooms which held all the suspected criminals. The jail was often located in a public building (i.e. a court house), a part of a castle or two or three dungeOns under the market house.3 Responsibility for the operation of the gaol rested on the local sheriff. The sheriff was appointed by the king and his position was one of dignity with a great deal of status but few actual duties. The sheriff contracted a keeper to actually manage the jail on a daily basis. The jail keeper was not paid a salary but the position was 3Ibid., p. 2. extremely profitable. Inmates had to purchase all of the services they received in the jail including their food, bed, matress, clothing, light, heat and 'turn key' services which had to be paid for before the prisoner could be released from the jail.“ Affluent members of the commu- nity could purchase more luxurious services in the jail including feather beds, liquor and beer, prostitutes and even private apartments outside the jail.5 New crimes were even planned in the jail and, for a fee to the keeper, inmates were allowed outside long enough to commit them and quickly return to jail.6 Most inmates, though, were unfortunate souls who had no money to pay for services and were forced to labor for the keeper in order to purchase the services that would enable them to survive in jail. Since the keeper was under no obligation to maintain the facilities (his only obligation was to prevent inmates from escaping) he did not expend any energy in maintaining a sanitary or therapuetic facility. This system of detention existed until the nineteenth century in England when convicted offenders began to be incarcerated to effect retribution. Jails began to receive both accused and convicted inmates in equal proportions. “Ibid., p. 3-5. SIbid., p. 5. Ibid., p. 5. When the early colonists came to America they brought along their old practices of criminal justice. Jails were established as soon as enough people settled in one geographic area. The purpose of the jail was once again detention instead of punishment. Incarceration as a form of punishment did not emerge until later in the nation's history.7 The establishment of the first jail in American coincides with the formation of the first colony in this country, Jamestown. Historians record that the first individuals incarcerated in the United States were, ironically, a few Indians in 1608 followed by a young German in 1609.8 For the next 150 years jails were used primarily for detention. Administrative responsibility for the jails rested with the sheriff who operated the facility in the same exploitive manner of his forefathers in England. The psoition of the jail keeper was, again, a very lucrative one and the sheriff usually awarded the job to the highest bidder. The keeper in turn, hired turnkeys, who were paid from the fees collected by the keeper. The Quaker colonies of Pennsylvania and New Jersey changed from corporal punishment to incarceration as 7Ibid., p. 7. 8Ibid., p. 8. punishment in 1718.9 In 1725 the first law creating a county jail was passed and it authorized each county to appoint a board of five county commissioners to fund a jail. Gradually jail reform in the U. S. began to take shape. For example, in 1730 new legislation prohibited the sale of any strong drink in the jail by the sheriff or any other person. But, reform was Slow and abuses contin- ued throughout the 18th and 19th centures. Robert Pursley states that during this time, "Normally no pretense was made of feeding or clothing the inmates, who were compelled to provide for their own needs. Whereas wealthy inmates might live quite comfortably, it was not unknown for prisoners who had no resources to die of starvation."lo Jails continue to be mismanaged, filthy and degrad- ing facilities even today. Since they are still operated on a local level they are frequently managed for political expediency with little regard for the rights and welfare of the inmates. Local political patronage became a part of the administration of jails very early and various con- tracts for jail services (i.e., food) are still dispensed today to political favorites. 9Ibid., p. 12. 10Robert D. Pursley, Introduction to Criminal Justice. (Glencoe Press, Encino, California: 1977), p. 372. Twentieth Century jails are constructed to house both convicted and unsentenced inmates. Thus, a paradox began to develop for modern jails when prisons began to build work and therapuetic programs for inmates and to become community oriented. Jails have remained essentially punitive in nature and do not provide unsentenced inmates the right to participate in work and therapy. The ration- ale for this phenomenon is that it would be unfair to force unconvicted inmates to work before they are sentenced. As a consequence, jail conditions today remain a disgrace and an embarrassment to modern 20th century 'civilized' man. The following are just a few examples from studies describing the current 'state of the art' in jails: (1) In 1975 the Law Enforcement Assistance Admini- stration found that out of 4,000 jails surveyed less than 4% had professional staff in such areas as social work, psychology or other treatment specialties.ll (2) In 1966, according to LEAA, only 8% of jail superintendents had a 12 (3) A 1972 survey indicated that college background. only 12% of all jails had any kind of vocational training. And, two thirds of jails offered no rehabilitation services at all (including group and individual counseling, education, 12 llIEid-, p. 382. Ibid., p. 382. Alcohol and drug related programs).13 (4) The 1975 LEAA jail census found that only 41% of all jails separated sentenced and unsentenced inmates.lu (5) The 1970 National Jail Census found that 25% of all jails were built before 1920.15 (6) The 1970 National Jail Census also found that 5% of all jails were overcrowded while many rural jails were not being used to capacity. For example, the survey found that 35 to 45% of Idaho's jails were unoccupied.l6 This report on 'the state of the art' is not a very encouraging one. Consequently empirical research must be conducted to enable professionals to suggest effective ways in which jail conditions can be improved. The National Advisory Commission on Criminal Justice Standards and Goals has made the following recommendations to alleviate some of the more salient problems that are currently associated with our nation's jails: Standard 9.3 State legislatures should immediately authorize the formulation of State standards for correctional facilities and operational procedures and State l31bid., p. 385. 1”U. S. Department of Justice, The Nation's Jails, (Washington, D. C.: U. S. Government Printing Office, 1975), p- 21- 15U. S. Department of Justice, The 1970 Jail Census, (Washington, D. C.: U. S. Government Printing Office, 1973): p. 4. 16Ibid., p. 4-5. lO inspection to insure compliance, including such features as: 1. Access of inspectors to a facility and the persons therein. 2. Inspection of: a. Administrative area, including record keeping procedures. b. Health and medical services. c. Offender's leisure activities. d. Offenders' employment. e. Offenders' education and work programs. f. Offenders' housing. g. Offenders' recreation programs. h Food service. i Observation of rights of offenders. Standard 9.6 Every jurisdiction operating locally based correctional institutions and programs should imme- diately establish these criteria for staff: 1. All personnel should be placed on a merit or civil service status, with all employees except as noted below assigned to the facility on a full-time basis. 2. Correctional personnel should receive salaries equal to those of persons with comparable qualifications and seniority in the jurisdiction's police and fire depart- ments. 3. Law enforcement personnel should not be assigned to the staffs of local correctional centers. 4. Qualifications for correctional staff members should be set at the State level and include requirement of a high school diploma. 5. A program of preservice and inservice training and staff development should be given all personnel. 11 Standard 9.7 A system of classification should be used to provide the basis for residential assignment and program palnning for individuals. Segregation of diverse categories of incarcerated persons, as well as identification of special supervision and treat- ment requirements, should be observed. Standard 9.10 A comprehensive survey and analysis should be made of criminal justice needs and projections in a particular service area. 1. Evaluation of population levels and pro- jections should assume maximum use of pre- trial release programs and posgadjudication alternatives to incarceration. If these recommendations are carried out conditions in jails should greatly improve. But, before major changes are made, research studies should be conducted to empiri- cally determine what specific changes would be most bene- ficial to the system. The study of self-concept in jails may be a very pertinent starting point in this research because it will determine the effects of the institutional milieu on the inmates' concept of self. At this point in the study a shift will be made away from examining the history and current situation of our nations jails to defining self-concept in an opera- tional manner. This must be explicitly defined to enable 17National Advisory Commission on Criminal Justice Standards and Goals, Corrections, (Washington, D. C.: U. S. Government Printing Office, 1973), p. 294-310. l2 researchers to measure change in this phenomenon over time in jails. As a theoretical concept the self has been studied, debated and analyzed since the seventh century. René’Descartes, the French mathematician and philosopher initiated this subject in which psychologists to this day find few areas of agreement. Self-concept has been defined in a variety of ways but most students of the field do agree that it referes to a persons perceptual field referred to as the "I" or "me." According to COOpersmith self-concept is defined as, "...an abstraction that an individual develops about the attri- butes, capacities, objects and activities which he posses— ses and pursues. This abstraction is presented by the symbol 'me' which is a person's idea of himself to himself. This concept is formed in the course of experience. ‘Direc- ted toward self-referent experiences, the process results in abstractions about the self, directed toward external experience, it results in abstractions about the physical d."18 and social worl According to Don E. Hamachek, author of Encounters with the Self, the manner in which we perceive ourselves is l8Russell Greiger, "Self-Concept, Self Esteem and Rational Emotive Therapy: A Brief Description," Rational Living, Vol. 10, No. 1 (Spring, 1975) p. 14. 13 extremely complicated and difficult to empirically ascer- tail." How we View ourselves is determined partially by how we perceive ourselves as we are really being, partially through how we view ourselves as ideally wanting to be, and partially through the expectations we perceive that others 19 have of us." These three processes combine to form a complex perspective that we call the self. It is very difficult to delineate between these three aspects of self but each individual possesses all three to some degree. The process of developing this sense of self takes place in every individual. According to Greenspan, "The self is something which has a development: it is not initially there, at birth, but ariSes in the process of social experience and activity, that is, developes in the given individual with that process."20 There are two basic theories on how self-concept is developed over time and the manner in which self-concept relates to anti-social behavior. .One theory, postulated by Walter Reckless, argues that a healthy self-concept will produce pro-social behavior by an individual. Another 19Don E. Hamachek, Encounters with the Self, (Chicago: Holt, Rinehart and Winston, Inc. 1971), p. 28. 20Barry Greenspan, "Differences in Self-Concept Identification by a Schizophrenic and Non-Schizophrenic," (Ph.D. dissertation, Michigan State University, 1970), p- 135. l4 theorist (i.e. Robert Culbertson et. a1.) feel that a healthy self-concept may or may not, depending on the degree of socialization, produce socially acceptable behavior. The rationale behind this position is that an individual may be totally socialized and integrated into the 'deviant' subculture and receive strong social and psychological rewards from that sub-culture when he behaves in an anti-social manner. Due to the fact that he is being positively rewarded for this behavior his self-concept will be very high. There is also the possibility that an indi- vidual in limbo between the deviant sub-culture and the normal culture would tend to have a poor self-concept which could result in deviant behavior. Both of these theories will be described in more detail and should be empirically examined in jails. Reckless and many other theorists believe that if an individual feels good about himself he will act in a good and socially acceptable manner. (Reckless empirically 'proved' this hypothesis in a study which will be describ- ed at a later point in this chapter.) Reckless states that "It is proposed that a socially appropriate or inappropri- ate concept of self and other is the basic component that steers the youthful person away from or toward delinquency and that those appropriate or inappropriate concepts rep- resent differential responses to various environments and 15 confrontations of delinquent patterns."21 Hamachek, who also believes in this perspective, states that "Healthy people, research shows, see themselves as liked, wanted, acceptable, able and worthy. Not only do they feel that they are people of dignity and worth, but they behave as though they were... It is not the people who feel that they are liked and wanted and acceptable and able who fill our prisons and mental hospitals. Rather, it is those who feel deeply inadequate, unliked, unwanted, unacceptable and unable."22 But there is now a growing number of pro- fessionals who do not agree with Reckless and Hamachek and who do not feel the above statements accurately illustrate the relationship between self-concept and delinquency. Hackler and Colbert contend that people tend to expose themselves to experiences in which they have a fair chance of success rather than those in which they may fail. Behavior that leads to withdrawal or indifference or rejec- tion will not be displayed by most individuals. Therefore, if one is raised in an environment where anti-social 21Sandra S. Tangri and Micheal Schwartz, "Delin- quency Research and the Self-Concept Variable," The Journal of Criminal Law, Criminology and Police Science, Vol. 58, No. 2, (June, 1967), p. 184. 22Don E. Hamachek, Encounters with the Self. Chicago: Holt, Rinehart and Winston, Inc., 1971), p. 251. l6 behavior is the norm (differential association) then this behavior will be reinforced by the surrounding social and psychological environment. Hackler maintains that, "The perspective presented here is that the self-concept develops out of the responses of others or, more accurately, out of the ego's perception of those responses. These self- relevant responses constantly indicate to the ego the type of person he is and what is expected of him. These self- categorizations and the concomitant perceived expectations in turn, influence roles he will seek to play in an effort to behave in ways compatible with his imagined characteris- tics and capacities."23 Both of these theories have some validity and both have been empirically tested and supported. Regardless of which theory is correct, it is important to note that experts do contend that self-concept can be altered in a manner which will produce positive behavior. Carl Rogers states that "One simple observation, which is repeated over and over again in each successful therapuetic case, seems to have rather deep theoretical implications. It is that as changes occur in the perception of self and in the 23James C. Hackler, "Boys, Blisters and Behavior-- The Impact of a Work Program in an Urban Area," The Journal of Research in Crime and Delinquency, Vol. 3, No. 2, TJuly, 1966), p. 156. l7 perception of reality, changes occur in behavior... Hence we find in therapy that as the perception of self alters, behavior a1ters."2u Now that self-concept has been defined and some of the theoretical assumptions pertaining to self-concept have been examined, previous studies can be reviewed with some degree of basic knowledge in this area. Most of the stud- ies on self-concept which have been conducted (these stud- ies have been conducted in prisons and juvenile institutions and not in jails) try to answer questions based on the theoretical assumptions that have been described above. Specifically, these studies ask the following research questions: 1. If self-concept is changed in a positive direc- tion, will behavior also change in a positive direction? 2. Are delinquents actually different from non- delinquents in their own perceptions of their self-concept? 3. Do inmates and delinquents who enter treatment oriented institutions have a greater increase in self-concept over time? 4. Is there a positive or negative relationship between delinquency orientation and the level of self—concept among juvenile delinquents? 2”Carl Rogers, "The Organization of Personality," The American Psychologist, Vol. 2 (1974), p. 361. 18 (The first and second questions will not be empirically examined in this study; the issues addressed are discussed in Chapter II. The third and fourth questions are empiri- cally examined in this thesis and detailed discussion of the research is also conducted in Chapter II.) The results of the studies that examined and utilized similar hypotheses have, in some instances, pro- duced contradictory results. The reasons for these contra- dictory findings may stem from inadequacies in the method- ological and statistical techniques. Consequently, the first study to be reviewed in this chapter will be cri- tiqued for methodological weaknesses as an example. Walter Reckless conducted the first major self- concept study and his hypothesis, methodologies and conclu- sions have since become the frame of reference from which other self-concept studies have evolved. Reckless compared a group of 'good' white sixth graders with a group of 'bad' white sixth graders in an area highly prone to delinquency to determine why some people are not delinquent even though their 'sub-culture' is delinquent. Reckless, in his hypothesis, states that, "In our quest to discover what insulates a boy against delinquency in a high delin- quency area, we believe we have some tangible evidence that a good self-concept, undoubtedly a product of favor- able socilization, veers slum boys from delinquency while a 19 poor self-concept, a product of unfavorable socialization, gives the slum boy no resistance to deviancy, delinquent companions or delinquent sub—culture."25 In order to test this hypothesis Reckless asked thirty sixth grade teachers in the highest delinquency areas of Columbus, Ohio to choose boys who would never, in their opinion, be in trouble with police or the courts. Of these nominated (N= 192) 27.3% could not be located due to summer vacations and 11.3% already had police records and were eliminated from the samples. This left 125 'good' boys who were interviewed and given three tests. Each was administered a delinquency proneness scale, a social respon- sibility scale and an occupational preference scale. The results indicated that the boys had low delinquency prone- ness, high social responsibility and very favorable percep- tions of family interaction. One year later a group of 'bad boys' (N= 108) were selected in the same manner by their teachers. Only seven boys could not be located so 101 boys were given the same three tests as the 'good' boys. The results indicated that the 'bad' boys scored 25Walter Reckless and Simon Dinitz, "Pioneering with Self-Concept as a Vulnerability Factor in Delinquency," The Journal of Criminal Law, Criminology and Police Science. Vol. 58, No._4, (December, 1967), p. 517. 