any!” 92!!» W, .q. , . ulnfifiiflfifivr .1 . ’E-I' I E. I... ‘ 1.... . (.3; :7. .r I. 91-33.... . v , .253... 5.15:2!»i t : Yrs 35.1,: In ‘1‘: .l. “xw Y: x: 2.1.... b V Nvlcl... \ieu» 2...}!!! , "no.3! gig-u... 3:2S397 :7 I: . .5 :u , 9'73? 0. .598 :3. #91! n 1: to. :1 .1 : Wu?!v?€9.4.d. .. .31 . . 'HA {-1 x: 4; \tvin’ .. 1...... a f... I... .1321: .. is... .531 . “1.5;! f... .1. 5...: 2.5... i:x!«?i.:.! E... .. ‘u1!!‘:1~.§ 2‘ i=1! '9}. 4' i z ;7 r s... V i - 1.3V! . xi“; "When". 57.. ‘ y .9‘337 all. I. :iyuh: .1]: .. . ‘ .r :4; fjrrtt- h...‘ I}: . . . '9 . r5289: .\.. i . . .2151 :5; ~ “-3.1 , II nob... ‘ n, i. r... 7335.0). «“3 .C. .. . 1.... £531....cr 9 .10.: 8.5;”! b. rahé vi. a ’VY‘CO‘ {E t, ‘ l I‘ .thDittu‘Kd . AttijxthuP , 9%... z a: r at. , . fr.» 3 .A .. $5.)... I135 .4. I (I. t: y... I v...‘.\ Ill?! 02‘ l .Kv 5..'\ 1.. I! p v 3. 3.},- ”760- 22 If? 13mm A / W Sgl G—I’ llllllllllllllllllllllllllllll1|.iIllll [a 3 1293 00910 {3725 This is to certify that the dissertation entitled ‘ PLACE Community Service for Juvenile Offenders: An TOAVC Experimental Evaluation ‘ ‘ presented by Carolyn L. Feis l s has been accepted towards fulfillment ‘ i of the requirements for Ph .D . degree in My Major professor William 5. Davidson Date 1/ 26/90 MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771 ‘ LIBRARY Michigan State 1 University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE l W l MSU is An Affirmative Action/Equal Opportunity institution c:\circ\ddedu.pm3-p.t COMMUNITY SERVICE FOR JUVENILE OFFENDERS: AN EXPERIMENTAL EVALUATION by Carolyn L. Feis A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1990 P‘” ‘.‘. "." Se. ‘ .'-.odun -. p r ‘H ..‘;’,I‘3‘S' H nmq"l. ' h "if T . " ~‘,..» a l . a». r a t: 39:. “ “ A 3.510.341? « ..T.ili...i 38": 1653395. if» ii‘lliéés tc'r g. A “3:39 of?“ :"HI‘I 6 "“‘0: a .41;- :;";-i C. J"‘ lan~~r.g ‘l..’ T‘Hc’IE .‘ ABSTRACT COMMUNITY SERVICE FOR JUVENILE OFFENDERS: AN EXPERIMENTAL EVALUATION by Carolyn L. Feis This study describes an experimental evaluation of a community service restitution program for juvenile offenders. This program was based on a philosophy of accountability which suggests that, in order to reduce recidivism, youths must feel responsible for the crimes they commit. Ninety-four youths originally ordered to perform community service were randomly assigned to community service and traditional service groups. Subjects were assessed on measures of accountability, bonding to convention norms, bonding to delinquent norms, and self- esteem. In addition, a variety of measures were used to monitor the intervention process, as well as to measure Indor and subsequent criminal history. This data showed no relationship in the expected direction between treatment condition and any of the intermediate or final outcome measures. However, experimental youths showed more positive attitudes toward community service over time while controls did not. Efforts to create typologies of YOUthS to predict (uncome were largely unsuccessful. Additional efforts to cOnfirm a theoretical model of delinquency theory produced ndxed findings. This intervention most likely failed to Iflpduce the expected results for a variety of reasons . ‘ '. ‘ I.F "F" a l u oJ-sU“-‘ 0‘ no 'FS""""' .A‘ -,« nnr - .,:. 1"6 "' C ausc15uL u e ‘- tessvoc Ln}... .......gs tr. ' in- a: ' . ,- .. trrse .-. .‘ll including a low level of service intensity, a short follow- up, inappropriate theory, small sample size, and the influence of social desirability on the intermediate outcome measures. Any or all of these may account for these findings, but further research would be needed to rule out any of these explanations. Copyright by CAROLYN LITTLE FEIS 1990 For my parents, Laney and Bill, who taught me the most important lessons. 0 F0 V" t . -‘n ”9,9 3. . O . ‘ '.l r! or; 4‘:S‘:, - I. 'h.. “' .. u- I. ..a- I o ’9 . . n::-~.V' .'. \ a l U I fl.- . u. an! ‘ - ::.:. a.“ «6” a”-". no If r «P .4 a. .\ un‘. b e: ' ‘ . . ”o 9 ‘I II“! "" 3:83.... 3.“ . ‘ .: ‘ .v-on a npn‘ .1 .....ale. 5”,; a. A '0 l u- n.,. -'. . 5‘ 53-9 :.4‘:- ‘- F . {5. ‘ PF 0 - .3“..~€9 I“... '.-. . I n‘ :_"'_ """ 3.... .01.. -- O "W‘ at 0". . v- .“.“5 VA .‘tgs L I :5: - ‘3 ”\p 4' 'P‘v ' an»- 4'4; .u r‘- 0", ,; ehLV ”Has ‘ an» M‘ .1-9 “I tears rat-*‘c’ . ..... {we a: fitted. Ti. 2 F s‘.:.ggies, it'.k :9; and Earth: ,.i . I": - r 51' P J (1 O .4. 5n ACKNOWLEDGEMENTS There are many people who had a part in the production of this dissertation, without whom this could not have been possible. The first thanks must go to Judge Owens, Shelly Bauer and Gail Moore of the Ingham County Probate Court. who not only cooperated, but also supported this research. My four research assistants -— Joy Pleiness, John Krapohl, Suzi Brundage, and Kathleen Cooper -- were invaluable assets. They gave more of themselves than I could have hoped for. My committee members -- Bill Davidson, Carol Mowbray, Tim Bynum, and Ralph Levine -- were instrumental in the fine tuning of this product. Jack Hunter provided essential assistance during the last weeks. Into this dissertation went a great deal of blood, sweat, and tears. I want to thank those who bandaged my wounds, wiped by brow, and kissed away my tears. My parents’ love and support taught me that I could do anything I wanted. This proves it. Despite all of the obstacles and struggles, it’s finally over! Thank you for your patience. Greg and Martha were always encouraging me and provided their share of assistance. And Steven, who’s smile could cheer me up on the worst days. May your life be filled with joy, love, and understanding. Kelly Hazel was with me for the best and worst of times. She is a true friend, in every sense of the word. vi 3! . a \ .- Sandy Herman tolerated my erratic hours in the latter stages of writing, without so much as a sideways glance. Jeff Knoll’s support, encouragement, advice, love and inspiration in the early days were unmatchable. And now there are my new friends -— Linda Callies, Phil Herr, Jacqui D’Alessio, Paul Yakoboski, Mark Rom, Joe Kile, Randy Wold, Paul Herrnson, Joe White —- who offered an empathetic ear in the final moments. I am thankful for and forever indebted to these people, and many others. vii 0. 4‘; S O . .-555J‘ ‘6. TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . Development of Restitution . . . . . . . . . Definitions . . . . . . . . . . . . . . . . . . History of Restitution . Philosophical Justifications Who Benefits . . . . . The offender . . . . . . . The victim . . . . . The community . . . . . . The juvenile justice system Traditional Aims of Sentencing . . . Accountability . . . . . . . . . . . . . . . Conclusion . . . . . . . . Research on Restitution . . . . . . . . . . . Reviews of Restitution Research . . . . . . . . . Attitude Studies . . . . . . . . Quasi-experimental Outcome Studies Crime Rates . . . . . . . . . Completion Rates . . . . . . . . . . . . One Group Post-test Only . . . . . . . . Variables Related to Success . . . . . . Organizational characteristics . . . . . . Personality variables . . . . . . . . . . Two-group Quasi- -Experimental Studies of Outcome Conclusions . . . . . . . . . . . . . . Experimental Outcome Studies Conclusions . . . . . . . . . . . eories of Delinquency . . . . . . . . . . . . . . 7m3train Theory . . . . . . . . . . . . . . . . . . Limitations . . . . . . . . . . . . . { Social Learning Theory . . . . . . . . Limitations . . . . . . . . . . . . . . . . . . \ Control Theory . . . . . . . . . . . . . . . . . \\ Limitations . . . . . . . . . . . . . The Multi-dimensional Theoretical Model . . . Implications for Research . . . . . . . . . . . . Goals of the Current Study . . . . . . . . . . . METHODS . . . . . . . . . . . . . . . . . Sample . . . . . . . . . . . . . . . . . . . Subject Mortality . . . . . . . . . . Design . . . . . . . . . . . . Procedures . . . . . . . . Processing . . . . . . . . . . . . . Intervention Alternatives Traditional Services . . . . . . . . . . . . . Community Service . . . . . . . . . . . . . . Post-test Assessment . . . . . . . . . . . viii V“ - A .— ‘-.a»c 9,... _.a.. ”'9'; \ ' ”P"a" o - r-n I Ypoe .n- A A u'vnl o.~.~u '0‘ kov-uny F‘; Ablb-‘ .1. — 1 _ . r. o C . . c C t. 3 i . . r r. v . 2 9,: ~ 1 a t . . . a: a. a 0. .. a .— ... . . . a . . . A v _ _ C. «u. o . V. F. . 0.. a. .2 .44 t. S .6 V. . . .... .1 S a V. t . ... a no. a . . n... r. V. .2 VI. ‘4‘ PA A...» I. Bu F; P“ n: .u VII: FM ha. ‘- F. 144 AA. 6\ “EL F. u. 7.— .syt F. FA wro. FIN P4 IN“ V: ¢.. 3. .9. o . U. 3 a. 3 V. . a. C ‘t . . .u 7. n... 3 o r... 3 . no a 0 Au : a O .J . . a .J .. a 05‘ we 1 ’& P. F- W! VIM. . Q ‘1‘ PU “In. VI... HM m V. ‘W 0" \. a "N Vt V.‘ Hutu .1 and P.— I'H Rm h... V4 ‘ c» z. e Axxv ..K a. E a: n. F s F c 9 ~ .k h c Q» h r ’& F a V a. PA VI t “ha .J‘ VII... 5- VIVA ~v‘ O . HK IL. 8 u \ he 3‘ 3.. . . a .r . I. .a v . . a $4 1 oh. mu 9 . s L A. be VA n F. E .J 9. v... a a 0. V... a 1 . fl. “w. .1 .. 4 s x In. V. 7- t. PU c. nuns. C» 9% .. t . 5-; a ‘3 IR.- P-u r~ qy t . “I; V‘. QC at e» um D... FD; CV .«A la. VA Pu. n\~ 9‘ . a . p u u u . . .w . . 9.. a‘. on Y. 9.; hi. Alia Cw .u E... Interview Training . . . . . . . . . . . . . . . . 76 Measures . . . . . . . . . . . . . . . . . . . . . . 76 Implementation of Community Service . . . . . . . . 77 Intervention Process . . . . . . . . . . . . . . . 77 The Youth Interview . . . . . . . . . . . . . . . . 79 Scale Development . . . . . . . . . . . . 79 Measures of Intervention Integrity . . . . . . . 81 Self— esteem . . . . . . . . . . . . . . . . . . . 87 Delinquent Associations . . . . . . . . . . . . . 90 Conventional Bonding . . . . . . . . . . . . . . 93 Social Desirability . . . . . . . . . . . . . . . 97 Interscale correlations . . . . . . . . . . . . 97 Recidivism . . . . . . . . . . . . . . . . . . . 100 Data Coding . . . . . . . . . . . . . . . . . . . . 102 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . 104 Intervention Description . . . . . . . . . . . . . . 104 Intervention Process . . . . . . . . . . . . . . . 104 Community Service Hours Ordered . . . . . . . . . 104 Additional Court Ordered Requirements . . . . . . 106 Missed Interview Appointments . . . . . . . . . . 108 Lag from Time of Offense . . . . . . . . . . . . 113 Pre to Posttest Interval . . . . . . . . . . . . 113 Group Equivalence . . . . . . . . . . . . . . . . 113 Summary of Intervention Process . . . . . . . . . 115 Implementation of Community Service . . . . . . . . 115 Summary of Implementation of Community Service . 124 Summary of Intervention Description . . . . . . . . 124 Outcome . . . . . . . . . . . . . . . . . . . . . . . 124 Intervention Integrity . . . . . . . . . . . . . . 126 Purpose of Different Sanctions . . . . . . . . . 126 Attitudes Toward Community Service . . . . . . . 130 Attitudes Toward Victim . . . . . . . . . . . . . 130 Accountability . . . . . . . . . . . . . . . . . 132 Response to Community Service . . . . . . . . . . 132 Exploratory Analyses . . . . . . . . . . . . . . 135 Summary of Intervention Integrity . . . . . . . . 135 Intermediate Outcomes . . . . . . . . . . . . . . . 136 Delinquent and Conventional Bonding . . . . . . . 136 Normative Values . . . . . . . . . . . . . . . . 137 Self-Esteem . . . . . . . . . . . . . . . . . . . 137 Exploratory Analyses . . . . . . . . . . . . . . 141 Summary of Intermediate Outcomes . . . . . . . . 145 Final Outcome . . . . . . . . . . . . . . . . . . . 145 Recidivism . . . . . . . . . . . . . . . . . . . 145 Exploratory Analyses . . . . . . . . . . . . . . 148 Summary of Final Outcome . . . . . . . . u . . . 148 Additional Tests of the Intervention . . . . . . . 150 Accountability . . . . . . . . . . . . . . . . . 150 Typologies . . . . . . . . . . . . . . . . . . . 152 Path Analysis . . . . . . . . . . . . . . . . . . 158 Summary of Additional Tests of the Intervention . 160 Summary of Outcome . . . . . . . . . . . . . . . . 163 ix .- - --— u .I ." x .. . I. . I -"‘. .lo' ....-‘. .u‘_ . h I. .-_|. .‘r.o‘ .- - u-.. ._ 0"..“ I..- _ o t.-- .- .-—c‘. . :. ._._ V_ \ ‘ .- . ._\ 5-- {u .I—o . 1 .. ‘ -- - .bau‘ - ‘. - n N \ o v w u‘ ' . -\~ .- u‘,‘ u... - \ 5 . .‘ -- --..\ .. ‘ . .‘n- --..\ ~. - a...— ..__|‘ s ' v . .u_‘\‘ A a I u. n...“ 5\ c ' v. -._.‘u ‘ . '§ .- I‘. --..\" a I‘ ‘ “-.|‘- 's,‘ o ‘ s u .- .Ivgl“ ‘ I _ _ ~ “ s ‘ ’ . V. v o 0 ML”? an. 9. o“ v 5-. I... .1 v.‘ A o '1' O-l v . n-. V I ;-<~1 O a a — O o a . Y V O #‘f D D O l D n O ) "1 I" It.) O’i <)'/;¢r‘m 'U t-«Q (I) :1" l) r—U 'D H) n q J In -,-. . \ l DISCUSSION . . . . Summary and Interpretation of Findings APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX A- '4 N€s Po . n 4‘. PA a. 6 . Ru .5 9. . . o... H . . O ‘58.. .pufi- ‘C .01.. V 6 p I. J 905'? 5“ .ha 1. U C. . F. .v). 0 1-. .. 3 r” r” .1 w. c .w L ._ Dy 9.. A. ~. 0 V a tn '3 r“ fin P. v. FF. 0. o» o. v.. 9.. V n v a O "c": ‘U~-...¢. Y '.:PFA ...-....4.‘.; a ._ 5.. . no 9‘ .1 VA .1 .‘ ..m o. t. A. 9.. r... “a ... u. h. a t L. .hu L1 in» ... .2 .1 11 9» FA Ft I. ‘1 .h.. 05 ‘5 Q. t» 0.» a: .3 a: .3 ~o .‘c .1 .a oz... F. u R J n. u n .1“ .un .... . u r . - . p,- . u o A. .... .wa . .u a. a c a. ,. CL 11 u u L- «6.1 6.. 6; D& o .. $. A. by C .8 F1 v h\ «.4 .14 c5 \. . LC .3 .h... w. uh... .bU ru e h». .V. .01 V. 0.. ‘A {c a: Him ’1 I. F: P» F. ‘A v. Amy . a .0 . D. 6.. Pt. 6 . 9t «nu In. P- . a: IV. to I .- nhu Cu “u h 1 . a V . . a . c ..a . u u FL .. . m u a... .r. 7.. ..fi 1.. .5“ . . .l. I . I. . l I In. . F I 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. LIST OF TABLES Demographic Characteristics of the Sample . . . . . 59 Distribution of Previous Offenses . . . . . . . . 60 Distribution of Offenses for Current Petition . . . 62 Internal Consistency of Attitudes Toward Community Service Scale . . . . . . . . . . . . 83 Internal Consistency of Community Service Outcome Scales . . . . . . . . . . . . . . . 85 Community Service Outcome Scale Intercorrelations . 86 Internal Consistency of Attitudes Toward Victim Scale 88 Internal Consistency of Accountability Scale . . . 89 Internal Consistency of the Self-Esteem Scale . . . 91 Internal Consistency of Delinquent Bonding Scales . 92 Internal Consistency of the Normative Values Scale 94 Internal Consistency of Conventional Bonding Scales 95 Specific Normlessness (Revised) Scale Intercorrelations . . . . . . . . . . . . . . . 98 Internal Consistency of the Honesty Scale . . . . . 99 Scale Intercorrelations . . . . . . . . . . . . . 101 Distribution of Community Service Hours Ordered . . 105 Total Number of Requirements Ordered‘, Not Including the Community Service Order . . . . . . . . . . . 107 Distribution of Court Order Requirements . . . . . 109 Distribution of Missed Pretest Appointments . . . . 111 Distribution of Missed Posttest Appointments . . . 112 Distribution of Days Participating in Study . . . 114 Distribution of Placements Used by Experimentals who Completed Community Service . . . . . . . . . . . 117 Distribution of Type of Community Service Tasks Performed . . . . . . . . . . . . . . . . . . . . 118 Distribution of Number of Days Between Intake Interview and Placement Interview . . . . . . . . 119 Distribution of the Number of Days at Placement . . 121 Distribution of Number of Days Between First and Last Day of Placement . . . . . . . . . . . . . . . . 122 Purpose of Criminal Sanctions - at Pretest . . . . 128 Purpose of Criminal Sanctions — at Posttest . . . . 129 ANOVA of Attitude Toward Community Service Scale . 131 ANOVA of the Accountability Scale . . . . . . . . . 133 Feelings About Community Service Experience . . . . 134 ANOVA of the Delinquent Associations Scale . . . . 138 ANOVA of the General Bonding Scale . . . . . . . . 139 ANOVA of Normative Values Scale . . . . . . . . . . 140 ANOVA of the Self-Esteem Scale . . . . . . . . . . 142 Correlations Between Process and Intermediate Outcome Variables . . . . . . . . . . . . . . . . . . . . 143 ANOVA of Recidivism . . . . . . . . . . . . . . . . 147 Correlations Between Process Variables and Subsequent Petitions . . . . . . . . . . . . . . . . . . . . 149 xi : uI I . , 4O *1 0 .-v- .[ -v-v' a... .- :4 ' I F"“. .95-”:- .. ' ".. Op ..hs~~b 0.1, I .0 ‘I It‘ 'P 9.- ‘1 ‘1 .4 . r0 . .0 ,0 1 t I A Pv (J a. fa 9U nu r4 f“. r. as. 11. an .44 .na 1.. .1... A- 1.. 41 .2“ on .u - ..- ..- .- r. .- n I v u \ r : u I o \ ‘I. . V r I f u r J . 1 f. v F a r a a a P ¢ r 2 Cu ..r 4 ‘1. .5 VI‘. w“. ‘1 \. .\. .\ ‘1 N , I Ida 4.4 I... I - Gnu 4... law In. law In. ”on ‘ . - - - - a . - . nfu. .1 .. ‘7 5 " .l .. l . “,1" I " I I I 5 7 c o - Q s c u o a Q . 5 ‘1 A. I N i ! § ‘ i Q Q 39. Description of Clusters . . . . . . . . . . . 40. Cluster Type and Subsequent Petitions Crosstabulation . . . . . . . . . . . . ££f-M -.._. ,_ combined with other sanctions, such as probation and parole (Chesney, Hudson, & McLagen, 1978; Galaway, 1977a). There has been reluctance to use any type of restitution as a sole sanction (Hudson, Galaway & Chesney, 1977; Galaway, 1977b; 5 "Restitution Sentences", 1977). Casson (1983) demonstrated the variety of forms that restitution can take in proposing a number of different such models. The amount of restitution to be paid is sometimes determined by the amount of loss suffered by the victim. In community service programs, this dollar amount may be I” translated to community service hours at the rate of minimum wage. Matrices have also been developed to compute community services hours. The sentence is then based on the severity of the offense with various add-ons and subtractions. These matrices serve the purpose of developing a disposition which offenders can see as proportional to the offense, and therefore perceive as fair. This method has been used in a number of programs across the country (Rubin, 1985—1986). History of Restitution The practice of restitution began thousands of years ago with the Code of Hammurabi ("an eye for an eye, a tooth for a tooth"), the Old Testament, and the Twelve Tables of Ancient Rome (Casson, 1983; Jacob, 1977; Schneider, 1985). In its earliest form, restitution was practiced by small— scale societies in order to prevent blood revenge. Because the process of making restitution was not institutionalized, the resulting payments often exceeded the actual loss (Armstrong, 1983). l 4.. i4 Uhxle the d?“ cultures has been fies-line of feudal:- seafrestitutiét :ec’evelcpment cf r but went largely u refsraer and mag;s Lifie of this can 351.? be a mean-s cf rehabilztating of: Recent interw lined to four {‘3‘ Efi’sand 1970’s Swenil e ‘ ' Just Ce udtlon that SET‘PJ "J“: the re L a “twists of pv- ‘5 inte PESt 1n M ., an; .11: and Cm din: . . yvat‘ ‘ ‘aotion Hgfi~ ~‘_*m 19ad Q t) 51'; Ja Ccyb .. . 19'7- :‘bh'r ' I 43’ ~e S r014 .. Eh the 6 While the development of restitution in Western cultures has been traced back to the 18th century, the decline of feudalism was accompanied with a decline in the use of restitution (Armstrong, 1983). Calls for the redevelopment of restitution began in the mid 19th century, but went largely unanswered. Margery Fry, a British penal reformer and magistrate, reintroduced the concept in the middle of this century, suggesting that restitution may not only be a means of compensating victims, but also a means of rehabilitating offenders (Brown, 1983; Jacob, 1977). Recent interest in restitution has been primarily linked to four factors (Armstrong, 1983). First, the late 1960’s and 1970’s were associated with major reforms in the juvenile justice system. Diversion and deinstitutionalization were just some examples of the recognition that institutional custodial care had failed. Second, there was a search for innovative programming. Criticisms of previous sanctions and treatments facilitated the interest in new programming, particularly in the context of family and community. There was increasing dissatisfaction with existing sentencing alternatives (Hudson & Galaway, 1977). Third, renewed interest in the victim lead to the call for restitution (Hudson & Galaway, 1977; Jacob, 1977). A variety of organizations and advocates brought the rights of the victim to the attention of the community_(Armstrong, 1983; Evans & Koederitz, 1983; W '0 n “rm-25's, 1 ‘ 'he federal g " restitut 1 J; v 1': ) ‘. ’45 . 'hnS. :.Sr‘_SAt 9AA \J‘ three Years 1, ’E'Stitutien I; tiOn e 3303 ' S the L ‘1 tEUthh I eshey 351.- r~u . thud]! llst ¢ 5 Mn «35 3 SurVn‘. f .‘5-1 ‘al'ltut' 10 Prg 7 Matthews, 1981). Finally, restitution was inherently appealing as a means of restoring equity. This renewed interest was associated with efforts from the federal government to support research and development of restitution as an alternative to traditional dispositions. In the late 1970’s, the Law Enforcement Assistance Administration (LEAA) spent $3.2 million in 11 states on 14 programs to develop and test restitutionler fladgltflgffgnders (Criminal Justice Research Center, 1982). A grant program was later sponsored by the Department of Justice’s Office of Juvenile Justice and Delinquency Prevention (OJJDP) in which $30 million was distributed over three years in 41 separate awards to study juvenile restitution (Armstrong, 1983; Evans & Koederitz, 1983; "Expansion", 1981). In 1978, 16 states were considering or had introduced legislation establishing a mechanism by which offenders could make restitution. In addition, there were 54 programs across the United States which had restitution as a primary focus (Chesney et al., 1978). A survey of courts on the National Council of Juvenile and Family Court Judges’ mailing list found that 86% of the courts surveyed had used restitution (Schneider, Schneider, Reiter, & Cleary, 1977). A 1983 survey found that 52% of the courts had a formal restitution program and that 97% had ordered restitution on occasion (Schneider, 1985). I'cfl Eel: as the oft} been tailcred t: érzstron . I??? .‘ESCLtutiCn {SC} .. a '5‘? "av-5";t an l [n “to . it“r‘i° I and ‘51 “ETA . :3}. T; O“ 0¢¢ ‘ I 'n’j(4 “UTETi? “1&3 Si’a- ‘ 8v \ ~‘ 1y' 197'.. 35‘ la, «‘1 . . -0n , & Srg v~-c 8 Philosophical Justifications Restitution is not only a common practice, but also has been justified on the basis of providing benefits to the victim, the community and the juvenile justice system, as well as the offender. Because restitution programs have been tailored to meet a number of different objectives (Armstrong, 1983), confusion has existed over the purpose of restitution (Schneider et al., 1977). Who Benefits Restitution may benefit offenders, victims, the community and the juvenile justice system through reduced recidivism, reduced intrusiveness, reduced sanction, victim restitution, equity restoration, victim satisfaction, fear/hostility reduction, alleviation of agency problems, and cost reduction (Beck-Zierdt et al., 1982). It has been argued that restitution fulfills the need for effective noncustodial sanctions which avoid the destructiveness of incarceration and is also less costly than imprisonment, has the possibility of helping the offender, and may bring compensation to the victim (Newton, 1979). The offender. Restitution was designed to make juveniles atone for their acts in a constructive way. Youths must be active in the carrying out of the sanction (Galaway, 1977a, 1977b; Gilbeau, Hofford, Maloney, Remington, & Steenson, 1980; Staples, 1986) and exercise a Q I élréxsai I .‘6 C . u .'...n .Dl"" .-". IQ‘S‘ ~ v- . r957 .4 V v V .-'r I." ._“.‘..~- . w -" .:.. -. ..ah...'. ¢k ..,..'o ,,..-,.. _,... -‘u an. a‘. CA \IA 9 sense of responsibility (Harding, 1982). Youths are introduced to the idea of regular work, whether or not they receive compensation (Brown, 1977). As the value of community service is recognized, the offender becomes a community resource and asset, rather than a community liability (Read, 1977). This allows the offender an opportunity to regain community standing ("LEAA’s JD Office", 1978) and become a productive member of society (Geis, 1977; Siegel, 1979). Restitution is related to the amount of damage done and may therefore be perceived as more just than other sanctions (Galaway, 1977a, 1977b). As such, restitution provides a clear and simple message of consequences for behavior (Gilbeau et a1, 1980; Maloney, Gilbeau, Hofford, Remington, & Steenson, 1982). It demonstrates that someone is concerned enough about their behavior to attach concrete consequences to it (Steenson, 1983). In addition, restitution provides an opportunity for offenders to pay for the offense; exposes the offender to the needs of others; and combines punishment with training or learning experiences such as work experience, occupational skills, and training. Participants may recognize that they possess skills they were previously unaware of and further, that these skills are valuable. Restitution was developed to provide the offender an opportunity to repay the victim and become integrated into ~ --’~I_ W‘--.- 4‘" ‘_~ A 4...! the conunity as the requirements accsaplishnent I Flaicney et al . , :hafaco . “3129:! a P- ullinals (p638 10 the community as a result (Harding, 1982). Completion of the requirements of restitution allows for a sense of accomplishment (Galaway, 1977b; Gilbeau et al., 1980; Maloney et al., 1982). Restitution has also been defended as providing a socially acceptable way of expressing atonement and guilt (Galaway, 1977b; Harding, 1982) and reducing the stress associated with inequity (Utne & Hatfield, 1978). Restitution helps offenders avoid the stigmatization or demeaning treatment of other programs (Harris et al., 1979). Restitution allows offenders to be characterized as debtors to society, rather than as criminals (Pease, 1981). Offenders are required to take positive steps toward compensating their victims (Balivet et al., 1975; Staples, 1986). However, some argue that restitution actually increases youth involvement in the justice system (Matthews, 1981) because of a widening of the net and the fact that restitution orders may take some time to complete. The victim. Restitution not only recognizes claims of the victim (Maloney et al., 1982; Staples, 1986; Viano, 1978), but also provides compensation to the victims (Siegel, 1979) who may otherwise feel estranged from the criminal justice process (Geis, 1977; Harding, 1982). Victims may also gain some satisfaction in having their say (McDonald, 1978). Restitution may serve to increase victim interest in the criminal justice system as well as to merease crime rep? certain (“Expansi-C.’ Some have 8??- really a benefit t’l are unsolved, makzr :2:;~er.sating vzctz: It has also been a nierstand the pur flaue. 19’8). ( .P .he COEer.§ 6 .- ““——~_:;;_; acceptable sanctxc .: an understanda‘: leasurable, al and lg H 1380; “alone. all... 'k8 the CC Y :tIIUI] .\ 6-393 ’ reStore e Ynez.1 . ”ghee svste ‘E “1' 1979) . r .3 .. , “unit-V 889 an . S (Brokn :55 _,' at 3 been a 8 a 5:Cn«- “5 -Llsn aVa' 11— st 11 increase crime reporting because sanctions become more certain ("Expansion", 1981). Some have argued, however, that restitution is not really a benefit to victims because the majority of crimes are unsolved, making restitution an ineffective means of compensating victims (Galaway, 1977a, 1977b; Stookey, 1977). It has also been argued that victims are often unable to understand the purpose and intent of restitution programs (Raue, 1978). The community. Restitution may make probation a more acceptable sanction to the public (Brown, 1977) because it is an understandable, objective, tangible, observable, measurable, and logical consequence to crime (Gilbeau et al., 1980; Maloney et al., 1982). Further, restitution allows the community an opportunity to restate certain values, restore equity, and develop a renewed confidence in the justice system (Harding, 1982). Restitution may also satisfy the desire for retribution (Harris et al., 1979). In addition to improving community relations with those _._. M 41. ., ~44“; w": v _ ,nmwc—axn who have been skeptical of the criminal justice system (Siegel, 1979), restitution provides a direct benefit to community agencies who receive valuable and needed services from the youths who participate in community service programs (Brown, 1977; Gilbeau et al., 1980). Restitution has also been argued to be one of the mgsldcgai:§££§9tive sanctions available (Brown, 1977; Gilbeau et al., 1980). in“! .JLJ . ‘ 1 The ’u‘~‘€.'.‘..€ . :rizinal justice 5 ZST'a. 1977b; 011‘. Eestitution not ",.' available to crux-1 1314'. also can servc 1:3} by prCVidW I “a.s Restitutirm a. a i u‘ .‘Y‘ Q ale kaat npn-\ Hcfmg Citizens e :v.. u v.1:lr.g the CC8 , 12 The juvenilegjustice:§ystem. Restitution is an easily administered sanction which is less demanding on the criminal justice system than are its alternatives (Galaway, 1977a, 1977b; Gilbeau et al., 1980; Maloney et al., 1982). Restitution not only increases the sanctioning options available to court workers (Galaway, 1979; Siegel, 1979), but also can serve to eliminate system overload (Siegel, 1979) by providing tangible closure for each case (Gilbeau et al., 1980) Restitution has been referred to as an efficient way to ensure that non-violent offenders become law-abiding, tax- paying citizens after sentencing ("Expansion", 1981) while avoiding the costs and other disadvantages of incarceration (Brown, 1977; Gilbeau et al., 1980). Traditional Aims of Sentencing Some restitution programs may simultaneously serve multiple goals (Harland, 1978; Hudson & Galaway, 1978; Staples, 1986; Van Voorhis, 1983) and as such, appeal to people with different goals ("Restitution Evaluation", 1983). Restitution may be viewed as punishment because offenders are made to assume responsibility for their actions (Armstrong, 1983; Schmitt, 1985). Restitution may also serve as a deterrent (Schmitt; 1985; Thorvaldson, 1980b; Tittle, 1978), particularly with the recognition that fulfillment of restitution requires a loss of liberty and property (Armstrong, 1983). LJ In addition, results, such as 1'. groviding socially festering atonemer. restitution privy-.4. can he used to a"; prz‘:lems, and t te r I“. v . '1'. a y b I ..