20 significantly higher on the delinquency proneness scale and significantly lower on the social responsibility scale than the 'good' boys. Reckless concluded that "...good self-concept is indicative of residual favorable socialization and a strong inner self, which in turn steer the person away from bad companions and street corner society, toward middle class values, and to awareness of possibilities of upward move- ment in the opportunity structure. Conversely the poor concept of self is indicative of a residual unfavorable socialization (by 12 years probably not the result of participation in delinquency sub-culture) and indicative of weak inner direction (self or ego), which in turn does not deflect the boy from bad companions and street corner society, does not enable him to embrace middle class values, and gives him an awareness of being cut off from upward movement in the legitimate opportunity system"26 This study was a pioneer effort in determining self-concept and the study has a great number ofmethod- ological deficiencies. For example, only 11.3% of the 'good' boys had court records but current victimization studies and self-report studies indicate that in actuality this figure would be higher which would mean this group 26Ibid., p. 517. 21 would be contaminated with 'bad' boys. Another problem with this study is that the various teachers may have had a variety of reasons for conceptualizing 'good' boys. Teachers may feel differently about the actual attributes that make up a 'good' boy and a 'bad' boy. For example, in some classrooms 60% of the boys were nominated to the 'bad' boy group while nine other teachers did not nominate one boy. Thus, there may not be a perfect correlation between teachers' evaluations of boys and their actual behavior. A final problem with this tudy was that the boys from the 'good' boy group who had records were eliminated from the sample but the boys from the 'bad' boy group who did not have records (76.8%) were not eliminated. Therefore a characteristic of one group was altered while the corre- sponding characteristic of the other was not. This meant the resultant groupings were no longer comparable. Methodological deficienies may be the primary reason why some studies reach conclusions that are in con- flict with those of Reckless. This author, though, is in no position to determine which specific research studies should be considered valid or invalid. The intent in this chapter is merely to examine the studies of other research- ers in self-concept and its relationship to deviancy. 22 One research study that did not agree with Reckless's conclusions was conducted by Eitzen. He studied delinquents in a community-based halfway house that used behavior modification techniques and for a control group he tested boys from the community. Eitzen found that the boys in the halfway house had an increase in self-concept which surpassed that of the boys in the community.27 Another research study which had perplexing and unexpected results (perplexing and unexpected only because it differed from Reckless) was conducted in Israel by Sholam and Shaskolosky. This project matched 100 delin- quent boys with 100 non-delinquent boys. The boys were given the RothstienSelf—Concept Inventory, the Pd Scale of the MMPI (the Pd Scale is a factor-analyzed instrument derived from the MMPI) and the Moral Judgment Scale. The delinquents consisted of 100 consecutive court referrals to the intake department of the Tel-Aviv juvenile court. The non-delinquents were taken from the same neighborhood as their delinquent counterparts. The method of pairing delinquents and non-delin- quents left many variables uncontrolled. Non-delinquents 27Stanley D. Eitzen, "The Effects of Behavior Modification on the Attitudes of Deliqnents," Behavior Research and Therapy. Vol. 13, No. 4, (October, 1975), P- 295-299- 23 tended to be more frequently from families that had migrated to Israel from Europe, settled prior to independ- ence in 1948 and whose fathers were white collar profes- sionals. The delinquents most often had fathers who were semi-or non-skilled workers and were non-western. The results of the study indicated that there were no differ- ences (statistically significant at the .001 level) in self-concept between the delinquents and the non-delinquents on these tests. According to the researchers the differ- ences in the above variables should have resulted in statistically significant differences in scores between the two groups. The researchers had no explanation for these perplexing results.28 Another very interesting research study was con- ducted by Scarpitti who formulated the following two hypotheSes. (1) Are delinquents actually different in their perceptions of values and opportunity? (2) Are negative perceptions of values and opportunities indi- cators of a totally unhealthy self-concept or can they be seen operating somewhat independently of other self image factors? In order to test these hypotheses Scarpitti 28Shlomo Sholam and Leon Shaskolosky, "An Analysis of Delinquents and Non-Delinquents in Israel: A Cross- Cultural Perspective," Sociology and Secial Research, Vol. 53, No. 3, (April, 1969) p. 7‘8‘.‘ 24 administered a questionnaire to three groups of boys, which purported to measure their degree of socialization, (this measure indicated whether the boys were currently veering toward or away from delinquency), self-concept, opportunity and value orientation. The first group consisted of 515 inmates at the Boys Industrial School in Lancaster, Ohio. The second group consisted of 61 ninth grade boys from a lower class (non-delinquent) school in Columbus, Ohio. The final group contained 68 ninth grade boys from a middle class school in Columbus, Ohio. The findings indicate that delinquents and non- delinquents perceive middle class values differently. Delinquents tended to reject middle class values and felt they had limited opportunity to achieve rewards associated with the middle class. Lower class non-delinquents felt the same way but not to the same extent as the delinquents. Lower class non-delinquents also felt they had more oppor- tunity than the delinquents but not as much as the middle class boys. Although the lower class non-delinquents shared the same level of self-concept as the middle class boys, the lower class boys' concept of self may be the ultimate factor influencing their social behavior. Accord- ing to the researcher "Negative perceptions of middle class values and a feeling of not having access to opportunity 25 are not powerful enough to cause delinquency among lower class boys if other aspects of the personality permit the boys to see themselves as non-delinquents. The socializa- tion of the lower class non-delinquents has been more adequate probably because he has experienced a more whole- some family life and as a result has a relatively positive concept of self."29 Another study which tested hypotheses similar to Scarpitti's was conducted by Healy. Healy developed and administered a test which examined two hypotheses. (l) Delinquency orientation is a better predicator of delinquency classification than self-evaluation and (2) There is a positive relationship between delinquency ori- entation and the level of self-evaluation among juvenile delinquents. In order to test these hypotheses the researcher used four sample groups which consisted of 23 non-delinquents, 26 self-reported delinquents, 39 delin- quents on probation and 43 delinquents incarcerated in a county institution. All of the boys were between the ages of 14-16, caucasian, from the same geographic area and from blue collar families. 29R. F. Scarpitti, "Delinquents and Non-Delin- quents Perceptions of Self, Values and Opportunity, " Mental Hygine, Vol. 49, No. 3, (1965), p. 401. 26 The researchers first hypothesis was not supported. Healy found no significant statistical differences between the groups in relation to self-concept. There was, however, a linear trend of negative self-conept from the non- delinquent group to the institutionalized delinquent group. Healy did find that his second hypothesis was statistically significant. That is the higher the delinquency orientation level, the higher the level of self-conCept a boy will have.30 Thus, as stated previously, it appears as if individuals will be attracted to groups that look favorably on their behavior. In this process the delinquent will begin to identify with delinquents and will differentiate themselves from non—delinquents. Delinquents who are highly identified and affiliated with the delinquent sub- culture will have a high self-concept while those with weaker identifications and those who are in transition from one group to another will have a lower self-concept. The last three studies to be examined compare self- concept levels among inmates in treatment and custody institutions. Only one of these studies controls for the 30Peter M. Hall, "Identification with the Delin- quent Sub-Culture and Level of Self-Evaluation," Sociometery, Vol. 29, No. 2, (June, 1966), p. 147. 27 individuals affiliation with the delinquent sub-culture (by measuring the number of times the delinquent has been incarcerated in the institution) that was mentioned in the above study. The proponents of treatment oriented facili- ties argue that if a healthy self-concept can be developed then delinquent behavior can be prevented. If self-concept is reduced by incarceration in a custodial institution then delinquency may continue with increasing frequency after the individual is released from the institution. The first study, conducted by Maskin and Flesher, examined one custodial and one treatment oriented juvenile institution. Sixty boys, who were first time offenders, were given the Tennessee Self-Concept Scale when they first entered these facilities. Although the boys were matched for ethnic origin, sex, age and educational achievement they were placed in two separate correctional facilities. Thirty boys went to a program which stressed physical duties such as ranch maintenance and sporting activities. The second group of thirty boys went to a program which emphasized group and individual counseling in order to develop better interpersonal and communication skills. After 120 days the juveniles were retested and at this time it was found that the delinquents in the treatment program scored significantly higher in self-concept than those youths in the work oriented group. 28 The purpose of the second study, conducted by Robert Culbertson, was to examine the effect of institu- tionalization on juvenile Offenders' self-concepts. The researchers hypothesized that self-concept will decrease over an increasing period of time in a custody oriented institution. The study was conducted at the Indiana Boys School and the researcher utilized a cross-sectional design to conduct his experiment. The Tennessee Self- Concept Scale was used to measure self-concept and was administered to the entire population (236) of the boys at the School. The findings of this study do not support the hypothesis. There was a very slight decrease in self- concept for the entire population but it was not signifi- cnat (.05). There was, however, significantly more decrease in self-concept over time for first time offenders than for boys who were previously incarcerated only one time (in the same institution) there was no difference in self-concept over time and for boys who were incarcerated two or more times the level of self-concept was slightly raised over time.31 31Robert G. Culbertson, "The Effect of Institu- tionalization on the Delinquent Inmate's Self-Concept," The Journal of Criminal Law and Criminology, Vol. 66, No. 1, (March 1975), p. 91. 29 This study utilized a variety of techniques which increased its validity. In addition to utilizing sophisti- cated data analysis the study began with an extensive review of the literature to determine the strengths and weaknesses of previous studies. Reviewing past studies enabled the researcher to improve his own design. Another strength of the research was that an entire population was examined thereby eliminating the need to sample (which is always a possible source of error). Finally, a standard- ized set of test instructions was developed and utilized to reduce bias on the part of the college students who administered the questionnaire. There were, however, a few methodological weaknes- ses in this study. The researcher utilized a cross- sectional design in his construct and dealt with a single time frame in the institution. This was an inadequate design for the purpose because the study attempted to explain a-causal process which takes place over an extended period of time. The conclusions drawn from this study, however, were reliable for only one point in time. Although it would increase the cost of the study, a longitudinal design would have been more appropriate in determining the effects of institutionalization of juveniles over time. The final major problem with the study was that the term 'custodial institution' was not operationally defined. 30 It was assumed that the Indiana Boys School was custodial but there was no attempt to list criterion which would define it as a custodial institution or compare it with a treatment institution. In the conclusion of the study the researcher specualted that the School might be in transi- tion from a custodial to a treatment orientation. This observation may explain the absence of operational defini- tion or maybe an attempt at a justification of the failure of study results to support the hypothesis. The final self-concept study this research will examine was conducted in two extremely different prisons in Michigan, the Ionia Reformatory (custody oriented) and the Michigan Training Unit (treatment oriented). Both institutions hold youthful offenders between the ages of 15 and 23. Ionia Reformatory is work oriented and the daily routines are dull for both staff and inmates. The relationships which prevail between staff and inmates are typical of a custodial institution and are highly disci- plined, impersonal and formal. Consequently, morale among staff and inmates is generally low. On the other hand, Michigan Training Unit (M.T.U.) offers academic education, vocational training including drafting, landscaping, auto mehcanics, auto body reconditioning, machine shop, cooking and baking and carpentry) and counseling. The social 31 climate in M.T.U. is generally collaborative and more informal and personal relations exist between the staff and inmates. To test self-concept, the first 70 men sent to each of these institutions during the same two month period (for a total of 140) were administered the Tennessee Self- Concept Scale. The researcher maintained that the two groups were fairly representative, in terms of age, I.Q. and recidivism, of the population in the institution to which they were assigned. It was then hypothesized that those inmates, who were sent to M.T.U., would experience a greater raise in their self-concepts than those individuals sentenced to Ionia. Although the two groups were very different in terms of I.Q., age and recidivism it was essential that they share the same measure of self-image in the beginning of the test so that effects of the institutions and any future differences in self-concept could be measured. These two groups, in fact, did not show statistical differ- ences (t test) in self-concept which indicates that they were similar with respect to self-concept during the initial period of their imprisonment. A post-test was administered after inmates had been in the institutions for six months. No significant differences in self-concept were found after 32 Six months (at .05 level) for those inmates incarcerated in the Ionia Reformatory. There was, however, a statisti- cally significant increase in positive self-concept after six months among inmates at M.T.U. This indicates that inmate self-concept did improve in the treatment oriented institution.32 There are discrepancies and similarities between the findings of the studies described here. The most important finding in this review is that there are discrep- ancies between the conclusions of studies which attempt to measure self-concept differences in treatment and custody oriented correctional institutions. If treatment programs in jails do, in fact, increase the inmate's self-concept and result in a greater degree of pro-social behavior, then implementing treatment programs should become a top priority for every jail in the nation. If treatment pro— grams in jails do not increase inmate self-concept then rehabilitative jail programs should be reappraised and new alternatives developed for improving the effectiveness of the nation's jails. 32Bruce J. Cohen and Arthur M. Vener, "Self-Concept Modification and Total Correctional Institutions," The Journal of COrrectional Education. Vol. 20, (January, 1968) . p79. 