-.. 'aicson, 198. se‘ .f-esteem (Hard EcSregor, 1978'. (I) L.‘ a). It has 31 restit ution is 1' I: . . .. restitut i on ; 1377‘. C 13 a oh 1' Further, . thereahzati -.e care Val. 1393‘s). t has :3. Sin. 13 In addition, restitution may have some rehabilitative results, such as instilling a sense of accomplishment, providing socially acceptable ways of expressing guilt, and fostering atonement (Armstrong, 1983). Furthermore, restitution programs can teach attitudes and skills which can be used to cope with social, emotional, and economic problems, and thereby reduce law-breaking behaviors (Thorvaldson, 1980b). Restitution may help offenders regain self—esteem (Harding, 1982; "LEAA’s JD Office", 1978; McGregor, 1978; Smith, 1977) and build character ("Kansas", 1975). It has also been argued that community service restitution is the area where the rehabilitative potentials of restitution programming can best be realized (Read, 1977). Restitution allows a community to demand that an offender do something he or she would not otherwise do, thereby satisfying the need for retribution (McAnany, 1978; Schmitt, 1985). Restitution is a form of incapacitation which is a cheaper alternative than incarceration (Klein 19 ). Further, such a sanction teaches moral values through the realization of the damage done while also maintaining the core values of the community (Eglash, 1977; Thorvaldson, 1980b). Accountgbility It has been repeatedly demonstrated that restitution is not simply punishment nor rehabilitation. Rather, 'estitutiC“ "35 a’-‘itudes such I sense 0f ”CC”? Tb redempticn- . v' [.31 be held 2‘6“- and effect r913 its ccnsequence A growing Varied goals, t *5 Juveni les Tiltitted (Schn iOE .act that lost a3~~Urt L; . akJ‘lltv I We. 14 restitution was designed to influence broad social and moral attitudes such as a sense of responsibility to others, a sense of reconciliation with the community, and a sense of redemption. The basic idea of restitution is that offenders must be held responsible for their crimes (Harding, 1982) so that the offender sees the connection between the harm done and the service to be performed (Thorvaldson, 1978, 1980a, 1980b). Juveniles may therefore become aware of the cause and effect relationship between their criminal activity and its consequences. A growing consensus in the 1980’s is that, despite varied goals, the underlying rationale of restitution is to hold juveniles accountable to the victim for crimes committed (Schneider, 1986). Seattle, Washington even included the word "accountability" in the title of their restitution program ("'Accountability’", 1977). The philosophy of accountability has emerged out of the fact that most restitution programs seemed to focus on accountability, rather than treatment or punishment (Schneider, 1986). Accountability was also recently found to be the most important goal of restitution as rated by 170 directors of restitution programs (Schneider, 1985). In 1984, OJJDP awarded $1.4 million for the development of the Restitution Education, Specialized Training & Technical Assistance (RESTTA) Program ("OJJDP Program", 1984) which was based on the premise that restitution teaches people to "I '1') be accountable 'CIJJDP Prcgram", The acccunta :espcnsibility an persm is held 3* “v-rq b L ' V r....'t.1'l0.".81 :C ‘ .‘ESpChSlbill'.y : O . .I.‘ .3 ‘ ...er.:er ‘s rp.‘;L“ :ther benefits tr 3:31 which shwli :tter benefits. The message .,‘_ yr 5&5: .‘ . s - . "Stltutlon is n. rJ-‘zlshmem ); nor ‘72:... "‘aie“". P... ‘ ‘ 3:0: 1 \vult, Offer-Id“ t. .9? the 1r acme”? .*' lbe SUE: :51!“ l A Q‘ \ V LmDnt Str, *‘3 15 be accountable for their behavior ("OJJDP Initiates", 1985; "OJJDP Program", 1984). The accountability perspective emphasizes individual responsibility and accountability for one’s actions. A person is held accountable to the victim in a way that is proportional to the amount of harm done and the level of responsibility for the crime (Schneider, 1985). The offender is required to repair the damage he or she has done (Armstrong, 1983). While such a perspective may result in other benefits to offenders, accountability is viewed as a goal which should be pursued whether or not there are any other benefits. The message of accountability is that the offender is responsible for what he or she did. Restitution is a sanction in which the offender takes an active role. Restitution is not something done 39 an offender (as in punishment); nor is it something done fgg_an offender (as in treatment). Restitution is something an offender does. As a result, offenders are more likely to accept responsibility for their actions (Barnett & Hagel, 1977; Schafer, 1960, 1970, 1975). The philosophy of the accountability perspective suggests that restitution should be viewed as a reinvestment strategy, rather than a rehabilitation strategy. There are four primary assumptions of the accountability philosophy (Schneider, 1985). First, the :ffenier 0‘95 3 shctuld be prCFQT 4. on the offens F'Ft‘v .sed ever t 'r. aéi‘fiuet ' ‘ L unab'lllt}. I: .een arguej e551. 5:: ”19 respc ts.) rel . iathqlv he .c «_re cla‘p be p r JL. 511m ‘ Ch 36'. 'a ”’2 -._nt -rat 9'1 (5». 35 e; 4“ "« ‘3 “EL. 5‘ de “‘ZEA . -3, a -al‘rl‘g 16 offender owes a debt to the victim. Second, the sanction should be proportionate to the offense. Third, the emphasis is on the offense, not the offender. Finally, the goal is an end itself, not just a means to another goal. Many other treatment and rehabilitative approaches to juvenile justice have found juveniles and their parents confused over the nature of the sanction imposed. All too often, the message of the court and the justice system have been lost on the offender (Steenson, 1983). A program of accountability makes these intents very clear. While it has been argued that many offenders are not equipped to readily assume responsibility for what they have done (Steenson, 1983), a well designed restitution program can ensure that they will be. Restitution, when guided by the ph11080phy of accountability, has been called the most effective sanction available to the juvenile justice system today (Gilbeau et al., 1980). Armstrong (1983) argued that it is reasonable to structure a restitution program around the goal of offender accountability. However, because accountability is a relatively new philosophy to be clearly operationalized, there have been few empirical studies of its effectiveness. Justice Springer emphasized that efforts must be concentrated on developing and implementing programs based on this new model of accountability (Armstrong, Hofford, Maloney, Remington & Steenson, 1983). Hofford (1983) also argued that rcgra nct consistently e and that successf: the restitution ~:. “ilteau et a1. , '7 .1. su::essf..l res . . 311293;) e t 8“ .4 ‘ I . For TGSt i t' :ature cf the r“ the victim's new the victim for c 7:. .' lee Piment- ShC\ A o I teas“table . an ‘3 tea“ RIB. P‘ a 4‘9. ”“3 SOme re ‘ n “at there “La“ ("Wing the .U‘fe,‘ ' ; H1 .qutanecu81‘ i: ‘ aka-19318 to is 17 argued that programs which have failed are those which have not consistently adhered to the philosophy of accountability and that successful programs are ones in which the terms of the restitution order are clear, measurable, and achievable (Gilbeau et al., 1980; Remington, 1979). A well articulated rationale has often been cited as one of the keys to a successful restitution program (Beck—Zierdt et al., 1982; Gilbeau et al., 1980; Schneider, 1985). For restitution to work, offenders must understand the nature of the relationship to victim, have some awareness of the victim’s needs, and appreciate the notion of paying back the victim for damages caused by the offender (Heide, 1983). The payment should be an effort, a sacrifice of time or convenience; the assignment should be clearly defined, measurable, and achievable without being too easy; the effort should be meaningful; and the assignment should produce some rewards (Keve, 1978). Conclusion It is clear that restitution is not a new practice and that there may be many beneficiaries of such programs, including the offender, the victim, the community, and the juvenile justice system. Further, because restitution can simultaneously address many different goals of sanctioning, it appeals to a diverse audience. Recent efforts to describe the important characteristics of successful programs have suggested that for restitution to be 18 effective, it must be built around the philosophy of accountability. Next it is important to examine studies on the effectiveness of restitution. Research on Restitution Research on restitution has lagged far behind public enthusiasm, political clamor and theoretical claims for its effects (Chesney et al., 1978; Miller, 1981). Not only is descriptive material about community service programs rare,but evaluation information is even more sparse (Harris et al., 1979; Hudson & Galaway, 1978; Miller, 1981). The research in the field has not been extensive and many completed evaluations are not available in published form. While most restitution research has emerged out of a theoretical vacuum (Harland et al., 1979; Van Voorhis, 1983), this is not a problem unique to restitution (Altschuler & Armstrong, 1989). However, there appears to be consensus among judges that restitution reduces recidivism and the victims who are compensated are more satisfied with the way the offenders are handled by the system (Evans & Koederitz, 1983). The economic and political benefits of restitution have also been proposed as adequate justification for restitution, even without clear results about its impact on recidivism (Keve, 1978). There is strong theoretical and some empirical support that requiring offenders to settle their gm ETC-3‘10 1 {18 a i which in turn a. In general. we been repcrt AQO‘J . I. ‘ e. sense abc at. r. Ligact of trese i «ere I s bpen l 3);. Cljrr‘erlt,l l‘.8] . AS rfin '-9nt1. a. Still 19 own wrongdoing and behave in a just way affects attitudes which in turn affect social behavior (Thorvaldson, 1980b). In general, empirical analyses of restitution programs have been reported only since the late 1970’s and most early studies looked only at its impact on victim attitudes (Schneider & Schneider, 1985). There is also little evidence about how offenders perceive restitution or the impact of these perceptions on success and failure (Van Voorhis, 1983). Harland and associates (1979) pointed to deficiencies in methodological sophistication. Keve (1978) argued there were no convincing reports from competent research about the rehabilitative effectiveness of restitution. Gendreau and Ross (1987) noted that most programs have not been described in sufficient detail to determine their integrity. Recent research has also failed to examine program components which may account for the success of restitution programs (Armstrong, 1983). Despite the interest in restitution, researchers and practitioners have generally not made attempts to build on the work of others who came before them (Hudson & Galaway, 1977). Further, there has been little effort to systematically integrate what is currently known about restitution (Hudson & Galaway, 1978). As recently as 1985, there were claims that restitution was still experimental and research was needed to explore .LA. Li its effectiveness .: lassachusetts EB? research on 3‘1““? Tze primary concet was done to then, experts think the 11an the offense Isney being paid cannot he expects A “Ulber Cf 9L F "9‘9 haVe been Mgr-ans. Huds r. EXP’eriiental et- a ‘Cnunity SQTVlC an; F fOUr Stud. e We involved a 20 its effectiveness. Andrew Klein, the founder of the Quincy, Massachusetts ”Earn-It" restitution program, argued that research on juveniles’ perceptions of restitution is needed. The primary concern should be the youths’ perception of what was done to them, regardless of what professionals and experts think they are doing to youths. If a youth does not link the offense with the service, or understand where the money being paid to the court is really going, restitution cannot be expected to be successful ("Growing", 1985). A number of previous reviewers have demonstrated that there have been few experimental studies of restitution programs. Hudson and Chesney (1978) uncovered only one experimental evaluation. In a review of restitution and community service studies, Hudson and Galaway (1980) found only four studies with experimental designs. Studies which have involved a control group have also shown little equivalence between the control and experimental groups (Gendreau & Ross, 1987). Two recent computer literature searches of the National Criminal Justice Reference Service (NCJRS) and Criminal Justice Periodical Index databases explored community service and financial restitution programs for both juveniles and adults. The literature on financial restitution programs and adult offenders was included because there are many themes which cut across all forms of restitution and therefore have relevance to a study of juvenile connunit literature 5355‘” available that 5‘. Citations. Stud. :nccrporated intc The researc.L fcra or through .\ categories: atti Studies are descr briefly and only are not directly Next, the outcom‘ apex-mental or e ~0nclusions of p. 21 juvenile community service. In addition, previous literature suggested that there was so little research available that such restrictions would result in few citations. Studies cited in other reviews were also incorporated into this review. The research studies which were available in published form or through NCJRS were divided into two primary categories: attitude studies and outcome studies. Attitude studies are described first. These studies are summarized briefly and only general results are presented because they are not directly relevant to the issues of this study. Next, the outcome studies were classified as quasi— experimental or experimental studies. First, the general conclusions of previous reviews will be discussed. Reviews of Restitution Research Chesney and associates’ (1978) review classified studies as descriptive, attitude, or evaluation studies. While their methods were not systematic, the authors identified two descriptive studies which suggested a high use of restitution by judges. The authors concluded that nine attitude surveys have shown that most judges, victims, offenders, community members, legislators and corrections workers were in favor of the use of restitution. Two major evaluation studies were also identified, both of which were residential programs. Neither study included restitution as the only treatment and the second study included no outcome late, illustratir- research on rest:' A later reVi' systematic and d4 :znaents to come restitution rese. 5:: their review selection. They had been publish . i'hnh included e Studies included evaluation studi P'TJEraI. The au5 of control, ther- stud: GS l8Cked j, The remain 3f restitution, pm resDense rt 22 data, illustrating some of the limitations of the existing research on restitution. A later review by Hudson and Galaway (1980) was more systematic and detailed. Perhaps one the most enlightening comments to come from this review of community service and restitution research was that decisions to include studies for their review were more a matter of acquisition than of selection. They identified 43 studies, only four of which had been published in professional journals and only four of which included experimental designs. None of the these studies included pretests. Further, none of the 31 program evaluation studies offered clear descriptions of the program. The authors concluded that these studies were void of control, there was no basis for comparison, and the studies lacked internal validity. The remaining 11 studies examined attitudes or opinions of restitution, eight of which relied on mailed surveys with poor response rates. However, the results of these studies consistently suggested that lay people as well as criminal justice professionals endorse restitution. Offenders also reported restitution to be a useful and fair sanction. Hudson and associates (1980) identified 336 articles, books, and reports about restitution. Included were 43 research studies of which 24 could be classified as outcome evaluations. While the authors acknowledged that the results of these studies were difficult to generalize from, they reported thz :‘eacnstrated this relatively low-c. addition, a 3:93 The results feicnstrate sup; :ffenders have ro :r rehabilitatrw punishment (Thor‘ Further, both \'1( 593 fair dispos; liaison, 1980; Th! lorS rat-Pd 23 they reported that the community service studies demonstrated that a large number of people were handled at a relatively low-cost and with few in-program failures. In addition, a great deal of work had been performed for the community. Attitude Studies The results of attitude studies collectively demonstrate support for the use of restitution. First, offenders have reported that restitution is more reparative or rehabilitative in its goals than it is a form of punishment (Thorvaldson, 1978, 1980a; Van Voorhis, 1985). Further, both victims and offenders perceived restitution to be a fair disposition and were satisfied with the sentence (Chesney, 1976; Hudson et al., 1980; Novack, Galaway, & Hudson, 1980; Thorvaldson, 1978, 1980a). Probation counselors rated the majority of their clients as somewhat or very cooperative in completing their restitution order (Steggerda & Dolphin, 1975). Criminal justice professionals have also reported their belief that restitution increased participants’ sense of responsibility and reduced recidivism (Chesney, 1976; Evans & Koederitz, 1983; Schneider et al., 1977). Quagi-experimentgl Outcome Studies Included in the group of quasi-experimental studies which examined outcome were a number of brief reports which provided recidivism rates of restitution participants or charges in crir: programs. The according to de rates, one grit variables rela‘ experimental 5' Crime Rates \ Marion Co. iecline in the F-‘C‘Eran for 1:: 961118, Kashj TEL-“A‘s had der restitution p {+72}. In ad 24 changes in crime rates of communities with restitution programs. The studies were grouped into five categories according to design: changes in crime rates, completion rates, one group posttest assessment of recidivism, variables related to success, and two-group quasi- experimental studies of outcome. Crime Rates Marion County, Oregon partially attributed an 18% decline in the crime rate in one year to a new restitution program for property offenders ("Juvenile Crime", 1978). Seattle, Washington reported that police contacts with youths had declined significantly (-11%) in the area of a restitution program as compared to the rest of the city (+7%). In addition, recidivism was two-thirds less than had been predicted during the 12 month follow-up ("'Accountability’", 1977). Completion Ratem A study of seven projects under the Community Service Restitution Program (CSRP) found that 87% of the offenders successfully completed all community service hours within a prescribed time frame and only 4% were rearrested during their assignments (Cooper & West, 1981a, 1981b). Schneider’s (1983) evaluation of 17,354 juveniles in the OJJDP sponsored programs found that 88% of those ordered to perform unpaid community service, 87% of those ordered to pay monetary restitution and 86% of those ordered to ._’ complete a conbi restitution suc: Nationally, com- iave been esti: initial projegq allcst 8E7. hat: l means i on" 7! rates of '7 'x I 'EEIngI-d , 15 The Jux ; “med as :2 restitutim “ o .ak Es “Ere and 19.x a. ‘ile a ll .‘ «‘3 COmmur‘ ‘C‘Q _. n‘lete, The 25 complete a combination of community service and monetary restitution successfully completed their orders. Nationally, completion rates of restitution participants have been estimated to be better than 80%. Results of the initial projects funded by OJJDP in 1978 found that 87% of the 12,000 juveniles completed their restitution program and almost 86% had no subsequent contact with the court ("Expansion", 1981). Other studies have found completion rates of 75% (Nelson, 1978), 91% (Maori, 1978), and 98% (Keldgord, 1978). One Gromp Post-test Only The juvenile restitution initiative which began in 1978 funded 85 programs across the United States. In the first two years of the project, 17,354 offenders were referred to restitution projects. The authors found that reoffense rates were 4% in the first three months, 8% at six months, and 14% at 12 months (Schneider, Schneider, Griffith, & Wilson, 1982). A study of 24,915 offenders who participated in community service in England found that only 9% had been convicted of another crime (Pease, 1981). The Charleston County Juvenile Restitution Program (JRP) was developed in 1979 to teach offenders accountability through the performance of community service. Follow-up reports collected on 112 participants 90 days after program termination found that 5% of the youths had henrearrested. Efiflxd,l981l. In addition. reported that the leldgsrd (1978) a agencies who part mt Some studies in restitution ar- personality \‘aris were “lemme to if?“ th‘. outco.w g33£l£e:;g; mined the diff C" u , “melerlstics the OJJDP JUVeni 313209 ssful Comp]. 0f ’h’) " “SQ . _ Rho did “3:13, to that 3.54; . 1“ ‘- son, 26 been rearrested, all of whom were incarcerated as a result (Hofford, 1981). In addition, participants of restitution programs have reported that the programs were helpful (Macri, 1978). Keldgord (1978) also found that nearly three-fourths of the agencies who participated said that their needs had been met. Eggiables Related to Success Some studies examined the relationship between success in restitution and characteristics of the programs or personality variables of the participants. These studies were attempts to examine variables which may differentially affect the outcome of restitution programs. Organizational characteristics. Schneider (1983) examined the differential impact of certain organizational characteristics of restitution programs which were part of the OJJDP Juvenile Restitution Initiative. He found that successful completion rates and in-program reoffense rates of those who did not participate in victim mediation was similar to that for those who did have victim mediation. In addition, the differences in completion rates and in-program re-offense rates for those in community service programs, financial restitution programs and combined programs were not functionally significant. However, participants in sole sanction programs had completion rates of 94% which were significantly greater than those of'participants who had a‘,.'rI-’4*.iorial Slitr ‘5‘é 5 programs had less than tha' {31L Persotal. relationship 1 fir49 partic prtgram. A m tcwards the s, restit ution. 122variables V- rOU lJa A summar- is in Figure 5“ ‘ 27 additional sanctions (85%). Further, those in sole sanction programs had a reoffense rate of 15% which was significantly less than that of youths in programs with other sanctions (22%). Personality variables. Heide (1983) examined the relationship between personality variables and recidivism for 49 participants in a post-conviction restitution program. A number of the variables which assessed attitudes towards the self were successful at predicting completion of restitution. However, it should be noted that only 20 of 122 variables assessed revealed a significant relationship with completion. Two-group Quasi-Experimental Studies of Outcome A summary of the two-group quasi-experimental studies is in Figure 1. Brown (1983) compared the success rates of offenders ordered to pay restitution with those not so ordered. Newton (1979) and Pease and associates (Pease, Billingham, & Earnshaw, 1977) conducted a one-year follow-up study of community service, assessing reconviction rates of offenders referred to community service compared to a group referred to, but not given, community service. Challeen and Heinlen (1978) evaluated a program alternative to fines, jail and probation for adult, first- time, non—violent misdemeanants. Experimental group participants had some say in the form of restitution while controls had been sent to jail during this same period. --.--< l...“ u - «‘3‘ :19 9‘; .. ‘ $1.81 3; ’~.-‘ .1.-'.-‘!. t a, I'c'm ‘~—- \A.‘ g, .3? 'QLM. .""- '1‘ 5:. ‘1 ..:81) ‘3 ‘ ‘flflfl ~ 135:) \‘i‘Y‘a A A‘ Q “Q 1:92} hug. Expat-i: "'V‘oa‘w 1 ..\.u.‘—. mane ‘J CS mone t. 58 W191. CS . b Wrist imap 28 Authors EXperimental Control Outcome Results Treatment Treatment Subjects Measures(n) Brown (1983) monetary none adult Recidivism ns Newton (1979); Passe et al (1977) cs no cs adult Recidivism ns Challeen & Heinlen (1978) monetary jail adult Recidivism ns Cannon & Stanford (1981) cs none juvenile Recidivism pos work none juvenile Recidivism pos Crotty & Meier (1980) monetary, cs, both none juvenile Recidivism pos? Bonta et a1 (1983) monetary and incar- incarceration ceration adult Recidivism ns Shichor & Binder (1982) monetary no petition juvenile Recidivism ns Miller (1981) monetary and probation probation Recidivism ns Heinz et a1 (1976) monetary parole adult Recidivism(2) pos,ns Employment pos cs = community service. pos = statistically significant results favoring the experimental group. pos? = the direction of the results favored the experimental group, but the statistical significance was not reported. ns = no statistically significant differences. Figure 1: Two-groupiQuasi-experimental Studies in Z we I inlvlfi .a....\v l l\-.- ,. Ink. AA. 3 " Pi 4! ‘Uocd o. it 2: ... C. L . .L 3.. an t. .L . . - E C .. S Q Co 1 t» f. 5. T 9. t _. u... T. .3 5.. 1‘ 1, ”4 n .~.d ‘ . I; Q.» h; in g. A i. C. \b‘ . . r“ .l. by.“ Q. 9 w :4 .d .u ole .H. 1» U a: s an. . u r P“ v . . 3 S . : ‘9. E t . .u . . . 4 . . l . . . . I . . w. . . .u.» an» i. C; am. . ‘ ‘. O. u x; Cu: x. .. .\. .l q. t. s. a: 4‘ he .. .u u... «Li. 3.. .a. no. 29 Cannon and Stanford (1981) studied a work and community service restitution program for youthful property offenders. The comparison group was selected from referrals received during the same period two years earlier. Subjects were followed for nine months after referral. Crotty and Meier (1980) tested a program where experimentals received probation and restitution while youths in the comparison group were selected from similar referrals made before the restitution program was implemented. Bonta and associates (Bonta, Boyle, Motiuk, & Sonnichsen, 1983) compared offenders who were incarcerated, willing to pay restitution, and eligible for placement at a community resource center to those in the center's Temporary Absence Program who had no restitution agreement. Shichor and Binder (1982) evaluated a community restitution program where youths referred from three police departments were compared to youths from a fourth department. Participants were youths who would not have been petitioned if there were no restitution program. Miller (1981) compared a sample of files of probationers ordered to pay restitution with a matched group ordered to probation only. Heinz and associates (Heinz, Galaway, & Hudson, 1976) conducted a 16 month follow—up study with new prison admissions referred to a restitution center who agreed to pay restitution and compared them to a matched group of men released on conventional parole. These st restitution ; for controls. :ztccee varia noyfidfne' r. .r'.~.\...'8n S ' ¢ ::ese studies ‘ generaiizat il .: ;arzs::n g: experiz. eta Y I \AI ‘ ‘3. . ‘.:~ ’I .LJ :0“ E 3.5" .11, to per 3:». 30 These studies found that the recidivism rates of restitution participants were equal to or lower than those for controls. The one study which assessed a different outcome variable found that significantly more restitution participants than controls were employed. However, five of these studies used adults as subjects, making» generalizability to juveniles questionable. Further, the comparison groups were often not comparable to the experimental groups. The quasi-experimental nature of the designs also restricts the validity of these findings. Conclusions Twenty-four quasi-experimental studies were described in this section. Only ten of these studies included a comparison group whose participants were not required to complete restitution (Bonta et al., 1983; Brown, 1983; Cannon & Stanford, 1981; Challeen & Heinlen, 1978; Crotty & Meier, 1980; Heinz et al., 1976; Miller, 1981; Newton, 1979; Pease et al., 1977; Shichor & Binder, 1982). Within these studies, the control groups were often not equivalent to the experimental group. For example, participants in the restitution sample were compared to probationers (Miller, 1981), to persons in jail (Challeen & Heinlen, 1978), and to persons who were referred during an earlier time periods (Cannon & Stanford, 1981; Crotty & Meier, 1980). In addition, all but one of the studies (Heinz et al., 1976) assessed recidivism as the only outcome variable and many ..s '39:: .: F. I ‘L »L‘ .“-o ‘Q . 