33 The second major finding of this review is that a high degree of affiliation with a deviant sub-culture results in a high level of self-concept. This phenomenon presents an interesting problem for program development in jails. In order to overcome the problem, treatment pro- grams will first have to reduce inmate self-concept by making them realize many of their problems. For example, when the alcoholic in a jail substance abuse program real- izes that he is an alcoholic his level of self-concept will invariably be greatly reduced. Consequently, the next objective of the program will be to appropriately enhance the inmates self-concepts by providing them with the proper social, psychological and technical skills to function satisfactorily in normal society. This chapter has attempted to provide the theoret- ical framework for the research to be conducted in this study. Self-concept, Which is more reliable than recidiv- ism) will be used as a measure to empirically test the effectiveness of treatment programs in jails. Further research in this area will increase the body of knowledge associated with jails for purposes of effective development and implementation of programs. The development of treat- ment programs in jails will not be a panacea for the multi- tude of complex problems facing correctional personnel but they may effect some improvement in an area currently in a tragic state of neglect. CHAPTER II METHODOLOGY A Description of the Institutions In order to understand the nature of this study it will be necessary to describe the two settings in which it was conducted. This section will describe the characteris- tics and attributes of Ingham County Jail in Mason, Michi- gan and Jackson County Jail in Jackson, Michigan. This description will be based upon the investigator's impres- sions of the institutions after hundreds of hours of observation and reserch in the two jails. The attributes of the jails that will be examined in this section are the physical aspects of the respective institutions, guards, policy and philosophy, food and the inmates themselves. There are distinct differences in the physical nature of the two jails. Ingham County Jail is about two and a half times larger than Jackson County Jail. The jail can hold 220 inmates and the average daily population is usually close to this figure. There are three floors in the jail which are used to segregate the inmates accord- ing to age, sex and severity of offense. The lower floor contains the sentenced inmates, the middle floor holds the older (more experienced) unsentenced inmates and the top 34 35 floor houses youthful unsentenced inmates and all females. Most of the cells in the jail are tanks or bullpens and hold 12 inmates with the exception of some maximum security single cells on the main floor for high security risks. There are offices, classrooms and counseling rooms in the jail which provide adequate physical space for the variety of programs the institution offers to its residents. There seems to be a great deal of movement and activity in the jail which provide a legitimate outlet for the inmates' tensions in an environment where liberties are curtailed. There is also a yard for outdoor activity during fair weather. The concrete walls are painted light colors in one more attempt to keep spirits as high as possible in this restricted setting. Although, the physical nature of the jail is one conducive to movement, flexibility and a relatively posi- tive atmosphere, there are two obvious negative aspects of the physical structure. The first is the lack of windows in the inmates' living quarters. The jail is completely enclosed providing no view of the outside world leading to a feeling of going "stir crasy" after a short time in confinement. Second, there is no contact with visiters in the institution. Plexiglass divides the inmate and his visitor creating a source of frustration for the inmate. 36 Visits mean a great deal to the incarcerated offender and separation by a piece of plastic greatly hinders the con- comitant nature of the visit. Many of the negative aspects of Jackson County Jail stem from its physical design. The jail can contain A approximately ninety offenders with the average daily pop— ulation between 70-80 inmates. Segregation is not as efficient in Jackson as it is in Ingham County Jail. Although all females are housed on the main floor and sen- tenced and unsentenced inmates are separated, there is little attempt to segregate young inmates from older resi- dents. Also, sentenced inmates and pretrial detainees are separated on a tank by tank basis and not by floors. This unsophisticated classification system can lead to a variety of abuses among inmates and is in direct conflict with the goals of our penal system. Jackson County Jail has a library about ten feet long and eight feet wide. This is the only room for the residents to use other than the tanks. There is little Opportunity for inmate activities other than sitting in the tanks. The concrete walls are painted grey and the inmates seem to be constantly sedentary and bored. The atmosphere is one of stagnation without mobility or activity. The level of tension and anxiety appears to be 37 much higher among the inmates in Jackson than in Ingham. There are no legitimate outlets for frustration in Jackson. The institution provides no outdoor or indoor recreational facilities. One particular instance can best illustrate the harmful potentialities of this situation. One day (it was 100 degrees on the outside and hotter inside the jail) this writer was administering self-concept tests to seven inmates in the library. The tension was very high in the room and tempers began to flare. At one point, after some pushing and shoving, a fight almost broke out. This writer was locked in the room with the irratible inmates and there were no officers within calling distance (the nearest officer was one floor and three heavy steel doors away). Luckily, the inmates began to regain their composure and tensions were eased. The example, however, clearly illus- trates the pressure, under which the inmates are living, due in part, to the physical structure of the jail. It is difficult to form general impressions of the guards in the two jails because each facility has a cus- todial staff comprised of individuals with distinct person- ality differences. Generally, though, the guards at the Ingham County Jail seem to perform their jobs at a profes- sional level. On the surface there is apparently little schism between the treatment staff and the custodial staff 38 in the jail. But, when one inquires at a deeper level, there does appear to be some animosity between these staff groups. The officers seem to perform their jobs adequately. One reason for this is the pressures exerted on them to comply with the treatment goals of the institution. The administration is genuinely motivated to provide jail residents with the fullest possible treatment to help them better cope with their problems in society. Unfortunately, the position of the officers is sometimes in direct con- flict with treatment objectives. This writer overheard guards express a wish, more than once, to physically coerce troublesome inmates into submission. But, this type of response is strictly forbidden by the administration. Another more subtle example of this custody vs treatment conflict can be illustrated by describing a directive from the Sheriff to all officers in the jail. This memo informed the officers that solitary confinement can no longer be mentioned as the "hole" but must be referred to by all guards as "solitary confinement." This may seem trivial but is actually a good illustration of officers put in an ambiguous situation because their traditional role conflicts with the goals of the institution. Most of the officers, however, do seem to get along well with the residents, empa- thize with them and attempt to make their stay in the insti- tution as beneficial and comfortable as possible. 39 The guards at Jackson generally are older than the officers at Ingham. Even though they are older the offi- cers still seem responsive to the inmate's needs in jail. It was noted that a number of guards display positive attitudes toward inmates and try to help them whenever possible. There is, however, a general lifelessness in the jail. Daily operations are rigidly established and strictly adhered to. Feeding, medication dispensation, receipt and release of inmates are all very routine and there is a notable lack of activity in the jail. The officers in Jackson appear to have an easier task in adhering to their roles in the jail than the offi- cers at Ingham. There is no ambiguity. These guards are there solely to provide custodial service and that is exactly what they do. They have no pressure from the administration to refer to the "hole" as "solitary confine- ment" or to refer to "prisoners" as "inmates." An apt illustration of the difference lies in an incident that occured last July in Jackson. Officers were searching a resident's incoming mail for contraband and an older guard was observed reading aloud an inmate's letter from a girlfriend. The letter contained lucid descriptions of how she would sexually satisfy the inmate once he was released and the guard took tremendous pleasure in reading 40 this letter and seemed to feel no compunction whatever in violating the inmate's privacy. In Ingham the mail cannot be read, only searched for contraband. This example illus- trates the differences between the roles of the officers in the two jails. It seems as if the jailer at Jackson is slightly more comfortable with his role than the jailer at Ingham. There are a number of policy differences between the Ingham and Jackson County Jails. The Ingham County Jail is a nationally renowned model for treatment oriented jails. The entire orientation of the institution is to help inmates learn to solve their own problems. There are a number of rehabilitative programs which are available to most of the inmates in the facility. Currently, there is a drug program, an alcohol program, education programs, (Lansing Public Schools and Lansing Community College) and numerous Alcoholics Anonymous volunteers who conduct ses- sions in the evenings. Thus, the inmates are offered a variety of rehabilitative programs in which they can parti- cipate. In the Jackson County Jail policy is based more on the deterrent model than on the rehabilitative model. Sen— tenced offenders are merely serving their time and imprison- ment is intended to deter them from committing another crime. 41 Pretrial detainees are held only to insure that they cause no further damage in the community and to guarantee their presence in court. Thus, the jail adheres solely to its custodial function and makes no attempt at being a rehabil- itative center. Another area in which there is a substantial dif- ference between the two jails is in the quality of the food served to the inmates. While jails are not noted for fine cuisine, the food in Ingham does appear to be better than that served at Jackson. At least two hot meals a day are served Ingham residents while at Jackson lunch is the only hot meal during the day. Sandwiches are served for dinner in Jackson. Differences in the quality of food are not reflected in inmate satisfaction. Like institutiona1~ ized people anywhere, both groups of inmates bitterly com- plain about the food in their respective institutions. There was one particular point of interest in the reactions of inmates in both jails to this researcher. The inmates at Jackson were usually very willing to take the individual self-concept tests while the inmates at Ingham were usually cooperative but not to the same extent as those in Jackson. For the inmates taking the test at Jackson it was a break from the jail's unrelieved idleness and boredom. Many wanted to take the test again the next 42 time I came to the jail. They also wanted to stay and "hang around" in the small library after they completed their tests. The inmates at Ingham did not need to take the test to relieve their boredom in jail. There are always a number of activities in this jail to keep the inmates busy during the days and evenings. Thus, although most inmates at Ingham were cooperative they seemed to be a little more smug about completing thetest because they did not need to take it in order to relieve their boredom in the jail. In sum, the attributes of the Ingham County Jail are relatively consistent with those characteristic of a treatment institution. Containment in Ingham County Jail is relatively unimportant, the emphasis is on changing the inmate's attitude by altering his or her psychological condition. The individual is seen as a highly complex being with highly complex problems. The inmate in Ingham County Jail must undergo an intra-psychic change to accom- plish the goals of the program. To facilitate such change and goal accomplishment the jail offers educational pro- grams and drug and alcohol counseling in individual or group sessions. In the custody oriented jail the major emphasis is placed on the need to protect the community by containing 43 the inmates. Inmates are seen as low-level, simple and unchangeable creatures who require only routine handling. The inmate must conform to the jail's structural and organ- izational norms to succeed. Staff-inmate conflicts and other organizational attributes in the custodial setting, force the inmates to adopt modes of adjustment to survive in the jail. In Jackson County Jail the staff members are perceived as generally more hostile toward inmates than the officers at Ingham and Show relatively little inter- action with them. The Method In this thesis the intention is to investigate the effects of jail type on inmate self-concept. Jail types investigated will be treatment oriented and custody oriented institutions. The treatment oriented institution will be defined as the facility which utilized a variety of treat- ment programs. The custody oriented institution will be defined as the facility which does not have any treatment programs. Therefore, this study will examine the relation- ship between inmate self—concept and the type of jail in which he is incarcerated. Due to inconclusive and contra- dictory findings in previous research in this area, no attempt will be made to predict the results or findings of this study. Consequently, open ended research questions 44 will be formulated instead of formal hypothesis. There will be no null hypothesis since the entire eligible popu- lation in both institutions is used as subjects in the research. Research Questions The following five research questions will be examined: 1. Is there a change in inmate self-concept in a treatment vs.a custodial jail within a given time period? 2. Is there a difference in self-concept, over time in the treatment institution, between inmates who participate in the treatment programs and inmates who do not participate in the treatment programs? 3. Is there a difference in self—concept between inmates who do not participate in programs in the treatment institution and inmates of the custodial institution? 4. Is a change in self-concept related to an inmate's prior criminal background? 5. What other extraneious variables (other than the jail type itself) may have an effect on inmate self—concept? (i.e. SES, Race, Education, Carge and Sentence, Age, Sex, Family Background.) 45 Before the research procedure is outlined, terms used in the research questions will be defined. The oper- ationalization of these terms will further clarify the purpose and use of the research questions. Self-concept will be measured on the Tennessee Self-Concept Scale developed by Dr. William H. Fitts. This test measures self—concept by calculating the subjects responses to 100 questions which compriSe this examination. It has been utilized in hundreds of research studies and is ranked in the top three among the fifty self-concept tests developed in recent years. Ingham County Jail possesses many of the attributes of a treatment oriented institution. Most notably, it provides a wide variety of treatment programs which are available to its residents. Jackson County Jail possesses many of the attributes characteristic of a custody oriented institution. Most notably, Jackson does not have any treat- ment programs at this time. Time in this study will be defined as three months. Previous studies indicate that three months is sufficient time for the change agent to effectively produce its results. A treatment program is defined as inclusive of one or more of the four rehabilitative programs operated in 46 the Ingham County Jail. These programs are the drug, alco- hol, education and AA programs. If an inmate participates in a program for eight weeks or longer he is considered to have received treatment. If an inmate was not in a program for at least eight weeks or if he was never involved in a treatment program he is considered to have received no treatment. Prior criminal background will be defined as the time, in months, that an inmate was previously incarcerated in this jail, in any other county jail or in a penitentiary. The Research Procedure In order to determine which of the two jail types produced a greater change in inmate self-concept, both jails were tested and the differences between the results analyzed statistically. This was done by testing the effects of the independent variable (jail type) on the dependent variable (self-concept) while statistically and individually controlling for other independent variables (i.e. race, recidivism rates etc.). The following illus- tration (Figure 1) may clarify the method of this design: 47 FIGURE 1 RESEARCH DESIGN Dependent Variable Y1= Self-Concept Independent Variable X = Jail Type Z's = Extraneous Variables x1 46 34791 /”4~“i Z1 Z2 Z3 etc. To accomplish the objectives of this project a quasi—experimental design was developed to test inmates in both types of jails as they enter the institutions and again three months later. The samples in both groups were unmatched. The following diagram (Figure 2) illustrates this design: FIGURE 2 RESEARCH DESIGN FOR THE TREATMENT AND THE CUSTODY JAIL Pre-Test Treatment Post-Test —) Experimental Group A Tl X T2 (treatment jail) Control Group T1 T2 (custodial jail) 48 The following diagram (Figure 3) illustrates the research design used at Ingham to determine self-concept differences between those inmates who participated in treatment programs and those inmates who did not partici- pate in treatment programs: FIGURE 3 RESEARCH DESIGN FOR TREATMENT AND NON-TREATMENT WITHIN INGHAM Pre-Test Treatment Post-Test Experimental Group T1 X T2 (treatment) Control Group T1 T2 (Non-treatment) The manner by which inmate self-concept was mea- sured was the Tennessee Self—concept Scale (TSCS).33 The TSCS consists of 100 standardized statements describing the self in eight different areas. There are 45 positive items, 45 negative items and 10 items taken from the Minnesota Multiphasic Personality Inventory. Self-Concept was derived from three types of information in horizontal 33The TSCS was used in the last three self-concept studies that were cited in the review of related litera- ture. These studies were conducted by Maskin and Flesher, Culbertson and Cohen and Verner. 49 categories and five types of information in vertical categores. The three primary categories (horizontal) are identity, self-satisfaction and behavior. The five second- ary categories (vertical) are physical self, moral-ethical self, personal self, family self and social self. Fitts, the developer of the TSCS, defines these eight categories as follows: Row 1 P Score — Identity These are the, "what I am" items. Here the indi- vidual is describing his basic identity--What he is as he sees himself. Row 2 P Score - Self Satisfaction This score comes from those items where the indi- vidual describes how he feels about the self he per- ceives. In general this score reflects the level of Self satisfaction or self acceptance. Row 3 P Score - Behavior This score comes from those items that say, "this is what I do or this is the way I act." Thus this score measures the individual's perception of his own behavior or the way he functions. Column A - Physical Self Here the individual is presenting his view of his body, his state of health, his physical appearance, skills and sexuality. Column B - Moral-Tethical Self This score describes the self from a moral-ethical frame of reference--mora1 worth, relationship to God, feelings of being a "good" or "bad" person and satisfaction with one's religion or lack of it. 50 Column C - Personal Self This score reflects the individual's sense of personal worth, his feeling of adequacy as a person and evaluation of his personality apart from his body or his relationships to others. Column D - Family Self This score reflects one's feelings of adequacy, worth, and value as a family member. It refers to the individual's perception of self in reference to his closest and most immediate circle of associates. Column E - Social Self This is another "self as perceived in relation to others" category but pertains to "others" in a more general way. It reflects the person's sense of adequacy and worth in his social interaction with other people in general.3 All eight of these categories taken together from the Total Positive (P) Score. This total P Score is the major score (the score range is measured from 0-99) used to determine selfeconcept in this study. It reflects the overall level of self-esteem. According to Fitts, "persons with high scores tend to like themselves, feel that they are persons of value and worth, have confidence in them- selves and act accordingly. People with low scores are doubtful about their own worth: see themselves as undesir- able; often feel anxious, depressed and unhappy; and have little faith or confidence in themselves."35 3“William H. Fitts, ManualeTennessee Self-Concept Scale. (Nashville, Tennessee: Counselor Recording and Test, 1965), p. 2—3. 35Ibid., p. 2. 51 The scale can be used for subjects from age 12 to adult on condition that the individual has attained a sixth grade reading level. According to Fitts, the scale is "applicable to the whole range of psychological adjustments from healthy, well-adjusted peOple to psychotic patients."36 Fitts found that reliability data (based on test-retest with 60 college students over a two week period) for the total P score was .74. Establishing internal validity for the TSCS requires a great deal of empirical study and logical thought. To determine content validity two devices were used. To examine which test questions actually belonged in the scale and in each specific category, a group of five psychologists had to unanimously approve the question and its place in the test before the question could be included. Furthermore, to determine average (mean) self-concept, a standardized group on which these norms were developed was taken from a sample of 629 individuals. The sample included an equal number of both sexes, Caucasoids and Negroids, subjects from 12 to 68 years of age and equal representation from all social, economic and educational levels between the sixth grade and Ph.D. level. (The mean 361bid., p. 24. 52 score of these 629 individuals was assigned the position of 50 on a scale from 0-99.) Hopefully, this large sample was adequate to determine a valid mean self-concept. One method of determining construct validity in prison and jail studies would be to collect data on future behavior and correlate the relationship between self- concept and anti-social behavior. It can be assumed that those with high self-concept will adhere to the rules of the normal culture and those with low self-concept will commit delinquent acts. A second method to determine con- struct validity is to correlate the test results with other tests. Fitts did correlate the TSCS with a variety of other tests including the MMPI, Edwards Personal Preference Schedule and the Inventory of Feelings. According to Fitts, in most of these comparisons, "it is apparent that most of the scores on the Scale correlate with other tests in ways one would expect from the nature of the scores."37 As a final measure in determining validity Fitts assumed that self-concept in 'normal individuals' and 'psychotic individuals' should be significantly different. He performed a statistical analysis using 369 psychiatric patients and 626 non-patients. Fitts concluded there were 37Ibid., p. 10. 53 highly significant (at the .001 level) differences between the psychiatric patients and the norm group in almost every category used on the TSCS. Before the main experiment was actually conducted in this study there was a pre-test to determine if the TSCS was reliable for a jail population. Eight inmates from Ingham County Jail were given the TSCS and two weeks later were re-administered the scale. The two sets of scores in this reliability test were correlated by utiliz— ing a students t-test. The results of this t-test indicate that there were no significant differences between the pre and post-test scores for this reliability group. The t-value was .37 (P = .725) which means that the TSCS was reliable for the jail population. Next, the jail population was administered the TSCS. The pre-test was given over a one month time period. In 1976 Ingham County Jail booked approximately 600 inmates in one month. Of this total monthly figure approximately half (300) are inmates arrested for alcohol related offenses (D.U.I.L., D and D, Intoxication). Consequently, these 300 inmates probably (based on previous experience) would not have been incarcerated in the jail for three months (the time needed for the independent variable to produce effects if any, on the inmate). These inmates were not tested. 54 This left 300 subjects (sentenced and unsentenced) who would be booked into the institution during the month. To determine suitable subjects (inmates who would remain in the jail for at least the three months necessary to enable them to take the post-test) the researcher visited each of the three floors in the jail once a week for a month and examined each floor's 'information books' which indicated the name of the new inmates, the release date (if known) and the bail amount (if unsentenced). If the release date revealed that a sentenced inmate would be in the institution for at least three months, he was adminis- tered the TSCS. Unsentenced inmates presented a problem in determining who would be in the institution for three months. On advice from the jail staff the researcher used the amount of $10,000 bail as an indication that it was probable that the inmate would be in the jail for the dura- tion of the study. Bail this high meant the crime was serious and that most pre-trial detainees would not be able to post the bond. Therefore, if an inmate was sentenced for at least three months or had bail set for $10,000 or more he was given the TSCS. This method of selection divided the inmates into two relatively even groups with respect to severity of offense. The frequency distribution revealed that 12 55 inmates in the sample committed or were charged with Part I crimes while 20 inmates committed or were charged with Part II crimes. The researcher originally estimated that one hundred inmates would fall into one of these two cate— gories at Ingham. But the actual number of inmates given the pre-test was forty-eight. The same process as in Ingham was used at Jackson County Jail. Since the population was smaller in the custodial institution than in the treatment institution, a five week pre-test period was used to increase the number of inmates being tested. The researcher originally esti- mated that sixty inmates would be administered the TSCS in the five week pre-test period, in actuality twenty-six inmates were given the test. When an inmate is first incarcerated in a jail he or she is usually emotionally upset and requires a few days to adjust to the new environment. Consequently, each inmate was allowed a four to seven day adjustment period in the jail before the test was administered.38 It is difficult to determine the point at which an inmate will become acclimated to his new surroundings (i.e. a murder 381n H. S. Sandhu's study on self-concept and prison impact, all inmates were administered the pre-test within the first week of confinement. 56- suspect may be under more emotional stress than an indi- vidual recently booked for shoplifting) but hopefully this time period was sufficient for adequate adjustment. Before the inmate was given the TSCS he completed an information face sheet. This provided the researcher with useful data relating to the extranebus variables that are pertinent to this study. (An explanation of the mean- ing and use of these variables will be discussed later in this chapter.) The information face sheet included the following: 1. NAME 2. AGE 21 or under 22-31 32 and above 3. SEX Male Female Nl—J DUMP 4. RACE White Black Chicano Native American tWNI—J 5. MARITAL STATUS Married Single Separated Divorced 6. PARENTS CURRENT MARITAL STATUS Married Separated Divorced One or both parents are deceased 4:me tWNH 7. WHAT WAS THE LAST YEAR OF SCHOOL YOU COMPLETED? 57 8. WHAT WAS THE CURRENT OR LAST JOB THAT YOU HELD? 9. WERE YOU MARRIED WHEN YOU COMMITTED THIS OFFENSE?____ 10. ARE YOU A SENTENCED OR AN UNSENTENCED INMATE IN THIS JAIL? 11. WHAT IS YOUR CURRENT CHARGE? 12. HAVE YOU EVER SPENT ANY PREVIOUS TIME, EXCLUDING THIS TIME, IN THIS JAIL? 13. HAVE YOU EVER SPENT ANY PREVIOUS TIME IN ANY OTHER COUNTY JAIL? l4. HAVE YOU EVER SPENT ANY PREVIOUS TIME IN A PENITEN- TIARY? Much of the information obtained from this face sheet was re-checked by reviewing the inmate's records in jail. Since the TSCS requires that users possess a sixth grade reading level ability, all inmates had to meet this requirement in order to take the test. Any inmate who had not completed the 10th grade was asked (by the researcher) to read and interpret the first test question to determine if he could read and understand the questionnaire. If the inmate had any difficulty reading or comprehending this question he was not allowed to complete the TSCS. It was assumed that inmates who had completed the 10th grad would possess a sixth grade reading competency. 58 It was assumed that some inmates would refuse to complete the information face sheet or refuse to take the test. In the original experimental research design it was assumed that if a relatively large number (over 30%) of inmates refused to take the test, positive rewards (i.e., one pack of cigarettes) would be used as an incentive to complete the exam. (In actuality, only 16% of the selected subjects refused to take the TSCS.) A positive incentive, though, may lead to the problem of a socially desirable response set so the original plan will not be carried out in the current project. According to Dungworth, "Response set is a tendency in the subject to answer in a fixed manner, either because of an answering style that is 'agreeable' or 'disagreeable' or because of a desire to give 'socially desirable' answers. In the former case the content of the items is ignored as the subject follows his own particular pattern, agreeing or disagreeing indiscrim- inately. In the latter, the subject searches for the response which will make him look the best. The presence of either phenomena hinders meaningful measurement of the subjects attitude."39 39Terence Dungworth, "A Methodology for the Measure- ment of Political Attitudes," (M.A. Thesis, University of Utah, Department of Political Science, 1971), p. 86. 59 Response set problems did not become a major factor in the TSCS for two reasons. First, ten items on the TSCS are composed of mildly derogatory statements that most people admit are true for them. This part of the test is called the Self—Criticism Score (SC). High scores on the SC indicate a normal and healthy capacity for openness and self-criticism. (Although scores above the 99th percentile indicate the individual may be totally lacking in defenses.) Inmates who deny these statements are often being defensive and are deliberately presenting themselves in a favorable light. Consequently, if an inmate is replying in a soc- ially desirable manner this response set problem will be indicated by the Self-Criticism Score. A second reason why response set may not be a severe problem is because the scale is composed of 45 positive items and 45 negative items. Since these items are divided into two response types, they are scored in a reverse manner which hinders subject attemtps to conform to desirable response sets. There is, though, one method that could be used to further prevent the problem of response set that cannot be employed in this study. This potentially feasible but unpractical solution in this particular study, inmate anonymity, will be discussed later as a limitation of the study. 60 Three months after the first test (t1) was admin- istered, a post-test (t2) was given to determine changes in self-concept. Previous studies“0 (such as those men— tioned earlier in this thesis) have shown that three months provides sufficient time for the effects of the independent variable (jail type) to act on the dependent variable (inmate self-concept). After the inmates at Ingham com- pleted their post-test they were asked to answer a question— naire which elicited information on their activities in the jail during the previous three months (See Appendix A). This questionnaire ascertained the type and extent of involvement of inmates in the treatment and work programs. The main objective was to determine which inmates had par- ticipated in treatment programs and which had been on work details for a statistical analysis of these two variables later in the study. The residents at Jackson did not complete this questionnaire because the jail has no treat- ment or work programs. It was assumed that a high percentage of inmates (this researcher originally estimated 50%) who took the pre-test would not be in jail at the time of the post-test. “OS. H. Sandhu's investigation and the Highfiels study both used a three month time period for their respective self-concept studies. 61 Although inmate loss was originally planned for in this design, the inherent unstable nature of the jail made it extremely difficult to predict which inmates would be incarcerated for a three month period. Jail inmate popu- lations are continually in a state of flux due to transfers, pre-trial diversion projects, early or late trials, various sentencing patterns, changing policy, etc. Actually, 26 inmates were administered the pre-test at Jackson and seven the post-test. At Ingham forty-eight inmates were given the pre-test and nineteen the post-test. This repre- sented twenty-seven percent of the original group in Jackson and forty percent of the original group in Ingham. Consequently, if an inmate given the first test was no longer in the jail at the time of the second test (three months later), the results of his first test were not compared with those of other inmates who took both pre and post-test. At this point it is necessary to discuss the extraneous variables measured in the study. The follow- ing nine variables were selected by the researcher because they may have had an effect (other than the effect of the jail type) on the inmates' perception of self. 62 These variables were broken down into the follow- ing categories. 1. Criminal Background 2. SES 3. Race 4. Education 5. Charge and Sentence 6. Age 7. Sentence Status 8. Marital Status 9. Parents Marital Status Low 0-4 months Med 5-9 months High 9 or more months Blue White Professional White Black White Chicano Non-white Native American Less than high school diploma High school diploma More than H.S. diploma Part I (Index) Part II 21 and under 22 - 31 32 and over Sentenced Unsentenced Single Married Married Divorced Non-married Separated Married Divorced Married Separated Non-married One or both Parents deceased Criminal background was used to measure degree of affiliation with the deviant subculture. It was assumed that the greater the involvement in criminal activities, 63 the greater the degree of affiliation with the deviant sub- culture. This variable was calculated by combining the amount of time (in months) that the inmate has spent in this particular jail, amount of time spent in any other jail and amount of time completed in any penitentiary. Socio-economic status was used to determine the effect of social class on inmate self-concept. SES was broken down into three categories, professional, blue collar and white collar workers. The inmate was asked to state his occupation just prior to his current confinement. The U. S. Bureau of Census lists hundreds of occupations and classifies each in one of the three categories used in this project. Thus, each subject was assigned to an appro- priate category in order to ascertain his particular SES class. Race, education, age and sentence status were used to determine what effects these variables may have had on an inmate's self-concept. To ascertain what effect the severity of the charge or sentence has on inmate self- concept, this variable was measured by using Part I (Index) and Part II crimes. Each inmate was assigned to one of these categories. An inmate charged with both a Part I and a Part II crime was placed in the Part II category. The final extraneous variable measured in this study was family background. The objective here was to 64 determine if the inmate was living in a stable or unstable family environment. It is very difficult to determine this variable accurately so its results can not be regarded as entirely conclusive. The traditional way to measure family background is to determine marital status, with divorce or separation signifying an unstable family life. Although there are merits to this technique there are also a number of disadvantages. For example, a home in which both parents are living together could be plagued with turmoil while a home with just one parent could be perfectly har- 7 monious and stable. Therefore, family background will be examined but not used as conclusive evidence in determin- ing its role in the development of inmate self-concept. At this point it will be helpful to briefly outline the type of statistical analysis used for each research question. A student's t-test (used for samples smaller than thirty) was used to compare changes in self-concept in the treatment vs.the custodial institution over three months time. This t-test was also utilized to compare the differences of the pre and post-test scores within each group. A t-test, an analysis of co-variance and a one-way analysis of variance were used to measure the differences in self-concept between inmates who receive treatment and inmates who did not receive treatment at Ingham. A t-test 65 and an analysis of co-variance were used to determine differences in inmate self-concept between inmates of Ingham who did not receive treatment and all inmates of Jackson. A one-way analysis of variance was used to deter- mine if a change in self-concept was related to the inmate's criminal background. This data was analyzed by combining both institutions and by comparing the two institutions. Finally, a t-test and a one-way analysis of variance were used to determine the effects of the extraneous variables. Limitations of the Study The most serious defect in the study is that the number of inmates tested was very small. This very small N means that the results and findings are not as strong as they would have been with a larger N. The size of the variable renders conclusive statements difficult, if not impossible. This does not invalidate the findings, but suggests that the results should not be taken as definitive. The findings cannot be interpreted as fact and must be examined with knowledge that the results may not accurately reflect the real life situation in the two institutions. The reasons for this small N will be explored in Chapter IV. The mistakes this writer made in data collection will be examined to enable future researchers to be aware of and avoid in the future the hazards of research in jails. 