31 authors did not report statistical tests of significance. Overall, however, these studies suggest that the recidivism rates of those who participated in a restitution program were equal to or lower than that for controls. Experimental Outcome Studies Few experimental studies were located through the literature search procedures described earlier. Those that were identified are summarized in Figure 2 and described below. Property offenders randomly selected from new prison admissions referred to a restitution center were randomly assigned to restitution versus parole or discharge (Minnesota Department of Corrections, 1977, cited in Hudson & Chesney, 1978). However, offenders were asked to participate in the study after they were told of their treatment condition. Four experimentals refused, with no comparable drop—out for controls. Similarly, the parole board could deny entry into the program, for experimentals only, which it did in nine of 72 cases in the first two years. Restitution was not the only difference in sentence between the groups. The level of parole supervision was also greater for the experimental group than the control group. Thus, it is difficult to determine if the differences in recidivism rates were due to participation in restitution or the level of parole supervision. O o.’ D‘ a. "NI‘U‘ 91:33} r. ”.8 4.... see a v '36 3.55:8 e p n he mm mcmp 1935) i: 1‘. r :3?“ s cm. he. “I“ at. 32 Authors Experimental Control Outcome Results Treatment Treatment Subjects Measures (n) Minnesota Department of Corrections monetary parole or (1977 ) discharge adult Recidivism pos Wax (1977) cs none juvenile Recidivism ns School atten ns Social beh ns Predict del ns Schneider monetary or (1986) cs detention juvenile Recidivism(4) ns cs or monetary with mediation probation juvenile Recidivism(4) pos,ns Monetary or traditional cs (alone or (probation or with counsel- incarceration) ing) and counseling & probation probation juvenile Recidivism( 4 )pos , ns monetary or cs, cs or monetary with probation traditional juvenile Recidivism(4) ns Wilson (1982) unknown probation juvenile Recidivism(2) pos? Koch (1985) cs diversion juvenile labeling ns or tradition bonding ns Recidivism(2) ns Davidson & cs diversion juvenile labeling(4) ns Johnson or tradition employ pos (in press) education ns parent involv ns prosocial ns Recidivism ns cs = commity service. pos = statistically significant results favoring the experimental group. pos? = the direction of the results favored the experimental group, but the statistical significance was not reported. ns = no statistically significant differences. Figure 2: Experimental Studies ‘. L j \‘iCe on 25' k . F Uta-- I::9‘,"'§ ‘ Ar .yhh“ 0 ‘-\»pa 6‘ II ”a 5.11- 5‘ ~ n? .c:; men 5? L, rDQ’ -H v' “tam C :.‘ “ fill bruouv" \ on: n27- ~. Qrefl 4 ‘r ‘4‘16 . Maths 33 Wax (1977) assessed the impact of 20 hours of community service on juvenile shoplifters. Thirty juveniles were randomly assigned to one of three treatment conditions: restitution with the victim present at sentencing, restitution without the victim present at sentencing, and no community service. A number of problems existed with this research. First, the sample size was extremely small. Second, those subjects who needed services in addition to community service were excluded from the study. Those who failed to complete community service were also excluded. While the intervention was completed in two weeks, the posttest was not completed for six months and there were no significant differences on any of the four outcome measures. Because of the delayed posttest, immediate effects could not be detected. Schneider (1986) examined four different studies. First, restitution was compared to short-term detention. Experimental youths were ordered to pay monetary restitution where there was a financial loss and community service restitution where there was no financial loss. Controls were ordered to an average of four weekends in detention and nine months probation. The second study compared victim- mediation restitution to probation. Because uncooperative youths could create problems during mediation, all youths assigned to restitution were allowed to reject their .i raised tr. The tn: :::r.sel:r.g. services figs 2: the first genera: of y "“erent fr ‘teA~s 1...“. ‘ ‘ A "32"EVus‘JJ I Estuutign C services. A; astlthtion C titer Alalf {)6 In all : ....eates of ~-.-1 that ’J 34 assignment and receive probation instead. Approximately 40% exercised this option. The third study compared four conditions: restitution, counseling, restitution and counseling, and traditional services (usually incarceration or probation). All youths in the first three groups were also on probation. Seven percent of youths were ultimately placed in a group different from that randomly assigned to because of judicial discretion. The majority of youths (60%) in the restitution conditions performed community service. The final study assigned all cases where a monetary loss was determined to restitution only, restitution and probation, and traditional services. Approximately half of the youths in each of the restitution conditions paid monetary restitution while the other half performed community service. In all sites, juveniles were randomly assigned to treatment conditions. However, all studies also allowed the local jurisdiction to change any assignment. All analyses were conducted on the groups to which participants were assigned, rather the type of service received. Four estimates of recidivism were used in all studies. Results found that the recidivism rates of experimentals were equal to or lower than those for controls. Wilson (1982) studied a program where offenders recommended for probation were randomly assigned to restitution or probation. Unfortunately, the report did not e . , “-alnin we 1 u‘ R:‘ 5‘ ¢ “U‘ k.1\-e 1:: 5-1. .. ‘UQDN‘. 146 'M '1‘?» - hf V I. .1 pl h at .‘5. “.71: 35 describe what services were provided in the various conditions, and the information which was provided suggested that subjects in all conditions were ordered to make some form of restitution. Further, less than two-thirds of the participants obtained from the probation group complied with the restitution order. This is substantially lower than most other studies. The low completion rates may be related to the high reoffense rates. However, experimentals had lower rates of recidivism than did controls. Koch (1985) and Davidson and Johnson (in press) compared youths assigned to diversion, diversion with community service arbitration, and tradition court processing. All youths in the community service program performed 3 to 4 hours of community service per week for 12 weeks, regardless of the offense. While the results pertaining to labeling and bonding were disappointing, the low reliability of the scales may contribute to the absence of positive findings. This was true for both the original and subsequent study. The only positive finding was for employment expectations (Davidson & Johnson, in press). Conclusions A total of 24 quasi-experimental outcome and 10 experimental studies were included in this review, the majority of which were plagued with methodological problems. For those quasi-experimental studies which did include a comparison group, there were often major differences between .33.. ””11 Incl, ”fig“ 3"“ .ollt‘al re 2:213 9“ a .p a 13’ that the r“ .east equal eere eontr' 36 the groups before the intervention began (Cannon & Stanford, 1981; Challeen & Heinlen, 1978; Crotty & Meier, 1980; Miller, 1981). Some of the studies had small sample sizes (Wax, 1977). The majority of studies relied solely on official records as the means for obtaining outcome data (Bonta et al., 1983; Brown, 1983; Cannon & Stanford, 1981; Challeen & Heinlen, 1978; Crotty & Meier, 1980; Miller, 1981; Schneider, 1985; Wilson, 1982). Few of the studies included an adequate description of services and often more than just completion of restitution differentiated the services received by participants in the different treatment conditions (Heinz et al., 1976; Hudson & Chesney, 1978; Wilson, 1982). A number of the studies did not make the restitution order related to the crime, a factor which has been cited as one of the keys to a successful program. Only four of the studies examined outcome variables other than recidivism. However, it is impossible to ignore the fact that a large number of studies have consistently concluded that the recidivism rates of restitution participants are at least equal to and sometimes lower than those of controls, where controls have usually received traditional services (e.g., parole, incarceration) or no service. Further, the political, economic and social benefits to the victims, community and justice system, have been proposed as adequate justification for restitution programs (Keve, 1978). benefit a n'. ‘ "1‘s Muted in 1’; '- “t 5,, to i 25:: 37 Theories of Delinquency The above review suggests that restitution programs can benefit a number of people and may reduce recidivism. While restitution has not emerged from a theory of delinquency, such a theory can be used to guide its evaluation. Theories are useful ways to think about not only how youths become involved in delinquency, but also how youths stop committing crimes (Fagan, 1988). As such, outcome measures can be developed or selected from a theory which is consistent with an intervention. This evaluation used this process to select intermediate outcome variables. The theory used is described below. I A recent effort to integrate three of the more lasting and supported uni-causal delinquency theories: strain theory, social learning theory, and social control theory (Elliott, Ageton, & Canter, 1979; Elliott, Huizinga, & Ageton, 1985) has been met with a great deal of interest. Each of these theories offers positive contributions to the understanding of delinquency, while each also has limitations. While this has not been the only recent attempt to integrate two or more current theories of delinquency (e.g., Hepburn, 1977; Simons, Miller, & Aigner, 1980; Thornberry, 1987), the studies on which it is based are the most rigorous and detailed. Results from these studies suggest a high degree of explanatory power and stable relationships between theoretical variables. It is therefore or I; :’fi ‘“‘“:19an -2€refcre, i. z!- ”5328 w ‘n d E r“ l :t 38 therefore one of the most appealing of the integrated models to be proposed. The three original theories will be presented and their limitations discussed. The multi- dimensional causal theory will then be presented. Strain Theory Strain theory argues that delinquency results from frustrated needs or wants. This theory suggests that delinquency is behavior oriented towards conventional goals. Therefore, when there are inadequate socially acceptable means to achieve these goals, alternative means will be used. This model emerged out of the hypothesis that all youths internalize conventional goals of success but that lower class youths, in particular, are denied access to these goals because of their social class (Cloward & Ohlin, 1960). Recently, however, theorists have tried to show how strain theory can also account for the fact that middle- class youths also become delinquent. It has been suggested that middle-class youths are just as likely to aspire beyond their means and because the difference between goals and opportunities are relative, middle-class youths may also engage in delinquent behavior to achieve these goals (Elliott & Voss, 1974). Nevertheless, this discrepancy is often greatest for lower-class youths. Further, what is important is the perception of youths that socially mail 1 reestabe n OvV ‘. \I‘Pk‘ fl “3.15.1.9 tk- .i;.e.,.,. as» Instant. ‘ \ h" 4. ¢ .2...‘3C 19; OE :“te, .. ‘\u~.‘v‘: are 39 acceptable means of achieving desired goals are not available to them (Elliott & Voss, 1974). Limitations The major problem with strain theory is that it is not able to explain why many lower—class youths do not become delinquent. The model assumes that the pressure to achieve a level of economic status above that which a person has is constant. If all economically disadvantaged persons are denied legitimate means to this goal, by definition, then all should become delinquent. However, this does not happen. Secondly, while recent theorists have attempted to demonstrate that this theory can explain middle-class delinquency, these efforts have fallen short. Middle—class persons are not denied conventional and legitimate opportunities for success. Third, this theory ignores individual values. What may appear to be normlessness by one person may actually be different norms. Further, lower- class persons may not strive towards conventional goals at all. It may be that lower-class youths have goals unique to their social status and that they are achieving status in their own subculture by acting out. Strain theory also ignores the importance of peer influences on delinquent behavior. . ‘1 Bhue s . t'. a .ueru. es, heavier is intimate gr: I: N , .. er. eras :‘ier definit 40 Social Learning Theory While strain theory ignores the importance of social influences, social learning theory argues that criminal behavior is learned in interactions with other persons in intimate groups and that a person becomes delinquent because of "an excess of definitions favorable to violation of law over definitions unfavorable to violation of law" (Sutherland & Cressey, 1978, p.81). Not only are techniques for committing crimes learned, but also definitions which make a person willing to break the law are learned. Akers (1977) added the notion of differential reinforcement; behavior is conditioned by the consequences it has. Given two choices, the act which is reinforced to the greatest amount, frequency and with the greatest probability will be maintained. Behavior is therefore determined by the expected rewards and punishments for engaging in certain acts, as well as the rewards and punishments anticipated with alternative acts. Therefore differential social reinforcement directs the decision to engage in conforming or deviant behavior. Akers, Krohn, Lanza-Kaduce, and Radosevich (1979) found that this social learning theory explained one half of the variance in drinking behavior and two-thirds of the variance in marijuana behavior of 3,000 youths. The importance of the components used in explaining this variance was ranked in the following order: differential association, :IELned so- d-::pwon91a IsnthsMHs ‘ PH \ dh“L O C “"14 flail .- . z":‘n..‘. ..