66 Another problem of this design is the four to seven day adjustment period in Jail (before inmates are admin- istered the test) may not have been long enough for some subjects. Such adjustment is dependent on the individual's own unique ability to adapt to new surroundings. Also, someone charged with murder and someone else charged with non-support may need vastly different adjustment periods. It is hoped, however, that this time period did allow sufficient time for adaptation. Another possible problem with this study is that the researcher is assuming that the treatment and the custody institutions used actually do possess all the attributes necessary in a treatment and custodial insti- tution. These institutions were selected because they apparently fit the definition of such facilities as des- cribed in the beginning of this chapter. But, it is beyond the scope of this study to empirically measure the attributes and climates of these institutions. This issue will be examined in more detail in Cahpter IV. Another limitation in this project is that there may have been difficulties in accurately measuring all of the variables utilized in this study. For example, it is extremely difficult to determine the 'real' level of self-concept due to problems in operationally defining 67 self-concept and due to unsophisticated and inaccurate measurement techniques. Although the TSCS is recommended more highly than most other self-concept scales (over 50 have been developed) and have been used in hundreds of studies, the scale still may have problems accurately measuring self-concept. For example, the TSCS utilizes a Likert-type answer format. A Likert Scale isia good design for determining attitudinal variables, but there are some problems with this test. A Likert Scale assumes that each item is as intense as any other item in the scale. In reality, though, each may have a different level of inten- sity. Consequently, during calculation, a variety of dif- ferent levels of items will be considered as one measure. A Likert Scale also assumes that all individuals share the same degree of agreement or disagreement; unfortunately, there are no universal definitions of these variables. The Likert-type format is only one example of the problems in accurately measuring self-concept. A further limitation in this study was that multi- variate analysis was not used to ascertain the strength of association between one or more pair of variables. A Bivarate analysis (which was used in thist study) has the disadvantage of only being able to compare the relation- ship between one independent variable and the dependent variable. Consequently this technique does not account 68 for the empirical effects of the various independent varia- bles on each other and on the dependent variable. The data analysis would have been much more effective in determining the strength of the relationship between variables if mul- tiple regression techniques were utilized in the study. Multiple regression would have been able to analyze the relationship between the dependent variable and a set of the independent variables. Therefore the various bivariate techniques that were utilized in this study only examined the relationships between the independent variable and the dependent variable while multiple regression techniques have the capability of examining the entire structure of linkages between the independent and dependent variables. In a study of this nature, that incorporated a large number if independent variables multiple regression should have been used to analyze the data. But the data in this study and the conceptualized model did not permit utilization of these techniques. A final problem with the test itself was that this researcher was unable to promise anonymity to the inmates to avoid resonse set problems. Problems of response set could have even further been reduced, if anonymity could be built into the design but names must be on the first test to facilitate a match of their pre and post-test scores. CHAPTER III DATA PRESENTATION AND ANALYSIS Analysis of Inmate Data The results of this study should not be considered conclusive because of the relatively small research sample. The number of inmates administered both the pre— and post- test in both institutions was twenty-six. Seven of these were inmates at Jackson and nineteen were inmates at Ingham. In order to have an adequate sample there should have been at least sixty subjects in a study of this nature.ul Other possible reasons for inconclusive findings in this study are that certain extraneous variables were not controlled (i.e. the effect of expected release on inmate self-concept) and the effect of the extraneous variables on each other was not taken into account in this analysis. Therefore, the results will be presented but must be looked upon as "possible indicators" of the actual situation and not definitive evidence. Further modifica- tions in research design will be needed in future study ulAs noted in Chapter I, Maskin and Flesher used sixty subjects in their study comparing treatment and custody oriented institutions. 69 70 to accurately solve the problems posed in this work. The implications for future research will be examined in the next chapter. For research question one, a t—test was used to determine if there was a significant difference in self- concept (dependent variable) between the two jail types (independent variables). The first analysis examined the relationship between these two groups at the time of the pre-test (Table l). A pre-test mean of 322.57 for Jackson and 318.42 for Ingham with a t-value of .26 (P = .80) indi- cated that there was no significant difference between these two means at the time of the pre-test. This indi- cates that inmates in both institutions had nearly the same level of self-concept at the time of the pre-test. Therefore, post-test comparisons can be made with the assurance that both groups began the three month test period at the same point. An examination of the post-test scores (Table 1) reveals that Jackson had a mean of 309.57 and Ingham had a mean of 328.26 with a t-value of -l.06 (P = .31). This finding illustrates that there is no association between the two post-test scores. In the comparison of the differ- ences of the pre- and post-test scores between the two jails we find that Jackson's mean self-concept score went 71 down -l3.00 and Ingham's went up by 9.84 (Table l). The t-value in this case -3.31 (P = .003), this finding indi- cates that there is a significant difference between the differences of the two jails pre- and post-test self- concept scores. This suggests that self-concept declines significantly over a three month time period in the custo- dial institution and increases over the same period in the treatment institution but that the increase is not statis- tically significant. Table l T-Test for Self-Concept Differences at Pre-Test, Post-Test and Differences Between Ingham and Jackson Pre-Test Post-Test Difference 3? SD '1? SD If SD Jackson 322.57 0.26 309.57 45.97 -l3.00 11.28 N=7 Ingham 318.42 39.78 328.26 37.5 9.84 23.64 N=19 t= .26 P=.801 t=-l.06 P=.3l t=3.31 P=.OO3 To further support this analysis a t-test was con- ducted on the pre- and post-test scores within each insti- tution (Table 2). Pairing pre- and post-test scores within each group is to reduce the influence of extraneous vari- ables on the variable being measured. The Jackson pre-test 72 mean was 322.57, the post-test mean 309.57 with a mean difference of 13.00. The t-value in this analysis was 3.05 (P = .023) indicating a significant difference between these two scores. In Ingham the pre-test mean was 318.42 and the post-test mean 328.26 with a mean difference of -9.84. The t-value in this instance was -l.82 (P = .086) which indicates that there was not a significant difference between these two scores. Table 2 T-Test for Self-Concept Differences at Pre-Test and Post-Test Within Ingham and Within Jackson Pre-Test Post—Test 2 SD 36 SD t Jackson 322.57 35.0 309.57 37.57 (3.05)* N: Ingham 318.42 39.79 328.26‘ 45.96 (-l.82)** N=l9 * At .023 Level of significance. ** At .086 level of significance. In sum, this analysis reveals a significant differ- ence between the self-concept scores of these two institu- tions over time. Furthermore, self-concept was signifi- cantly reduced in the custodial setting but not significantly 73 increased in the treatment setting over a three month time period. This finding indicates a significant difference between pre- and post-test scores at Jackson, but no signi- ficant difference between the pre- and post-test scores at Ingham. Treatment vs. Non-Treatment in Ingham An analysis of variance was conducted for inmates at Ingham to determine if inmates who received treatment would have a greater or lesser change in self-concept than inmates who did not receive treatment (Table 3). In this case the effects of treatment had a F-value of .293 (P = .596). Table 3 Analysis of Variance for Treatment vs. Non-Treatment at Ingham Source SS df MS F Treatment 38029.68 1 180.52 .293* * At .596 level of significance. This indicates that treatment did not have a statis- tically significant effect on inmate self-concept. This finding suggests that participation in treatment programs offered (or non-participation) will not significantly affect 74 self-concept. Although the N is small in this case, it can be inferred from the data that jail climate or the institu- tional milieu may be a more crucial variable than treatment in the jail. In order to further analyze this variable a student's t-test was conducted. It was found that there was no significant difference in self-concept for those inmates who participated in treatment programs and those inmates who did not participate in treatment programs. (Table 4). Table 4 T-Test for Self-Concept Differences at Pre-Test and Post-Test and Difference Between Treatment and Non-Treatment Groups at Ingham Pre-Test Post-Test Difference '1? SD R SD if SD Treatment 299.1 45.9 305.9 51.2 6.75 27.5 N=8 Non-Treatment 332.5 29.3 344.5 35.7 12.09 21.5 N=1l t=-1.80 P=.099 t=-l.84 P=.9l t=-.46 P=.655 It is intersting to note that inmates who received treatment had a pre-test score of 299.1 and those who did 75 not had a pre-test score of 332.4. This 34 point differ- ence was not significant but indicates that those inmates who participate in treatment programs did tend to have a lower concept of self when incarcerated. This phenomenon may mean that inmates at Ingham who receive treatment do so out of greater need for therapeutic services as their initial self-concept was found to be lower than that of inmates not receiving treatment. This may also mean that diagnostic services in the jail are effective. Treatment is provided on a voluntary basis, but each inmate is diagnosed at admission and a treatment plan is recommended if there is an apparent need for services. A further discussion of this finding will be conducted in Chapter IV. A t-test was conducted on the pre and post-test scores within each group (Table 5). For inmates who received treatment the pre-test mean was 299.1 and the post-test mean was 305.9 with a mean difference of -6.75. The t—test was -.69 (P = .51) indicating that there was not a significant difference between the two scores. For those inmates who did not receive treatment the pre-test mean was 332.45 and the post-test mean was 344.54 with a mean difference of -12.09 (P = .092) which indicates that there was not a significant difference between these two scores. Since the difference in the pre- and post-test scores for these two variables was not 76 significant, this further emphasizes the point that treat- ment itself did not affect self-concept but that the insti- tutional milieu did significantly affect self-concept. Table 5 T-Test for Self-Concept Differences at Pre-Test and Post-Test Within Treatment and Non-Treatment at Ingham Pre-Test Post-Test x SD R SD t Treatment 299.1 46.0 305.9 51.2 (-.69)* N: Non-Treatment 332.5 29.3 344.5 35.7 (-l.86)** N=ll * At .51 level of significance. ** At .092 level of significance. Jackson vs. Non-Treatment at Ingham An analysis of variance was also conducted to determine if there was a statistically significant differ- ence in self-concept between inmates who did not partici- pate in treatment programs in Ingham and inmates at Jackson which does not offer treatment programs. It can be seen from Table 6 that the F value was 7.311 (P = .016). This indicates a statistically significant differ— ence between these two groups. A t-test was performed to 77 determine the direction of this difference (Table 7). The mean at Jackson went down -l3.00 while the mean at Ingham went up 12.09 with a t-value of -3.23 at the .005 level of significance. Table 6 Analysis of Variance for Self-Concept Differences for Jackson and Non-Treated Inmates at Ingham Source SS df MS F Comparison 26452.94 1 2633.37 7.3ll* * At .016 level of significance. Table 7 T—Test for Self-Concept Differences at Jackson and Non-Treated at Ingham Difference f SD t Jackson —13.000 11.284 N=7 Ingham 12.09 21.53 (-3.23)* N=11 * At .005 level of significance. This finding indicates a statistically significant difference in self-concept between these two groups. 78 Self-concept declined in Jackson and increased in Ingham. Once again, this analysis indicates that institutional milieu has a significant effect on self—concept. The Effect of Criminal Background on Inmate Self-Concept A one-way analysis of variance was conducted to determine any statistically significant differences in self-concept between inmates based on low, medium or high affiliation with the deviant subculture. An inmate with no previous time in a jail or a penitentiary was classified in the "low" group. If an inmate had been previously incarcerated one to four months he was classified in the "medium" group and if previously incarcerated for nine or more months he was classified in the "high" group. This analysis was first performed on the groups together (Ingham and Jackson) and then on each group separately. Table 8 illustrates that at the time of the pre— test, for Ingham and Jackson, prior criminal background was a significant factor relating to inmate self-concept. (It must be reiterated that the variable is small and this analysis can only be considered suggestive.) The F ratio in this case was 4.517 (P = .027). It is important to note that not only is the difference between these three groups significant but inmates with the least identifica- tion with the deviant subculture when jailed had the 79 highest self-concepts. The medium group was in the middle range and those with the longest records had the lowest self-concept in the pre-test period. Table 8 One-Way Analysis of Variance for Ingham and Jackson at.Pre-Test Controlling for Criminal Background Source SS df MS F Criminal Background 8533.66 2 4266.83 4.517* Group Mean SD Low 345.20 34.27 (N=5) Medium 339.85 21.53 (N=7) High 300.12 35.00 (N=8) * At .027 level of significance. At the post-test period the difference between groups was not significant, having an F ratio at 2.52 (P = .11). 'It is interesting to note that this progression of high to low self-concept scores was still existent even though they were no longer significant (Table 9). The difference between the pre- and post-test scores (from this point on this variable will be referred 80 to as the "difference") was not significant, with an F ratio of .052 (P = .949). Table 9 Mean Scores for Ingham and Jackson at Post-Test Controlling for Criminal Background Group Mean 1 SD Low 350.80 36.43 N=5 Medium 344.42 40.39 N=7 High - 301.75 51.51 N=8 A one-way analysis of variance was also conducted for Jackson and Ingham separately while controlling for criminal background. Table 10 reveals that at the time of the pre-test at Jackson, prior criminal background was a significant factor in determining the inmate's self-concept. (The N in this case was very small.) The F ratio was 269.35 (P = .004). Once again the scores indicate that inmates with limited criminal back- ground had a higher self-concept at the pre-test than inmates with more lengthy criminal backgrounds. 81 Table 10 One-Way Analysis of Variance for Jackson at Pre-Test Controlling for Criminal Background Source SS df MS F Criminal Background 7003.20 2 3501.6 269.35* Group Mean SD Low 370.00 N=l Medium 323.00 3.60 N=3 High 253.00 N=1 * At .004 level of significance. Prior criminal background for inmates at Jackson during the post-test also had a significant effect on self-concept. The F ratio was 192.83 (P = .005), change in self-concept was in the same direction as in the pre- test (Table 11). The difference in pre- and post-test scores at Jackson was not statistically significant. The F ratio was .277 (P = .783). It is interesting, though not statis- tically significant, that the decrease in self-concept over the three months was progressively greater in direct rela- tionship with the length of criminal record. (Table 12). 82 Table 11 One-Way Analysis of Variance for Jackson at Post—Test Controlling for Criminal Background Source SS df MS F Criminal Background 7842.13 2 3921.06 192.83* Group Mean SD Low 361.0 N=l Medium 310.33 4.509 N=3 High 273.00 N=l * At .005 level of significance. Table 12 Mean Scores for the Difference at Jackson Controlling for Criminal Background Group Mean Difference SD Low -9.00 N=l Medium -l2.66 6.65 N=3 High -l6.00 N=l 83 At the time of the pre-test, post-test and differ- ence, criminal background was not statistically significant for the inmates at Ingham. The pre-test F ratio was 3.21 (P = .076), the post-test F ratio was 2.54 (P .12) and the F ratio for the difference was .382 (P = .69). It is interesting to note (although not statistically signifi- cant) that the inmates with moderate criminal records (middle group) had the highest self-concepts at incarceration and also three months later and experienced the greatest increase over the three months. (Table 13, 14 and 15). Table 13 Mean Scores for the Pre-Test at Ingham Controlling for Criminal Background Group Mean SD Loy-17Ll 339.0 36.19 Medium 352.5 20.53 High 306.85 31.72 N=7 84 Table 14 Mean Scores for the Post-Test at Ingham Controlling for Criminal Background Group Mean SD Low 348.25 41.55 N=4 Medium 370.00 34.86 N=4 High 311.00 47.93 N=7 Table 15 Mean Scores for the Difference at Ingham Controlling for Criminal Background Group Mean Difference SD Low 9.25 15.19 N=4 Medium 17.50 26.81 =4 High 4.14 26.72 N=7 85 One-Way Analysis Test for Extraneous Variables To determine the effects of certain extraneous variables on self-concept, a one-way analysis of variance was conducted for race, education, SES, inmate marital status and parents marital status, sentence status and current charge. This analysis is divided into three sections with three sub-division in each section. The first section examinedJEckson'sresultsand specificially analyzed pre-test effect, post-test effect and the differ- ence between pre- and post-test for each variable. The second section performed these analyses on Ingham's results. The final section performed the same three analyses but Ingham and Jackson inmates were combined into one group. If an extraneous variable had a F ratio with a significance of .05 or less, then a t-test was conducted to determine the strength and direction of the variable. In Jackson none of the extraneous variables had a statistically significant effect on inmate self-concept. This lack of significance may have been caused by the small number of subjects tested in this facility. This phenom- enon may also be attributed to the fact that the effect of the custodial institution and prior criminal background are the two primary factors having a statistically significant effect. 86 In Ingham the extraneous variables age, race, educa- tion and marital status did not have statistical signifi- cance on the pre-test, post-test or difference. The one- way analysis of variance for parents' marital status at pre-test and post-test was significant (Table 16) but the difference was not significant. Table 16 One-Way Analysis of Variance at Ingham for Pre- and Post-Test Controlling for Parents' Marital Status Pre-Test Source SS df MS F Parents'Marital Status 15320.47 3 5106.82 5.584* Post-Test Source SS df MS F Parents'Marital Status 18748.89 3 6249.63 4.447** * At .011 level of significance. ** At .023 level of significance. The F ratio for the pre-test was 5.584 (P = .011) and 4.447 (P = .023) for the post-test. Parents' marital status was coded as married and non-married and a t-test conducted on this variable to 87 determine if the values significantly affected self-concept. Table 17 illustrates that there was a pre-test mean of 345.25 for inmates with married parents and 290.33 for inmates with non-married parents with a t-value of 3.67 (P = .002). Table 17 T-Test at Ingham for Pre- and Post-Test Controlling for Parents' Marital Status Pre-Test Post-Test Y SD t '1? SD t Married 345.25 29.15 359.75 38.06 N=8 Non-Married 290.33 32.57 (3.67)* 295.55 34.31 (3.63)** * At .002 level of significance. ** At .003 level of significance. The post—test mean for married parents was 359.75 and 295.55 for non-married parents with a t-value of 3.63 (P = .003). There was no statistically significant increase in self-concept in the married group or in the non-married group. T-values were -1.84 (P = .108) and -.59 (P = .571) respectively. This may be due to the small N at Ingham. But, it may also mean inmates at Ingham with married par- ents entered the institution with a significantly higher 88 self-concept than those with non-married parents. Further- more, differences within each group were not significant but the differences between groups in pre— and post-tests were statistically significant. Those with married parents scored higher on self—concept than those with non-married parents. At Ingham the extraneous variables SES and sentence status were not statistically significant in the pre-test, post-test or in the difference. The one-way analysis of variance did conclude that current charge had a significant effect on self-concept in the post-test (Table 18). The F ratio for the post-test was 4.447 (P = .05). There was not significance on the pre-test or in the difference for current charge. Table 18 One-Way Analysis of Variance at Ingham for Post—Test Controlling for Current Charge Post-Test Source SS df MS F Current Charge 7885.26 1 7885.26 4.447* * At. 05 level of significance. 89 A t-test was conducted on current charge to determine if this variable had a significant effect upon inmate self- concept. Table 19 illustrates that for inmates who commit- ted Part I crimes in the post-test the mean self-concept was 304.37 and for inmates who committed Part II crimes the mean self-concept was 345.63 with a t-value of -2.10 (P = .05). This finding means seriousness of crime correlates positively with low self-concept. Further analysis reveals that self-concept did not change signifi- cantly over the three months for inmates who committed Part I crimes but did increase Significantly for inmates who committed Part II crimes (Table 20). Table 19 T-Test at Ingham for Post-Test Controlling for Current Charge Post-Test f SD t Part I Crimes 304.37 42.97 N=8 Part II Crimes 345.63 41.49 (-2.10)* N=ll * At .05 level of significance. 90 Table 20 T-Test at Ingham for Post-Test Within Groups Controlling for Current Charge Pre-Test Post-Test T SD T SD t Parg I Crimes 305.12 37.50 304.37 42.97 (.08)* N: . Part II Crimes 328.09 40.26 345.63 41.49 (-3.20)** N=ll * At .940 level of Significance. ** At .009 level of significance. Thus, inmates who committed Part II crimes had significantly higher self-concepts than those inmates who committed Part I crimes as measured by the post-test. Also, inmates who committed Part II crimes experienced a signifi- cant rise in self-concept during the three month period. Furthermore, inmates who committed Part I crimes did not Show a significant increase in self—concept over time. In fact, their mean self-concept went down .7500 during the period. The figure is not significant in itself but does contrast sharply with the rise in self-concept experienced by inmates who committed Part II crimes. The final section of this analysis involves Ingham and Jackson together while controlling for the extraneous 91 variables. In Ingham and Jackson the independent variables education, age, race, marital status, sentence status and current charge did not have statistical significance on the pre-test, post-test or difference. The variance for parents' marital status at the pre-test and post-test was significant but the difference was not. (Table 21). The F ratio for the pre-test was 3.130 (P = .049) and the F ratio for the post-test was 2.923 (P = .05). The F ratio for the difference was .142 (P = .934). Table 21 One-Way Analysis of Variance at Ingham and Jackson for Pre-Test and Post-Test Controlling for Parents' Marital Status Pre-Test Source SS df MS F Parents' Marital Status 11117.65 3 3705.88 3.13* Post-Test Parents' Marital Status 14202.10 3 4754.03 2.92** * At .049 level of significance. ** At .05 level of significance. Parents' marital status was recoded into married and non-married and a t-test conducted to determine if these values had a significant effect upon self-concept. 92 Table 22 illustrates that there was a pre—test mean of 342.20 for inmates with married parents and 300.78 for inmates with non-married parents with a t-value of 3.15 (P = .005). Table 22 T—Test at Ingham and Jackson for Pre- and Post-Test Controlling for Parents'Marital Status Pre-Test 1' SD t Married Parents 342.20 26.77 N=10 ' Non-Married Parents 300.78 37.56 (3.15)* N=14 Post-Test Married Parents 349.00 40.50 N=10 Non-Married Parents 301.07 37.84 (2.94)** N=l4 * At .005 level of significance. ** At .008 level of significance. The post-test mean for inmates with married parents was 349.00 and 301.07 for non-married parents with a t-value of 2.94 (P = .008). There was no statistically significant rise in self-concept within the married group or within the 93 non-married group. T-values was -.83 (P = .426) and -.05 (P = .963) respectively. Once again these findings may be attributed to the small variable at Ingham. But it may also mean that inmates in these two facilities who had married parents entered the institutions with a higher self-concept than those whose parents were not married. Furthermore, the differences within each group were not statistically significant. The married group had a higher self-concept than the non-married group. Analysis of Jail Officer Data A questionnaire was developed and administered to the officers in each jail to ascertain their attitudes toward their roles and respective institutions. Eight officers in Jackson and sixteen officers at Ingham com- pleted this questionnaire. The following section will list each of the eighteen questions in this survey and provide an explanation of the findings. Q. #1. "I give the residents in this jail very little responsibility." Sixty-two percent of the Ingham County officers agreed with this statement 37% disagreed. The mean response for the guards at Jackson was 2.5 which is inconclusive considering the fact that each of the four possible 94 responses had 25% of the answers in each category. There- fore, it can be concluded that a majority of officers in Ingham agreed with this statement while officers in Jackson have mixed opinions. Q. #2. "When a new inmate is booked in the jail I try to help him get acquainted with other inmates in his tank." There was a sharp difference between the responses of the two groups. At Ingham the mean was 2.4 (disagree) while at Jackson the mean was 3.4 (agree). The responses in Ingham were nearly Split evenly with 43% agreeing and 50% disagreeing with an average deviation of .55. At Jackson 50% agreed, 12% strongly agreed, 25% disagreed and 12% strongly disagreed and there was a very high average deviation of .925. It can be concluded that the officers at Jackson make a strong attempt to help inmates become acquainted, while in Ingham about half do and half do not. Q. #3. "Working in this jail is challenging and rewarding." Both groups agreed with this statement although officers in Jackson agreed more strongly than those in Ingham. At Ingham the response was 3.1 (agree) while at Jackson the mean was 3.5 (agree). The average deviation for Ingham was .6 and the average deviation for Jackson was .5. At Ingham 6% strongly disagreed, 12% disagreed. 50% agreed and 31% strongly agreed. At Jackson 50% agreed 95 and 50% strongly agreed. It can be concluded that most officers in both jails feel that their jobs are challenging and rewarding although only in Jackson were all responses positive. Q. #4. "Almost all inmate grievances are petty in nature." Both groups agreed with this answer although offi- cers in Jackson agreed more strongly than officers at Ing— ham. At Ingham the mean response was 2.6 (agree) while at Jackson the mean response was 3.1. Ingham had a fairly high average deviation of .763, while 5% strongly disagreed, 50% disagreed, 25% agreed and 18% strongly agreed. Thus, the relative frequency for this response indicates that 56% of the officers disagreed to some degree while 44% agreed to some degree. At Jackson 22% of the officers disagreed, 44% agreed and 33% strongly agreed. In Ingham the responses to this item indicate about half of the officers feel inmate grievances are petty while the other half do not. Officers at Jackson feel strongly that inmate grievances are petty in nature. Q. #5. "Most inmates can change their ways and become better citizens in society." The responses at Ingham were evenly split (7 = 2.5) while officers at Jackson (Y = 2.8) agreed with the state- ment. Six percent of the officers at Ingham strongly 96 disagreed, 46% disagreed, 33% agreed and 13% strongly agreed. In Jackson 33% disagreed, 55% agreed and 11% strongly agreed. Q. #6. "When you give a resident an inch he will take a mile." Both groups agreed with this statement although the agreement was stronger among Jackson officers. The mean response for Ingham was 2.7 and the mean response for Jackson was 3.1 while both groups had rather high standard deviations of .86 and .778 respectively. At Ingham 13% of the respondents strongly disagreed, 26% disagreed, 33% agreed and 26% strongly agreed. At Jackson 37% disagreed, 12% agreed and 50% strongly agreed. Thus guards at Jackson agreed more strongly with this statement than guards at Ingham who agreed, but less positively. Q. #7. "If an inmate has a problem he should be able to talk to a guard immediately." Both groups agreed with this statement although the officers at Jackson agreed more consistently and strongly than officers at Ingham. The mean response at Ingham was 2.8, with an average deviation of .823 and the mean response at Jackson was 3.3 with an average deviation of .589. At Ingham 6% strongly disagreed, 37% disagreed, 25% agreed and 31% strongly agreed. At Jackson 11% disagreed, 44% agreed and 44% strongly agreed. Thus, officers at Ingham generally agreed with this statement although there is a great deal 97 deal of variance in their responses. Officers at Jackson tended to agree with this statement more strongly. Q. #8. "Inmates have a right to 'treatment' in jail." Both groups agreed with this statement although the degree of agreement was stronger for the guards at Jackson. The mean response at Ingham was 3.0 and the mean at Jackson was 3.3. Both groups had similar average deviations of .429 and .433 respectively. At Ingham 21% disagreed. 57% agreed and 31% strongly agreed. At Jackson 66% agreed and 33% strongly agreed. Q. #9. "I feel most inmates in this jail cannot be trusted." The officers at Ingham had a mean response of 2.4 which meant that they generally disagreed with this state- ment. Twelve percent strongly disagreed, 50% disagreed, 25% agreed and 12% strongly agreed. The officers at Jackson agreed with this statement; mean response was 3.3. At Jackson 22% of the officers disagreed with this state- ment, 55% agreed and 22% strongly agreed. Q. #10. "Inmates in this jail are terribly stupid." Both groups answered almost identically to this statement. The mean at Ingham was 1.7 and the mean at Jackson was 1.8. In Ingham 43% strongly disagreed, 43% disagreed and 12% agreed. At Jackson 33% strongly dis- agreed, 55% disagreed and 11% agreed. 98 Q. #11. "I do not have much patience with inmates who demand alot of attention." Both groups disagreed with this statement although the officers at Jackson disagreed more strongly than the officers at Ingham. At Ingham 5% strongly disagreed, 62% disagreed and 31% agreed with this statement. In Jackson 33% strongly disagreed, 55% disagreed and 11% agreed. The mean response at Ingham was 2.3 while the mean response at Jackson was 1.8. Q. #12. "I get along with quite a few of the inmates here." Both groups had almost identical responses to this statement. The mean response at Ingham was 3.1 and 3.2 for Jackson. At Ingham 6% of the officers strongly disagreed, 68% agreed and 25% strongly agreed. In Jackson 77% of the guards agreed with this statement and 22% strongly agreed. Thus both groups perceived themselves as being well liked by the inmates. Q. #13. "I like to try and make the inmates here feel comfortable." Both groups agreed with this statement although the officers at Jackson agreed to a greater extent than the guards at Ingham. The mean response at Ingham was 2.8 while the mean at Jackson was 3.1. At Ingham 31% disagreed, 62% agreed and 6% strongly agreed. At Jackson 88% agreed and 11% strongly agreed. 99 Q. #14. "The guards in this jail should have more authority and power when dealing with inmates than current governmental laws allow." The guards at Ingham agreed with this statement while the guards at Jackson disagreed. The mean response at Ingham was 2.8 while the mean response at Jackson was 1.9. At Ingham 6% of the officers strongly disagreed with this statement, 37% disagreed, 31% agreed and 25% strongly agreed. At Jackson 22% strongly disagreed, 66% disagreed and 11% agreed. Q. #15. "When inmates from this jail are released into society, I would not be opposed to becoming friends with some of them." Both groups had very similar answers to this ques- tion with Ingham's mean at 2.5 and Jackson's mean at 2.6. The responses of officers at Ingham to this item were split on this question with 6% strongly disagreeing, 43% disagree- ing, 43% agreeing and 6% strongly agreeing. At Jackson 44% disagreed and 55% agreed with the statement. Q. #16. "The overall morals in the Shriff's Department is good." Both groups disagreed slightly with this statement; the mean response at Ingham was 2.4 and the mean response at Jackson 2.3. Thirteen percent of the officers at Ingham strongly disagreed, 33% disagreed and 53% agreed. At Jackson 11% strongly disagreed, 33% disagreed, 44% agreed and 11% strongly agreed. 100 Q. #17. "The overall morale in the jail is good." Both groups disagreed slightly with this statement although the officers at Ingham disagreed to a little more than the officers at Jackson. The mean response at Ingham was 2.1 while the mean at Jackson was 2.4. At Ingham 31% strongly disagreed, 25% disagreed and 43% agreed. At Jackson 55% disagreed and 44% agreed. Q. #18. "Inmates and guards talk to each other more than they Should in this jail." Both groups disagreed with this item. the mean at Ingham was 2.0 and the mean at Jackson was 2.1. At Ingham 31% oftflnaofficers strongly disagreed with this statement, 43% disagreed, 18% agreed and 6% strongly agreed. At Jackson 33% strongly disagreed, 33% disagreed, 22% agreed and 11% strongly agreed. A proper statistical analysis was not feasible with this data but, in examining the raw data there were some very interesting and unexpected results. The data indicates that the officers at Jackson had more of a treatment orientation than the officers at Ingham. This is surpris- ing because it was assumed the officers in the treatment facility would more likely favor a "treatment philosophy" and that officers in the custodial facility would favor a "custodial philosophy." 