““‘LL . o o" naqu'ah‘ VA )T‘le ‘ c LEE-1c; ‘~-Vus ‘} " A .n‘ "-lf-m. "149“,? . ‘:-‘ "\ I ~Fa1n ur I f‘ “Ring Wig. L .1“ h Quad |.. “i u, 41 definitions favorable to or unfavorable toward drug use, combined social/nonsocial differential reinforcement, differential social reinforcement, and imitation. Limitations There are a number of criticisms which have been levied against Sutherland’s theory. First, not everyone who comes in contact with a criminal becomes a criminal themselves. Further, the theory does not explain why people associate with criminals in the first place and does not identify the source of definitions favorable or unfavorable to the law. Additionally, the theory has been criticized because of the difficulty in operationalizing "excess of definitions", "favorable to", and "unfavorable to". Akers’ (1977) reformulation of this model does not adequately address these limitations. Control Theory While social learning theory proposes that there are social patterns which favor delinquency, control theory suggests that it is the absence of controls which permits delinquency and the strength of social controls which restrain unconventional means of achieving goals. People conform because social controls have been effective, but criminality will emerge when these controls break down. Control theories have focused upon the process of social bonding as a means of social control. The stronger that these bonds are, the more a person’s behavior will be :ortrelled. gerscn is in :2: are not estrcng ccrr. attitudes to teri'entionai “o 1. .ecnuler Md be jeo :“ad e “w ¥ hat (3 42 controlled. When these conventional ties do not exist, a person is free to engage in deviant acts. Therefore, youths who are not attached to conventional groups, who do not have a strong commitment to prosocial goals, have negative attitudes toward obeying the law and are uninvolved in conventional activities are more likely to become delinquent (Altschuler & Armstrong, 1989). Hirschi (1969) identified four elements to this bond: attachment, commitment, involvement, and belief. Attachment is defined as a moral link to others, particularly family, peers, and school. Commitment is described as a rational investment in conventional goals, especially those that would be jeopardized by deviance. Briar and Piliavin (1965) noted that commitment is not an irreversible process. They asserted that those with low levels of commitment may have experiences which increase their stakes in conformity and lead to conventional behavior. Involvement suggests that time spent in conventional activities restricts the amount of time available for delinquent activities. Belief in the moral values of society is the last element of these bonds. These elements are all interrelated so that the weakening of one is accompanied by the weakening of another. Social control theory suggests that delinquency is largely the result of inadequate or nonexistent social integration from various groups, such as family and peers. For youths who are involved in relationships that would be threatened b1 social norms ‘:-:6:O;fips o.l-.u-45A --I .- 1 ‘I titer)" 1331-13 :t‘ NEED: to a d-n\ l.‘,‘ W .‘3..:‘C (Ahoy . r . ”van“, "I'Li‘uthtve tc. ic‘. «‘qu ‘9 MP -.io ~ 4. Elm “lie 8 3": “‘il,‘ “t fir l have Etta g e “sins of 43 threatened by delinquency, or who are uneasy violating social norms, the costs of delinquency are high. Limitations One limitation of control theory is that delinquents may not share conventional morality. Further, control theory lacks a motivational component. This theory does not attempt to account for external pressures which generate delinquency. While a lack of controls may be a factor conducive to delinquency, it is not a generative factor. Nevertheless, social bonding has been supported by a good deal of empirical evidence (Akers, 1977). The theory, however, lacks information about the role that sanctions may play in the development of delinquency. The Multi-dimggsional Theoretical Model By acknowledging the limitations of the individual theories, Elliott and associates (Elliott et al., 1979, 1985) have developed a hybrid model which is based on the strengths of strain, social learning and control theories. They suggested that there is support for the notion that strain has a direct cause on delinquency and that strain is also moderated by controls. In other words, not only does strain lead directly to delinquency, but strain may also weaken social controls which, in turn, may lead to delinquency. While social learning and control theories emerged from different traditions, there are some common assumptions of salt of 3 ‘ipqnarfl\' e....-,...i -. "6 Fr VI ‘ S a ...9 C ‘4‘ ny‘v-n‘ .iflvh¥4 . ‘ia. -¢A find with 44 these models. First, both models argue that behavior is the result of a rational weighing of costs and rewards associated with the various behavior options. Second, delinquency is viewed as the result of differential socialization and the most important source of social controls are social relationships. Control theory asserts that the content of socialization is constant and what varies is how well it works. In other words, the ability of the child to internalize the norms, the ability of the parent or teacher to teach the norms and the various circumstances under which this process occurs determines the effectiveness of the socialization process. On the other hand, social learning theory postulates that the content of the messages of socialization are not constant and therefore there is variation in what is internalized. This model was tested by means of path analysis on data from 1,725 youths (Elliott et al., 1985). The resulting model is presented in Figure 3. The authors found that youths with high conventional bonding were lowest in terms of delinquent behaviors. Results also suggested that conventional bonding insulates against bonding to delinquent peers. Further, low conventional bonding, when combined with high delinquent bonding, was associated with more delinquent behaviors. They concluded that strong conventional bonds reduce the likelihood of developing strong delinquent bonds. Further, when associations with Ammcfiv scuow< w .mwcwwflaz .qufiHHm aoum Hmcos Hmcowmcmefieapass .m museum .. Sausage - 45 =38 mafia - . r 832.8 “egos—on 3.98m — cot-£7.98 Hucoscczoa 2 2.93m .323350 7 . — 8363.5 29.5 .303 r ‘ 5.5m — agave . Bananas» .8...— e v 1",gquenu ’ :tgoud‘ fl re.:ce .‘ ' i. ,....‘ "e e..r...... un‘ "‘=zrated a .5. nieis on .. n Y .28 S. I ..Vlo-" " :3': .d.. VH7 «I OPI- .:._ h -. n . . . '- ‘ a. ‘ '9 ‘:‘A|:,4J . r H: .I'Iie" k...“ .‘ \ . '- . ‘ ‘ “M. . ‘e“”““e. ‘ I no .. ‘u‘ U‘ .I.‘ .5. ‘ ‘ A I‘ .“ 1::0\ aa. .. ‘ + i H... ‘7 .r .i":‘«¢.a..\ . U~~ 1 35d . c “HI-.1 . -... "¢S' S :‘p‘.’ ‘ ‘u‘ ‘ “eggs 5 \L‘ ““S ['1‘ 32:} ‘ .n‘. ‘n e". 5..“ 1%” E‘llr a -. . '5‘ . K ”‘5’ 8‘. b ‘. 46 delinquent peers are developed, strong conventional bonds will reduce the strength of the pro-delinquent influences of the delinquent group. The predictive ability of this integrated model is greater than that of the individual models on which it was based. In a similar path analysis test of differential association and social control theories, Hepburn (1977) found that while delinquency may be the result of an absence of constraints on behavior, particularly when there is a lack of perceived family support, it may also be the result of delinquent associations. The data support the notion, however, that delinquent definitions precede the development of delinquent associations. The multi-dimensional causal model suggested by Elliott and associates simultaneously addresses issues of socialization, peer group influences, social bonding, and opportunities for achievement through conventional means. This model suggests that delinquency emerges from weak controls, strain, and peer influences. Intervention strategies should therefore be directed at these variables. Implications for Research This multi-dimensional causal model of delinquency is useful in evaluating a restitution program in a number of ways. Elliott and associates (1985) showed that weak controls, strain, and peer influences were related to delinquent behavior. Figure 4 shows the relationships II EE 47 Hypothesized Impact Theoretical of Restitution Variable Activities Positive relationship Conventional Involve youth with the community Bonding with positive social influences (worksite supervisors) Sense of accomplishment Conventional Completion or success Bonding letters, Less free time to spend with delinquent friends Delinquent Bonding certificates, praise from supervisor Less free time to spend with friends Figure 4: Relationship Between Hypothesized Impacts of Restitution and Theoretical Variables Being Tested ear, the n." r n ~- 3 p0snr ”. \..~a'as| v , u..l. 1.“ ....._‘E 5.. 48 between the expected outcomes of restitution and factors in this theoretical model, and restitution activities that address these variables. First, restitution is aimed directly at the issue of conventional bonding. Restitution programs, particularly in the form of community service, involve youths with positive social influences in the community, which may in turn strengthen conventional bonding and encourage conventional beliefs. On the other hand, it could be argued that because youths do not volunteer for community service but are coerced, there will be no commitment and therefore no bonding. In order to test this, conventional bonding should be assessed. Restitution may also reduce the impact of negative peer influences. Efforts to fulfill restitution minimize the amount of interaction with delinquent youths so as to avoid the negative consequences of delinquent bonding. Because youths may spend a substantial portion of their free time meeting the terms of the restitution order, the relationships with delinquent peers may break down. Therefore delinquent bonding should also be assessed. Goals of the Current Study While research into a new area often begins with less sophisticated evaluations, the issue of restitution can no longer be considered a new area of study. It is time that evaluations become more sophisticated and directed. A uvc0uu~ht' 1 ‘ 0‘. pm. '0: ""“ ‘\-~\.‘ . ,‘ "f 95 c' .4..u.‘_ U ead‘. .' - . "Pr-1:" "F .n.._k V“ :2““,r\ . _, .‘.1. -AU..‘..‘ . "25‘ c A .i. V-“ ”C. 49 number of authors indicated the need to develop programs based on what has already been learned and to not only construct, but also evaluate such programs in light the theory which is known (Hudson & Galaway, 1977, 1978). This study was therefore designed to evaluate a community service restitution program based on what is known about restitution and to examine intermediate outcome variables suggested by current delinquency theory. As a result, this study evaluated a community service program, founded on the accountability philosophy with proportional sanctioning. The goal of this research was to evaluate the impact of a community service program guided by the philosophy of accountability and with concrete operationalization of each program component. What also makes this research unique is its examination of the impact of such a program in relation to contemporary theories of delinquency. This research utilized a unique set of measures, including some developed by and used to test certain dimensions (see Figure 5) included in the multi— dimensional causal model of Elliott and associates (1979, 1985). Research questions to be answered were of four different types. First the intervention alternatives needed to be documented so as to confirm that the only differences between the two groups was performance of community service. 4.. "Anni. é ."‘ en; ‘.a 55“ 50 Variables in Theoretical Model Measures Used Prior Self-reported Delinquency Weak Conventional Bonding Strong Bonding to Deliquent Peers Associations Scale Delinquent Behavior Court petitions related to prior offenses General Bonding Scale Specific Bonding Scale Delinquent Normative Pressure Scale Subsequent court petitions Figure 5: Variables in the theoretical model measured in this study 1. 53' 'J-‘C :ndit: servzce h :riered re ”Ha-«1'. A , ‘ men. ousane .. 51 1. What was the intervention process for each of the two conditions? Specifically, the number of community service hours, the number and type of additional court ordered requirements, the number of pretest and posttest appointments that were missed, the time from offense to pretest, the time from offense to community service, the time from pretest to community service, and the time from pretest and posttest were used to describe and compare the groups. 2. To what extent was community service implemented? Second, the integrity of the community service program was determined by the extent to which it met the goals of accountability. 3. Were community service participants more likely than controls to understand the purpose of community service to be reparation? 4. Did community service participants perceive their community service order to be more fair than did controls? 5. Did community service participants feel more favorable toward their victim than did controls? 6. Did community service participants have greater feelings of accountability than did controls? Third, it was important to examine the impact that participation in a community service program had on the youths. Each of the remaining research questions examined p'u. ‘ 0.;- p u .- -.. E‘ a - u I; -pa ‘sooinu . . kappa u "I ~ ~--ao-A: on (O Y e a... 1., 52 constructs suggested by part of the multi-dimensional causal theoretical model described. 7. Did community service participants have less delinquent friends than did controls? This addressed bonding to delinquent peers. 8. Did community service participants feel less pressure from their friends than did controls? This question also addressed bonding to delinquent peers. 9. Did community service participants have more positive attitudes towards traditional beliefs than did corrtrols? This question served to address conventional bonding. 10. Did community service participants report greater self-esteem than did controls? Restitution is thought to give participants a sense of accomplishment and therefore improve self-esteem. 11. Did community service participants have lower recidivism rates than controls? This addressed delinquent behavior. Finally, it was necessary to determine the extent to which the intermediate variables of accountability and conventional and delinquent bonding were related to recidivism. 12. Did subjects who had high scores on accountability measures at posttest have more positive outcome results than those with low scores? fl.u . .‘4 .C. Alli ‘5 A.-. ..A a. . man in lit .01. 5'5 u-- A... A» o . a. 14. a'd o t ‘u .3 II. ‘V 0 u on. .v «d . 55 community service is considered a violation of probation, which could result in additional court proceedings where ‘programming is reconsidered. The distribution of Informals sand Formals was similar to that of previous years (43% and 557%, respectively, in 1987). Because informals were usually ffiirst-time offenders and often had no sanction other than cuammunity service and formals had both prior contact with 'tlie court and were on probation, all outcome analyses cnampared subjects on this "status" variable. The number of participants required for this study was (digrected by the need to achieve adequate statistical power. Itemzidivism rates from previous studies (Cannon & Stanford, 319E31; Challeen & Heinlen, 1978; Crotty & Meier, 1980; Heinz et; al., 1976; Hudson & Chesney, 1978; Miller, 1981; Scflineider, 1986; Schneider & Schneider, 1985; Shichor & Bidnder, 1982; Wilson, 1982) were used to create an average Effect size (.1) of .50, which was then used to determine the: sample size. While this estimate of the effect size Seenm:to be high, it is the best estimate available. In Order to achieve 70% power with two equal groups, a minimum of 82 youths were required to participate in this study (Kraemer & Thiemann, 1987). However, given the fact that these original studies had mixed findings and often found no differences between experimentals and controls, .50 is likely an overestimate of effect size. Therefore, 82 subjects will likely not be enough to achieve 70% power. Tze restric‘. irsfificien‘ Ail su' treatment a tie Cmnuni sere assign service obi EIC‘J;S were 0f the Success fu 1; 5’3? refus ”P‘Erinent; 1|" “ian . 101 3'1" : . " *amilr $4;. a-‘litlon’ .S,’ HE: E: ‘A Erkgv . “‘91", 0 ’fire; . radiilpa‘ Ff“ qu 1n , “t Su «L ’s. 56 The restriction on statistical power that is created by an insufficiently large enough sample cannot be ignored. All subjects were randomly assigned to one of two ‘treatment alternatives. Forty-seven youths were assigned to time Community Service Program (experimentals) and 47 youths evere assigned to traditional services without a community seervice obligation (controls). The youths in these two groups were very similar. Subject Mortality Of the 94 youths who agreed to participate, 81 were sniccessfully posttested. One youth in the experimental group refused to complete the posttest, while four other e’ijerimentals and five controls had moved out of the area to Unknown locations (no forwarding address and/or caseworkers arui family members were unaware of their location). In addition, one youth in the experimental group did not cCnnplete the posttest prior to project completion. One zidditional experimental youth did not complete the community Service prior to project termination and this subject and his/her matched control were therefore also not posttested. HOwever, official records were examined for all Participants. Four of the original 47 community service participants did not successfully complete their community service obligation. These subjects were excluded from outcome analyses because they did not receive the specified 5 ‘: ir'erx'er... E as very s be ' 4"'9'?” .'C Hal‘- ’n '5: fC b_..-"‘ S‘ ‘l‘; r‘.‘tisns' E'3v: ‘:‘|‘ 57 intervention. The number of subjects for whom this was true was very small and analyses which included these showed that the direction of findings did not change (see Appendix A). .Figure 6 shows the flow and attrition of subjects throughout ‘the course of the study. Background characteristics are presented in Table 1. (Iverall, 68% were male, 60% were white, 19% were black, 17% eveme mexican, 42% lived with one parent/guardian, and in 27% c>f'the cases, no parent/guardian in the household was snorking. One-third (38%) of the youths had held a job in tJie past or at the time of the first interview, 12% had Irreyiously performed community service work, and the average age of the subjects was 15.6 years. The only significant (iinfferences between the groups was when whites were compared tC) all persons of color, combined (X3(1) = 5.34, p < .05). The majority of youths had no previous offenses, dfisfined as contacts with this particular court (See Table 2) . Fifty—seven percent of the experimentals and 61% of the (NDntrols had no prior petitions with approximately one- quarter of each group having one prior petition. The average number of prior petitions was .74 for experimentals and .57 for controls. There was a significant relationship between status (formal and informal) and number of prior petitions, such that youths with formal court orders had significantly more (F(1,84) = 27.99, p < .001) prior an A 5 ’ 1 O >-“ r" ‘1, C) C 58 Referrals (n=106) Accepted Refused and (n=12) Pretested (n:94) / \\ Experimental Control Group Group (n=47) (n=47) Ccnnpleted Did not "lost" Completed Posttest "lost" Community Complete (n=4) (n=41) (11:6) Service Community (11:39) Service ‘ (n=4) Completed Did not complete POsttest Posttest (n=38) (n=1) Figure 6: Flow of subjects through study, including attrition Title 1 u ‘ ' €2-’.’3“.’.; '_.L_L._ '37: {.8 C. o... t-Uhd: Table 1 59 Demographic Characteristics of the Sample Experimentals Controls Variable n (%) n (%) X2 (df) Status 0.00 (1) Formal 25 (53%) 25 (53%) Informal 22 (47%) 22 (47%) Gender 1.22 (1) Male 29 (62%) 35 (75%) Female 18 (38%) 12 (26%) Race 7.38 (3) White 34 (72%) 22 (47%) Black 5 (11%) 13 (28%) Mexican 7 (15%) 9 (19%) Mixed 1 (2%) 3 (6%) Number of "parents" in household 1.58 (1) One 16 (34%) 23 (49%)" Two 31 (66%) 24 (51%) Number with no "parents" working 10 (21%) 15 (35%) 0.87 (1) Youths with job - experience 19 (40%) 16 (35%) 0.12 (1) Previous Community Service 4 (10%) 4 (10%) 0.00 (1) Mean Age 15.6 years 15.5 years E(1,88)= 0.19 Title 2 ‘.h.r;'h-e ‘1] l ..=...Ld A' titer of F' _v ”he 0. ”.01. pet. Table 2 60 Distribution of Previous Offenses .Number of Experimentals Controls IPrdor petitions n (%) n (%) F(1,85) Zero 24 (57) 27 (61) One 9 (21) 10 (23) Two 6 (14) 6 (14) Three 2 (5) 1 (2) Fkaur 1 (2) 0 MEAN (SD) 0.74 (1.04) 0.57 (.82) 0.71 NOTE: Due to missing court records, available for all subjects. this data was not rer't‘ons p-vo \vL ',.o-. l '- ....ah6 Let The c 15.-12'“ “at 1.,“ ziztr a”, L.‘ 61 petitions (1.06) than did youths who were referred from the Intake Department (0.14). The current offenses for all youths are presented in Table 3. The offenses are listed roughly according to the seriousness of offense. Those youths with single charges are listed in the top portion of the table. Those youths who were referred to community service with multiple charges are listed at the bottom of the table. The distribution demonstrates that most youths were charged with relatively minor offenses and that most of these were crimes against property, particularly shoplifting (covered under larceny and retail fraud categories). Design An experimental design was used to assess the impact of community service participation on recidivism and the intervening variables discussed earlier. Subjects were nested in time and crossed on treatment condition. This experiment was a single factor repeated measures design. Pre—tests of all independent and dependent variables were used to assess the equivalence of the groups. Post- tests included these same measures. Official court records were examined as a measure of intervention integrity and recidivism. \,.07:L' t....b“ ---( r "‘ .“ —ell‘ . ' A ‘ u ‘A a 62 Table 3 Distribution of Offenses for Current Petition Experimentals Controls Offense n n Assault 1 2 Assault and Battery 1 1 Illegal Possession of a Credit Card 0 1 Larceny Under $100 11 11 Retail Fraud — 2nd Degree 2 4 Attempted Larceny 0 1 No Operator’s License 1 0 Trespassing 1 0 Aggravated Assault 0 2 Entry Without Breaking 1 0 Larceny Over $100 3 1 Possession of a Controlled Substance 1 1 Receiving & Concealing Stolen Goods Over $100 1 0 Unarmed Robbery 1 0 Malicious Destruction of Property Over $100 1 1 Violation of Probation or a Court Order 3 1 (table continues) 0:9, ...--‘ A. ..u 3.. .i- . FP" A p a u w A '- vzm a: u a \ o It .rJu . v y “44 V. :5 Ifiw N F. en.» 63 Table 3 (continued) Experimentals Controls Offense n n Breaking & Entering 4 7 Carrying a Concealed Weapon 1 2 Criminal Sexual Conduct 1 2 Unlawful Driving Away an Automobile (UDAA) 1 0 UDAA - Attempted 1 1 MULTIPLE CHARGES Assault & Battery - 2 Counts 1 0 Carrying a Dangerous Weapon AND Larceny Under $100 1 0 Retail Fraud II AND Malicious Dest. Property Under $100 0 1 Unlawful Use of Motor Vehicle AND Violation of Probation 1 0 Unarmed Robbery AND Violation of Probation 0 1 Breaking & Entering AND Breaking & Entering a Coin Machine/Box 1 0 Breaking & Entering AND Larceny Under $100 0 1 Breaking & Entering AND Receiving & Concealing Stolen Goods Over $100 1 1 Breaking & Entering AND Violation of Probation 1 1 UDAA AND Receiving & Concealing Stolen Goods Under $100 0 1 it?! 3‘- P: -L “‘--c.‘ 6.. w... I. -1 L» :o ... et at .va u. a. at .3 D. an. hm an .«a a,» ... :n u . r a (be. .H... . 1. .3 i. 2‘ .5. A . e .. s .. 61. .- _ 1 .q :. a: are 64 Procedures Referral Subjects were referred by the judge or the referee by means of a referral form (See Appendix B). The referral form included information about the youth as well as the number of community service hours to be performed and the date they were to be completed by. The number of community service hours was determined from a matrix which specified three classes of crimes and a range of appropriate hours (See Appendix C). Beginning with the maximum, a specified number of hours could be subtracted for youths who were in school full-time, had a job, participated in extra- curricular activities (related to both school or family and the court), were new referrals to the court, or met all of the above criteria. These procedures were adopted to ensure that the amount of community services was proportional to the harm that was done (Galaway, 1977a, 1977b; Schneider, 1985). It is important to acknowledge that this hierarchy of sanctions may, despite its design, result in disproportional sentencing because of variations in the way charges are made, either at the time of arrest or processing by the court. For example, one youth may be charged with assault (Class A) and another with assault with a deadly weapon (Class C), having committed very similar offenses. While the system for assigning community service hours is not 1 .rr “.905 Ina-9‘ ‘ f 53”» -“ .1 .~|(¢. o.s. (a II we. I.\ e .1 I . RI... n. a c. A. .. .V)‘ FA I . alv r. vh.u .. a ‘.\ uh 4am .\ A O s I?“ P. Au. A~£ O 5 VI PM L” F“ a“ VI \ pH. 0. «a. ‘.b 0. any C. a“ He V. s .~« .3 ..4 . . . . . . e . 14‘ 3. 9t 4 -. ii a . _ .7 use. 2. «3 p t .5. .0 . n . .F. n u .p u LL. t e as a C an n)... .1 rt .0 .e. .l. n: I p. . . «an o . s . 2‘ as .n . 1. an. nu. U . 65 arbitrary in that sanctions are based on the charged offense, it may not necessarily result in proportional sanctioning for the severity of the offense. Completion dates (the dates by which community service was to be completed by) were usually set shortly before the youth was scheduled to reappear in court. Referral forms were submitted to the Community Service Program shortly after completion and applications were then reviewed by project staff to ensure program eligibility. Processing Youths referred to the program were contacted by a letter from the Community Service Program, addressed to them and their parents (see Appendix D). This letter informed the youth and the parent that the youth had been referred to the program and that at least one parent and the youth must attend an initial interview at the court. The date and time of this interview was specified in the letter and they were told to contact the court if the interview was at an inconvenient time. At the time of this interview, the research project was explained to the youth and the parent by the interviewer. Families were told that the project was designed to examine the effectiveness of various sanctions available to the court; the disposition would be determined by random assignment; participation would involve the youth completing an interview immediately and again a short time later; the 66 youth would receive coupons from local merchants, at posttest, for their time in completing these interviews; if the youth or the parent refused participation, the youth would be directed to perform the community service hours as ordered; and that participation was completely voluntary. At that time, the youth and the parent were asked if they agreed to participate. If either party refused to participate, the youth was excluded from the research project. If both parties agreed to participate, they were asked to sign a consent form (See Appendix E). This form again described the project, outlined the consequences of non-participation, explained that participation is voluntary, and specified that the youth and parent agreed to these terms. These forms and the consent procedure were approved by the University Committee on Research Involving Human Subjects (UCRHIS) (See Appendix F). After the consent form was signed, the youth was interviewed alone for 30 minutes. At the completion of the pretest interview, the parent and youth were reunited and the youth was randomly assigned to one of two treatment conditions (community service or traditional service). The treatment condition to which the youth was assigned was determined by lottery and was stratified by referral status (i.e., intake or formal), age (i.e., 16 and over or under 16), and sex. Slips of paper were prepared with one of the treatment conditions written on it. These were then 67 randomly placed in envelopes prepared by the researcher. There were eight sets of envelopes, one for each combination of the stratification variables. Each set contained equal numbers of envelopes with each treatment option. The envelopes were numbered and labeled on the outside with the stratification combination (e.g., intake, male, under 16). The interviewer did not know the contents of any envelope until it was opened in the presence of the youth and his/her parent. Envelopes were initially prepared in batches of six for each stratification set. As envelopes were used, they were replaced with new envelopes prepared, in pairs, in a similar manner. A minimum of four envelopes were presented and the youth or the parent drew an envelope determining the treatment condition. This reassured both the youth and the parent that the process was random. Intervention Alternatives After pretesting, the caseworker was informed of the treatment condition. Subsequent procedures varied according to the condition to which the youth was assigned. Below is an explanation of the intervention activities associated with each treatment alternative. It should be remembered that this study was designed to test the marginal effects of community service. The only difference between the experimentals and the controls was that controls were not h . ' a!" F .91“‘ AA"'~" age‘dfi ‘- A'QQ' .uA-lh-Ir ;" ....3 file 11 eul- '. h A A H‘fl :nte u y ‘ e. K art: 68 required to fulfill the community service component of their disposition. Issditionsl Services Youths assigned to traditional services were handled according to the original disposition, excluding the community service obligation. For youths referred from the Intake Department, this often involved dismissal or a warning. For formally processed cases, traditional services most often involved probation and referral back to the caseworker. stmunity Service Youths assigned to community service were referred to the Community Service Program. The program was established in 1984 with the goals of increasing accountability and making compensation for property loss or damage caused by minors. The initial contact for youths was an immediate intake interview, designed to discuss work placement and program responsibilities. First, the youth was asked about his or her knowledge