101 In this guard questionnaire sixteen of the items attempted to ascertain the officers' attitude towards their respective jobs, the jail and the inmates. Two of the items in the questionnaire elicited the officers percep- tions of morale in the sheriff's department and the jail. On twelve of the sixteen attitudinal items the officers at Jackson indicated that they had more of a treatment orienta- tion than the officers at Ingham. In many of the questions, though, the statistical difference between the two groups was very small. Since statistical analysis was not done this researcher cannot say definitively that the officers in Jackson possess more of a treatment orientation than officers at Ingham. But, by examining the raw data, at face value, it does appear that the custodial officers have an orientation that was not expected by this researcher. One possible explanation for this phenomenon may be the fact that officers at Jackson are more comfortable with their roles as jail guards than their counterparts at Ingham. As previously discussed in Chapter I, the officers at Ingham are under a great deal of pressure to behave in a manner that is conducive to a rehabilitation setting. But, the officers at Ingham are also expected to fulfill their custodial obligations in the jail. These two roles are contradictory and the conflict may produce anxieties 102 and eventually lead to negative attitudes toward the inmates and their own roles in the jail. This effect can be illus- trated with the response to an item on the questionnaire. Question number fourteen asked if, "The guards in this jail Should have more authority and power when dealing with inmates than current governmental laws allow." The offi- cers at Ingham agreed with this statement (X = 2.8). This indicates that officers at Ingham feel the need of more authority to effectively perform their duties. Therefore, one possible explanation for the results is that this forced dichotomy of the guard role has produced an effect in direct conflict with the goals of a treatment institu- tion. The officers at Jackson appear to be quite comfort- able with their roles as officers. They fulfill their tasks as custodians very effectively because they are only required to perform one role. For example, their response to item fourteen was disagree (I = 1.9). They do not feel they need more power to adequately perform their jobs because they already have enough authority, while the officers at Ingham feel they lack adequate power and authority. This rationale is one possible interpretation of these findings. While it does seem feasible, there may 103 be other possible explanations for these findings. For example, the officers in the two facilities may be working under two different management styles. The differences in these styles could produce these attitudes in the officers. For example, the officers at Ingham felt the morale was not good in their jail (Y = 2.1). The officers at Jackson also felt that the morale was not good in their jail but not to the same extent as the officers at Ingham (I = 2.4). This issue should be studied in more depth in the future in order to explain the differences in orientation between these two groups. CHAPTER IV CONCLUSIONS AND IMPLICATIONS Conclusions based on the data analysis and implica- tions for policy development in county jails will be form- ulated in this chapter. These conclusions will be derived from the statistical findings in this study. It must be reiterated that the findings in this study cannot be con- strued to be definitive fact for a number of reasons, one of which is that the number of inmates in the sample was very small. Although conclusions and implications will be derived from the data, they must be interpreted at face value and not as definitive evidence of validated phenomena, in the jail setting. Findings in this investigation indicate that there was considerable difference in inmate self-concept in two institutions over a three month time period. Self-concept significantly decreased in the custodial facility and increased in the treatment facility but, that increase was not statistically Significant. What is important, is that there were significant differences in self-concept over time between the treatment and custody jail. This is an important finding since the emphasis of contemporary correc- tions now seems to be moving away from the rehabilitative 104 105 model and toward the deterrence model. The rehabilitative model has been losing favor in corrections at an increas- ing rate over the past few years. Many practitioners and other professionals who had previously adhered to these techniques have "given up" on the rehabilitation model because "it doesn't work." In other words, the goals of the rehabilitation model are not being attained through correctional treatment and other rehabilitative modalities. The results of this study indicate that use and further analysis of treatment programs should be continued. Previous studies that have attempted to evaluate the effec- tiveness of treatment programs (i.e., Martinson) have pri- marily used recidivism as the dependent variable. Recid- ivism, however, may not accurately measure the effective- ness of treatment programs for a number of reasons. First, there is no universally accepted definition of recidivism. For example, does recidivism include the re-arrest rate, re-conviction rate or the rate of re-incarceration? Also there is no agreement on the time that should elapse between release and recidivism calculations (i.e. six months, one year, five years?). Finally, recidivism may not accurately ascertain the effectiveness of treatment programs because it does not measure the severity of crim— inal activity. For example, the recidivism rate for two 106 exconvicts who committed the same initial crimes (armed robbery) may be identical but one may have committed three more armed robberies while the other may have committed three car thefts after release from prison. Thus, their recidivism rates would be identical but the severity of the crimes committed would not be reflected by those rates. Consequently, since the recidivism rate may not accurately measure the effectiveness of programs, other variables should also be employed during an evaluation. The self—concept level may be one of-the most salient variables in evaluating the effectiveness of pro- grams in jails and prisons. This assumes that an increase or decrease in self-concept may have a direct effect upon behavior. Certainly, if any policy implications can be derived from this Study, it should be that we must not completely jump on the bandwagon for deterrence facilities and programs and abandon the treatment modalities in correc- tions. Treatment programs have been ineffective because they have only been in use a short time and have not devel- oped fully to achieve optimal effect. Most importantly, treatment programs have "failed" because they have not been properly and effectively evaluated. This approach must not be abandoned until all possibilities, in both program pro- gram development and evaluation, have been exhausted. 107 The results of this study add to findings in pre- vious self-concept studies conducted in treatment and custodial institutions. The findings of the last three self-concept studies cited under related literature con- flict with each other. The study just completed supports the findings of Maskin and Fletcher. The results also partially support the findings of Cohen and Verner who found that Self-concept significantly increased in the treatment setting but that self-concept did not signifi- cantly change in the custodial setting. Finally, the results were in conflict with the findings of Culbertson which concluded that self-concept did not change over time in a custodial setting. As stated previously in this report the findings of the study are not conclusive for a number of reasons. The N was small, effects of the extraneous variables on each other were not taken into account, not all possible extraneous variables were examined and the variables all possessed inherent measurement difficulties. There were, though, some interesting findings in the results of the examination of variables that were analyzed. An analysis was conducted to determine differences in Self-concept between inmates at Ingham who participated in treatment programs and those who did not. The analysis 108 indicated that treatment did not have a Significant effect on inmate self-concept. This finding means that institu- tional milieu may be a more crucial variable in determin- ing self-concept than treatment. A further analysis was conducted to determine the differences in self-concept between the inmates of Ingham who did not participate in treatment programs and the residents at Jackson. A statistically significant differ- ence in self-concept was found between these two groups. Also self-concept was significantly decreased at Jackson and increased at Ingham during three months of incarcera- tion. It is very difficult to interpret and construct policy statements from the results of the last two analyses. It can, though, be assumed from this analysis that treatment is not the only factor that produces a rise in self-concept. The entire institutional milieu affects the self-concept of inmates. This indicates the atmosphere of the treatment institution does generate a healthy ambience to help residents confront their own problems in a positive and effective manner. It also indicates that the ambience of the custodial institution tends to decrease an inmate's perception of self. It can be concluded that incarceration in a treatment oriented jail may be more beneficial for an inmate than incarceration in a custody oriented jail. 109 The analysis indicates that an evaluation of the treatment programs at Ingham is in order. If they do not significantly affect self-concept, are they necessary and what treatment objectives do they reach? This question is beyond the scope of this study although a few possible explanations can be postualted. In the pre-test, the self- concept of inmates who later received treatment was far below those who did not receive treatment. This phenomenon could mean that inmates who participated in treatment were more emotionally distraught from the beginning than inmates who did not participate in treatment programs. Therefore, depending on the severity of their problems, it might be difficult in three months to help them significantly increase their perception of self. Another explanation lies in the fact that in the first phase of treatment the inmates recognizes and admits he has a problem. During this phase it is reasonable to assume that self-concept would not increase and could even diminish. Therefore, if the self-concept test was admin- istered later, after the inmate had proceeded to a more positive point in his treatment, self-concept might be enhanced. The mere presence of treatment programs may also explain the phenomena. Residents at Ingham may respond 110 positively as a result of general staff concern with their welfare. This Hawthorne effect may produce positive results whether or not the residents participate in treat- ment. A discouraging explanation for these findings may be that treatment programs are ineffective in improving inmate self—concept. Policy maker must consider this possibility but not without a more detailed examination to accurately ascertain treatment effects. An extraneous variable that affected inmate self- concept was the marital status of the inmate's parents. Inmates who had married parents scored a significantly higher self-concept on the pre- and post-test than inmates whose parents were not married. (This was significant for the inmates at Ingham but not significant for inmates of Jackson when the analysis was conducted separately.) This finding supports the theory of Walter Reckless who postu- lated that a healthy self-concept was a product of favorable socialization. Favorable socialization, according to Reckless, is facilitated by an emotionally stable and lov- ing family life. Unhealthy self-concept is a product of unfavorable socialization caused, in part, by an emotion- ally unstable and unloving family. The difference in self-concept between inmates with married and those with non-married parents was not statis- tically significant. This indicates that the effect of the 111 institutional milieu does not change the self—concept of inmates over a three month time period while controlling for this variable. The effect of parents marital status on self-concept appears to be stronger than the effect of the institution on the inmate's self-concept. The next variable that had statistical significance on self-concept was the inmate's criminal background. This analysis was conducted separately on each facility and it was conducted with both facilities combined together. Prior criminal background was not significant in the pre- test, post-test or the difference at Ingham. Criminal _background was significant at Jackson in the pre-test and post-test but was not in the difference. When the groups were combined, prior criminal background was significant at the pre-test but not at the post-test or the difference. One interesting aspect of this is that for the combined groups, and for Jackson, the inmates who had the least identification with the deviant subculture at pre- test and post-test had the highest concept of self. The medium group was in the middle range and the group with the highest amount of criminal background had the lowest self-concept at the pre-test and the post-test periods. Furthermore, inmates with the greatest increase in self— concept (or the least decrease at Jackson) were those with 112 short criminal backgrounds. The middle criminal background group had a moderate increase and those with the longest criminal history had the least increase in self-concept. The difference, however, were not statistically significant. This analysis does not support some findings of other researchers,such as Healy mentioned in the related literature. Healy found that self-concept increased as the degree of criminal background increased. Once again, though, this investigation does support the findings of Walter Reckless. The results of Reckless and this study indicate that the greater the degree of involvement in criminal activities the lower the concept of self. This finding is important because there has been a great deal of literature (i.e. Sutherlands--Differential Association) to indicate that people who expose themselves to the deviant subculture will internalize this culture and in this pro- cess they will gain status and attain a higher concept of self. Findings in this study indicate that Sutherland's theory is invalid. It appears that inmates with the least identification with the deviant subculture possess the highest self-concept and inmates with greatest identifica- tion possess the lowest self-concept. The analysis of another extraneous variable, current charge, is similar to criminal background and supports the 113 above hypothesis. Current charge at the post—test in Ingham did have a significant effect upon inmate self- concept. Inmates who committed Part I crimes had a Signi- ficantly lower self-concept at the post-test than inmates who committed Part II crimes. Furthermore, self-concept did not change significantly over three months for inmates who committed Part I crimes but did significantly increase for inmates who committed Part II crimes. It can be assumed that many of the inmates who committed Part I crimes had a greater identification with the deviant subculture. Conse- quently, this further supports the hypothesis that inmates .with the lowest prior criminal background possess the highest self-concept and inmates with the greatest prior criminal background possess the lowest self-concept. It is important to note that inmates who commit less serious crimes can be helped through incarceration in a treatment oriented jail. The apparent implications of this finding are that inmates, with only slight identification with the deviant subculture, can change their own lives in a positive fashion if given the right opportunity in the proper set- ting. At Jackson there was no Significant difference in the difference in self-concept between the pre- and post- test scores for those inmates who committed Part I and Part II crimes. This indicates that inmates who commit 114 Part II crimes and are incarcerated in a treatment oriented jail can significantly improve their self-concept and hope- fully direct their future energies and behavior in a more positive and useful manner. The results of this study cannot be generalized to other treatment and custody oriented jails in this country. The analysis can only pertain to Ingham County Jail and Jackson County Jail. Some of the characteristics of these two facilities may be found in other county jails. There- fore, jail administrators may be able to gain some valuable insights into the effects of their institutions on resi- dent's self-concept. CHAPTER V PROBLEMS OF DATA COLLECTION AND IMPLICATIONS FOR FUTURE RESEARCH In this chapter the study will be critically examined and alternative methods of jail research suggested for the furure. The first section of this chapter will deal with problems of locating a suitable research site. The second section will include a discussion of the small number of available inmates who could be used as subjects, the problems the writer encountered in data collection and possibilities for increasing N size proposed. The inherent methodological problems are examined in the third section and alternative research designs proposed for alleviating these problems in the future. In reviewing the literature for this thesis it was observed that there were no empirical studies conducted in a jail setting. This does not mean that none have been done. Just that no studies were uncovered by this researcher in his review of the literature. Our knowledge of the effects of local correctional institutions on inmates is limited in comparison to the data we have on the operation and concomitant effects of penitentiaries. The prison has been studied, debated and analyzed for decades, yet it was not until 1970 that the first jail census was conducted in 115 116 this country. Consequently, it was not until this date that concerned practitioners and professionals had knowledge of the most rudimentary demographic data about jails. It was this dearth of knowledge that inspired this writer to explore this uncharted area in the criminal justice system. The first major problem in this project was find- ing suitable sites to conduct the research. Prison admin- istrators are accustomed to "academics" utilizing their institutions to conduct their research but sheriffs typically are not and may find it an intrusion into their domain. Sheriffs are locally elected officials who may tend to feel sensitive about outsiders who enter their facilities and may view the activities in the jail with unprofessional bias. These fears may indeed be valid. Prisons are operated by the state and a large number of administrators absorbtfluepressure if controversial conclu- sions or results are arrived at by the researcher. The sheriff, on the other hand, is politically vulnerable to his local detractors if detrimental conditions in the jail are exposed through research. It was in the face of these political realities that I attempted to locate both a suit- able custodial facility and a treatment oriented jail. Fortunately there were no real problems in finding and utilizing the treatment oriented facility. The jail 117 has a very good reputation locally and nationally the administrators were very willing to allow the researcher to show, empirically, that they operated a superior facility. On the other hand, locating a suitable custodial institu- tion was difficult. Nearly every county in southern and mid-Michigan was contacted (I even contacted one or two counties in western Ohio) concerning use of their jails for this study. All indicated that it would not be possible for a number of official reasons (i.e. liability, currently in the process of changing administrators, not enough staff, already have pending lawsuits, etc.) which indirectly alluded to the political precautions addressed above. Finally, one sheriff in southern Michigan relented and granted permission to conduct the study in the Jackson County Jail. Were it not for the open-minded sheriff and undersheriff in Jackson County, this study would have never been completed. The problem of obtaining research sites is diffi- cult to overcome. One method used when approaching the jail administrators was to present the research package in a manner to make it appear favorable to the institution. In other words, "if the project was conducted in this particular facility it would expirically confirm and greatly enhance the very fine qualities and attributes of the facility." When the project is presented thusly the 118 officials may relax, feel less defensive and may even see value in having a research study conducted in their insti- tution. The major problem afflicting this study was the sample size. Because the N was so small, the overall quality of the study was seriously effected. The results of data analysis could not be determined to be con- clusive. Obviously, this problem should be avoided in the future. The reasons for a frequently changing population in jails has been discussed previously in this thesis. It is extremely difficult to predict who will be in the instia tution for any specified amount of time. One very effec- tive way to overcome this difficulty would be to conduct the research in a setting where there is a very large population. Ingham County Jail has an average daily pop- ulation of 220 and this was not sufficient for the purposes of this study. Therefore, a facility should be located that houses at least 600 sentenced and unsentenced inmates. At Ingham 48 inmates were administered the pre-test and 19 the post-test. Using this proportion with a jail of 600, we would find that approximately 150 inmates would be pre- tested and 60 inmates would be in residence for the post- test. This sample would be adequate for proper and effec- tive data analysis. Obviously only a small percentage of 119 jails have populations of this size. The 1970 local jail census indicates that there are only thirty institutions with inmate populations over the 600 mark.‘42 But, if the resources are available, these would be appropriate places in which to conduct a longitudinal study. An alternative research design would be a pre-test, post-test and a post hoc test. This design would be effec- tive in a large jail setting as described above. The pre- test and post-test would be conducted in the institution and the post hoc test would be administered at a specified time after release. The post hoc test would measure the change in self-concept after release from the institution. Thus, the effect of the jail milieu on future self-concept and future deviant behavior could be explored while simul- taneously controlling for extraneous variables in this design. Another possible design could include a pre-test and a post-test at a future point in time. The post-test would be given at this specific time (i.e. six months) regardless of whether the inmate was or was not still in jail. If the inmate was not incarcerated during the post- test, he would be administered the TSCS and a structured M2U. S. Department of Justice, The 1970 Jail Census, (Washington: U. S. Government Printing Office, 1973). 120 interview both of which could be conducted in his community. The structured interview would ascertain his attitude toward the jail at the time of post-testing as well as during his period of confinement. If the inmate was incarcerated during the post-test he could be given the TSCS and a structured interview, both conducted in the jail. The structured interview would ascertain the inmates current attitudes toward the jail and his own confinement. If a researcher decided to test the effect of a jail on an inmate's attitude toward some subject (i.e. self-concept) a longitudinal design should be used. But one method of avoiding this process of jail inmate sample selection, (choosing inmates who will be incarcerated for a specified amount of time) is to use a cross-sectional design. Robert Culbertson did this in his cross-sectional study on self-concept in delinquent youths.”3 This is not ideal as it is difficult to accurately measure a causal process over time with a cross-sectional study. Even though it is not as suitable for this research as a longi- tudinal design, it may be the most practical way to avoid the pre- and post-test sample selection problems in jails. "3Hobert G. Culbertson. "The Effect of Institu- tionalization on the Delinquent Inmate's Self-Concept," The Journal of Criminal Law and Criminology, Vol. 66, No. 1 (March, 1975), p. 90. 121 Two other alternatives exist to help solve the sample selection problems in jails. The alternatives are very similar, a variation on the same theme. This first design (See Figure 4) provides for both an experimental and a control group. It will employ a post—test but pre— testing will not be necessary. FIGURE 4 RESEARCH ALTERNATIVE Pre-Test Treatment Post-Test Experimental Group X T2 Control Group T2 The results of Culbertson's study and this thesis reveal that pre-test scores at both institutions showed no significant differences. This indicates the groups had essentially the same level of self-concept at the pre-test. This is accomplished by randomized sampling techniques which insure that, at the time of the assignment, the groups were equal. Therefore, in this design, it may be assumed that both the experimental and the control group would have had similar self-concept scores if pre-tested, so there will be no need to conduct that test. The next step in this design is simply to test all inmates who have been incarcerated for a pre-determined 122 amount of time (i.e. three months). Finally, to determine the differences in self-concept between the control and experimental groups, a variety of statistical techniques are used. This approach would enable researchers to determine differences in self-concept between the two groups over time. One advantage of this design is that is enables the researcher to give the self-concept test to every inmate who has been incarcerated in the institution for three months. This researcher may have overlooked a number of inmates who eventually did remain in the jail for the three months. This is disturbing and inefficient, allowing eligible inmates to completely bypass the study. If only a post-test is used, including all eligible inmates, it becomes an efficient matter. Another advantage of this design is that it avoids the costly and time consuming process of collecting pre-test data. Using only a post- test will keep costs lower. There are a few disadvantages to this research design. One is that it can only determine statistical differences in self-concept, over time, between the con- trol and the experimental group. In this thesis other factors were examined which could not be explored with the above design. For example, the effects of extraneous 123 variables could not be analyzed since demographic data would not be collected at a pre-test. In this thesis, for example, it was found that inmates with longer criminal records had significantly lower self-concept, at the pre- test period, than inmates with shorter backgrounds in crime. These facts could not be determined in this design because there is no pre-test which means that there can be no post hoc comparisons. Another disadvantage is that the design assumes a pre—test commonality without any verification that that commonality actually existed. The following research design would overcome this problem. The design is very similar to the design above except that the experiment is actually conducted twice. (See Figure 5) The first time a pre- and post-test are FIGURE 5 RESEARCH DESIGN Pre-Test Treatment Post-Test Pre-Tested Experimental Tl X T2 Pre-Tested Control Tl - T2 Unpre-Tested Experimental X T2 Unpre--Tested Control - T2 124 used with both the experimental and control groups. The second time, just a post-test is used with the experimental and control group. In other words, this design assigns the subjects into four groups. In the first two groups a pre- test and a post-test are used but only a small number of subjects are administered these tests. In the third and fourth groups only a post-test is used and a large number of subjects are assigned to these two groups. This design and the research procedures for these last two groups are exactly identical to the research design described in the first research alternative. A great deal of confidence can be placed in findings for groups 3 and 4 if they are consistent with the results of groups 1 and 2. Thus, it is possible to assume that the pre-test scores for groups 3 and 4 would would be similar to groups 1 and 2. This design insures that post hoc com- parisons can be conducted with assurance that the pre-test scores were similar but without the burden of conducting extensive pre-test examinations. Before concluding this chapter some of the method- ological weaknesses in this study will be examined altern- atives proposed to improve this type of research in the future. 125 One critical flaw is that the study did not measure self-concept following the inmates' release from the insti- tution. The entire purpose of the jail is to effect changes in the inmate in order to diminish his deviant behavior in the community. It is assumed that increase or decrease in self-concept while in jail will directly effect his behavior in society. This being so, it is imperative that the former inmate's self-concept be measured again, after a specified amount of time in the community, to determine if the effects of the institution are long lasting or only temporary. It would also be of interest to determine whether self-concept varies in a direct relationship with rates of recidivism. Another critical flaw resulted from insufficient data collection. Although data on nine extraneous vari- ables was collected in this study, one additional variable should have been included. Length of sentence remaining, or the amount of time left to be served in jail is a factor which may have significant effect on inmate self-concept. It is possible that an inmate's self-concept will increase if his release is imminent. It could even decrease if impending release caused him to view his chances "on the outside" in competition with others but labeled an "ex-con." This data could have been collected easily for sentenced inmates by examining their jail booking records. An unsentenced inmate taking the pre-test and post-test could 126 have been asked when he expected to be released. Since release date (or expected release date) is vitally impor- tant to an inmate, its inclusion as an extraneous variable could be extremely beneficial. A final problem with this project is the assumption that the Jackson and Ingham county jails actually do possess all of the characteristics inherent in the concept of "custodial" and "treatment oriented" jails. From personal observation they appear to possess characteristics consis- tent with concept models but this was not empirically determined. In the interest of improved reliability, the social climates of the institutions should be examined and measured, particularly as this study found evidence that milieu may outrank treatment program participation in affecting self-concept. One method to accomplish this 44 This would utilize Rudolf Moss's Social Climate Scale. scale would be administered to administrators, staff and inmates in both facilities. The Social Climate Scale is an excellent research instrument to ascertain the aspects of a jail which are charateristic of a custody environment uuRudolf Moss, "The Assessment of Social Climates in Correctional Institutions," The Journal of Research in Crime and Delinquency, Vol. 5, (July, 1968), p. 174-188. 127 and those attributes which are consistent with a treatment oriented facility. The Social Climate Scale can identify which milieu characteristics are being emphasized in the jail, if characteristics are changing over time and the areas of discrepancy if management, staff and inmates do not agree. The use of this scale would determine empiri- cally if Jackson, in fact, fits a custody-oriented model and Ingham, in actuality matches a treatment-oriented model. At this point it becomes appropriate to list research questions not examined in this study but relevant to self-concept studies conducted in jails. The following questions should be examined in future research to more accurately assess the effects of jails on inmate self- concept: 1. What effect does the expected time of release from jail have upon inmate self-concept? 2. What effect does the philosophical orientation of the facilities' correctional officers have upon inmate self-concept? 3. If the jail produces a change in inmate self- concept, what effect will this change have on inmate deviant behavior after the offender is released to the community? 4. If the facility produces a change in inmate self-concept, what will the duration of this change be after release? studies be more ticular tion of concept 128 Do some treatment programs produce a greater increase in inmate self—concept than other Speci- fied treatment programs within a treatment oriented facility? What effect do the jail administrators and/or the sheriff's management style have on inmate self- concept? If these research questions are taken up in future the effects of the jail on inmate self-concept may accurately ascertained. The results of this par- study must not be viewed as a definitive illustra- the effects of the two institutions on inmate self- but should be viewed as a starting point for further research. With the groundwork laid and some obstructions defined, some of the difficulties encountered in this study Should be avoided. Hopefully, this project will inspire researchers to further investigate our nations jails in order to determine the effect of these institutions upon their residents. APPENDIX A II. APPENDIX A POST-TEST QUESTIONNAIRE HAVE YOU BEEN ASSIGNED TO A WORK DETAIL WHILE YOU HAVE BEEN IN THE JAIL? YES NO A. IF YES, WHAT TYPE OF WORK HAVE YOU BEEN DOING? B. HOW MANY WEEKS HAVE YOU BEEN WORKING? HAVE YOU BEEN INVOLVED IN NAY OF THE TREATMENT PROGRAMS IN THE JAIL? YES NO IF YES, PLEASE CHECK THE APPROPRIATE PROGRAM(S). DRUG PROGRAM FOR HOW MANY WEEKS? ALCOHOL PROGRAM FOR HOW MANY WEEKS? EDUCATION PROGRAM_____FOR HOW MANY WEEKS? AA PROGRAM FOR HOW MANY WEEKS? 130 BIBLIOGRAPHY Burns, Henry J. Origins and Development of Jails in America. 1974. Cohen, Bruce J., and Verner M. Arthur. "Self-Concept Modification and Total Correctional Institutions." The Journal of Correctional Education, 20 (January, 1968): 113-125. Cohen, Howard C. The Relationship of The Prison Program to Changes in Attitudes and Self-Concept of Inmates. Saratoga, California: R and E Research Associates, 1975. Culbertson, Robert G. "The Effect of Institutionalization on the Delinquent Inmates' Self-Concept." Journal of Criminal Law and Criminology, 66 (March, 1975). Dungworth, Terence. "A Methodology for the Measurement of Political Attitudes." M.A. Thesis, University of Utah, 1971. Eitzen, Stanley D. "The Effects of Behavior Modification on the Attitudes of Delinquents." Behavior Research and Therapy, 13 (October, 1975), 295-299. Fitts, William H. Manual—Tennessee Self-Concept Scale. Nashville, Tennessee: Counselor Recording and Test, 1965. Goldfarb, Ronald. Jails: The Ultimate Ghetto of the Criminal Justice System. Garden City, New York: Anchor Books, 1976. Greenspan, Barry. "Differences in Self-Concept Identifica- tion by a Schizophrenic and Non—Schizophrenic." Ph.D. Dissertation, Michigan State University, 1970. Greiger, Russell. "Self-Concept, Self-Esteem and Rational Emotive Therapy: A Brief Description." Rational Living, 10 (Spring, 1975), 13-17. 131 Hackler, James C. "Boys, Blisters and Behavior-~The Impact of a Work Program in an Urban Area." The Journal of Research in Crime and Delinqpency, 3 (July, 1966), 155—164. Hall, Peter M. "Identification with the Delinquent Sub- Culture and Level of Self—Evaluation." Sociometery, 29 (June, 1966). Hamachek, Don E. Encounters with the Self. Chicago: Holt, Rinehart and Winston, Inc., 1977. Isaac, Stephen. Handbook in Research and Evaluation. San Diego, California: Edits Publishers, 1971. Moss, Rudolf. "The Assessment of Social Climates in Correctional Institutions," The Journal of Research in Crime and Delinquency, 5 (July, 1968), 174-188. National Advisory Commission on Criminal Justice Standards and Goals. Corrections. Washington: U. S. Government Printing Office, 1973. Nie, Norman H. et. al., Statistical Package for the Social Sciences. New York: McGraw-Hill Book Company, 1970. Pursley, Robert D. Introduction to Criminal Justice. Encino, California: Glencoe Press, 1977. Recklees, Walter and Dinitz, Simon. "Pioneering with Self- Concept as a Vulnerability Factor in Delinquency." The Journal of Criminal Law, Qgiminology and Police Science, 58 (December, 1967), 515-523. Rogers, Carol. "The Organization of Personality." The American Psychologist, 2 (July, 1966), 358-367. Sandhu, S. H. A study on Prison Impact. Delhi, India: Bharat Bharat: Press, 1968. Scarpitti, R. F. "Delinquents and Non-Delinquents Percep- tions of Self, Values and Opportunity." Mental Hygine, 49 (Fall, 1965), 399-404. 132 Sholam, Shlomo and Shaskolosky, Leon. "An Analysis of Delinquents and Non-Delinquents in Israel: A Cross-Cultural Perspective." Sociology and Social Research, 53 (April, 1969). Tangri, Sandra S. and Schwartz, Michael. "Delinquency Research and the Self-Concept Variable." Tpe Journal of Criminal Law, Criminology and Police Science, 58 (June, 1967), 182-190. U. S. Department of Justice. The 1970 National Jail Census. Washington: U. S. Government Printing Office, 1973. U. S. Department of Justice. The Nations Jails. Washington: U. S. Government Printing Office, 1975. Weiss, Carol H. Evaluation Research, Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1972. Wylie, Ruth C. The Self-Concept, Lincoln, Nebraska: University of Nebraska Press, 1974. 133 "111111111111ES