of the purpose of the Community Service Program. To the extent that the youth was unable to articulate this purpose, it was explained to him or her. The goal of accountability was stressed (Harding, 1982; Schneider, 1986). The offense was outlined, the number of community service hours to be completed was discussed in relation to the offense, in order to increase the youths’ ' v... a ‘< ‘a 3v- ‘1‘ L . yd: 69 perceptions of the fairness of the sentence, and therefore accountability (Schneider, 1985), and the youth was told that there are costs to the community even when property is recovered or the damage is of a small amount (Heide, 1983). The youth was told that the community would be repaid through the completion of community service activities (Armstrong, 1983). Youths who were referred from intake were also told that the offense would not go on their official record if they successfully completed their community service order. To locate an appropriate placement, the youth was asked about responsibilities and obligations which would restrict his or her availability for such work, including school, after-school activities, jobs, and family responsibilities. Commitment to complete community service would result in less available free time for youths to spend with their friends, and therefore reduce delinquent bonding. The youth was also asked about available transportation for getting to and from work sites. If the public transportation system was the youth’s only means of transportation, bus tokens were available to those who could not afford the costs. It was stressed to youths that it was their responsibility to get to the work site, not that of their parents or friends. The youths also completed a form which allowed them to record their interests and talents (See Appendix G) so that their placement could be most meaningful and commitment 70 would be increased (Keve, 1978). Human service placement requests were carefully screened to exclude those youths with violent histories. While the interests of the youths were used to facilitate identification of an appropriate placement setting, geography, transportation, and hours of availability often controlled the selection process. There were over 60 cooperating agencies where youths may have been placed (See Appendix H). These agencies have been selected because they offer the youths an opportunity to experience positive social influences and thus increase conventional bonding. Job tasks typically include light maintenance; outdoor maintenance; clerical duties; parks conservation and clean-up; housekeeping; pre-school assistance; nursing home and hospital care; and assistance to museums, libraries, churches, and food and clothing banks. Only one youth was placed at a work site at a time unless there were separate areas in which youths may have worked. An exception to this was the work crew program, operating mainly in the spring and summer months, when groups of approximately eight youths were driven from the court (by court staff) to a local park where they performed eight hours of community service and were returned to the court. These types of placements were often used with more violent offenders and difficult to place youths. U" .0- -A..anl . - \— 4" 71 The youth was told that a placement interview would be scheduled. If possible, a potential work site was contacted and an appointment was scheduled immediately. If this was not possible, the youth was told that he or she would receive a phone call or letter about the placement interview. Youths were told they would sign a contract outlining their responsibilities and that they must present a work permit to be signed at that meeting. Youths were also told that they would need to bring to that meeting a letter of apology to their victim. This letter was to explain to the victim that the youth would be completing community service to repay the community. The purpose of this letter was to reinforce to youths that they must take responsibility for their actions. They were also told that they would have to tell the work site supervisor what their offense was, because the supervisor had the right to know. Supervisors were also often reassured by this information, particularly when the crime was of a non-violent nature. This was another way that youths were required to take responsibility for their actions. Youths were told when and where the interview would occur and that they must arrange appropriate transportation to the interview. The youth received the "Placement Interview Instructions" which described the information to include in the letter of apology; the date, time and tn 1n !-l w (4! 72 location of the placement interview, and a reminder to bring a work permit from school (See Appendix I). At the placement interview, the community service advisor, the youth and the work site supervisor discussed the work responsibilities and the dates and times the youth was to work. A starting date and projected ending date were determined. The youth and work site supervisor were reminded of the number of community service hours to be completed. The youth was also told that time spent at the work site which was not productive time would not be counted as hours toward completion. The youth and the work site supervisor were invited to ask questions to help clarify this agreement. Once an agreement had been reached, a contract outlining the number of hours to be completed, the date they must have been completed by, the dates and times the youth was to be at the work site, the name and phone number of the work site supervisor, a list of expectations (e.g., promptness, appropriate dress, etc.), and an explanation of the consequences for failure to meet this obligation (e.g., additional community service hours, in-home detention, juvenile home detention, etc.) was signed by the youth, the work site supervisor, and the community service advisor. The contract also stated that the court and the work site supervisor were to be notified in case of absence. In this way, there were clear goals and youths could see that the 73 assignment was achievable. The youth, the work site supervisor, and the community service advisor each received a copy of the contract (See Appendix J). The work permit application was completed and signed by the work site supervisor and returned to the youth to be exchanged for a work permit prior to the first day of work. The youth also explained to the supervisor what the offense was. Work site supervisors were contacted by telephone after the youths’ first few days of work to ensure that the youth had attended and to remind the supervisor to contact the office in case of any problems or concerns. Weekly attempts to contact work site supervisors were made to monitor the progress of each youth. All contacts with the youth or the supervisor were recorded in the youth’s file on the contact sheet (see Appendix K, a sample is also attached). Youths were not contacted during placement unless they failed to fulfill the obligations of the agreement. Youths who failed to perform their duties according to their contract were contacted by the Community Service Program to determine the reasons behind the problems. Youths were reminded of their commitment and changes in the work site contract (e.g., times to report, work site, etc.) were made if necessary. If a youth continued to refuse participation, he or she was referred to the caseworker who determine the appropriate action. Sanctions usually included a 74 progressively intrusive set of responses, from increased community service hours to in-home detention to juvenile home detention. Specific sanctions were determined on an individual basis for each youth. When the specified number of community service hours were completed, the work site supervisor returned to the Community Service Program a completed schedule which included the number of hours the youth worked and the supervisor’s comments about the quality of the work (See Appendix L). On occasion, however, work site supervisors notified the Community Service Program of completions without returning these sheets. At completion, youths were sent letters of completion from the Community Service Program (See Appendix M). A certificate of special recognition was offered to six youths who exceed their responsibility in performing their community service duties (See Appendix N). This way, youths would experience as sense of accomplishment and perhaps increase their conventional bonding. In addition to completing the community service requirement, youths in this group were also required to complete the other orders in their disposition. Post-test Assessment At the time of random assignment, youths in the traditional services condition were yoked with experimental youths to determine the pre-post interval. Each youth in rhi- A 9 s v ...-. a- N 5...!— ant-p ui 1.. o. .L. ... FF. 9. .4. “a -. .~ .5. p.“ L. v. .. ... .. .sa .. “I :o ..- no. N. :a . 2.. s . V 1 2‘ .1. 'Q .4‘ As.‘ v 1 ‘ - 75 the control condition was paired with an experimental youth whose intake date was as close as possible to his or her own. On occasion it was possible to match a control with one of several experimentals with the same pretest date. In these cases, the other criteria used for matching, in order, were number of community service hours ordered, status, sex, and then age. The date of the posttest for the control youth was then determined by the date of program completion for the matched experimental youth. Youths who successfully completed community service were contacted for a posttest appointment at the time that the completion letter was mailed. Youths who failed to complete the community service agreement and were referred back to a caseworker were contacted for a posttest appointment at the time of that referral. Controls were contacted for posttest at approximately the same time. The posttest interviews were scheduled to occur at the court and lasted 30 minutes. Occasionally, interviews were conducted at the youth’s home, when requested. At completion of the interview, youths received their coupons as compensation for their time and to reduce subject mortality. Most youths were offered a free movie pass and a coupon for a free beverage from a convenience store. The posttest appointment letters did not specify the nature of these coupons, however, after a youth missed one posttest appointment, a note was added to subsequent appointment 76 letters specifying the specific types of coupons he or she would receive. Court records were reviewed by project staff blind to the treatment condition one month after posttest to determine in-program recidivism. A random sample of 10% of these records were coded a second time, one month later, for test retest reliability. Interview Training Interviews were conducted by undergraduate students recruited from the psychology and criminal justice departments. Students were told that they were expected to make a three-term commitment to the project. The first term involved 4-5 weeks of training, and the remainder of their time was spent interviewing youths. Four students were selected as interviewers and spent an average of 4 hours a week conducting interviews. The initial training of the interviewers involved role— plays and tape-recorded practice interviews, as well as practice scoring of the interviews. Inter-rater reliability of the interviewers reached 80% before interviews were conducted. Measures The Community Service Program was built on the model of accountability. It was proposed that participation in community service activities would strengthen conventional bonds and reduce delinquent bonds and further reduce the '1 ..u an. .da -.~pa ~~.4. -1‘ uk i. .u wan Cu as ubk “a. a: .4 1 . P. am a: I” .1 .u a. .4 n4. 5. .3 . . s... on 77 likelihood of further delinquent activity. Measures were designed to assess both the integrity of the accountability model of community service and the impact that it had on these other variables. Figure 7 lists each of the measures used at pretest and posttest, according to treatment condition. l_plementation of Community Service A number of measures were used to assess the integrity of the community service intervention. This included observation of a 10% random sample of the intake (see Appendix 0) and placement interviews (see Appendix P) to ensure that all steps were discussed. In addition, a sample intake interview was transcribed to further document this process (see Appendix Q). Extensive records were also maintained to monitor the progress of the youths (see Appendix R). Included were the number of community service hours ordered, program completion status (successful or unsuccessful), the type of placement, and the amount of time spent completing the order. Intervention Process Court records were used to describe the intervention alternatives. First, the researcher maintained a record of the dates of all pretest and posttest appointments for each youth so as to count the number of appointments that were missed. The dates of the first pre and post-test II.~ C~ ct LD~ 78 Experimental Control Measures Pretest Posttest Pretest Posttest 1. Official delinquency (# of petitions, seriousness of offense) X X X X 2. Referral (hours and other services) X X X X 3. Pretest Interview X X 4. Posttest Interview X X 5. Specific Community Service Outcomes (tasks, scales) X 6. Community Service Process (placement information, completion status) X 7. Process data (# appointments missed, pre-post interval) X X NOTE: "X" the presence of a measure for a given condition and testtime. Figure 7: Measures Used at Each Time Period, By Treatment Condition C I. .. . 9.. C .. G. a. C V . . 4 . . a . E P. .- . T S . A .l u a. V. a” . . .2 .4 . — .. a. . . . . u. C» . A l . t . 3.. 9. v“. .3 .. . '7 a. .4 .e fi. 4. y»; .3 .b.. PU. .p.. Q. CL PM 3‘ s. r. .J a... 3.. .u z «— 14¢ .3 l.. . .n .v‘ t . )Ha “I . ¢ 1 ~ n\. .u; . . -~ n. and 79 appointments as well as the dates these tests were actually completed, were also recorded as part of the project monitoring forms (see Appendix R). In addition, court records were used to determine the current sanctions imposed on each participant so as to describe the services received by the control group and the additional services received by the experimental group (see Appendix 8). The Youth Interview The youth interview was divided into 10 main sections (see Appendix T). Each section was identified by a letter which proceeded the question number. Section A therefore included questions A1 through A4, section B contained Bl through 813, and so on. There was also an opening section designed to gain background information on the youth, as well as to help the youth get comfortable with the interviewer and to save the more difficult questions for later. §g§lg,Develonment While the interview consisted mostly of measures previously developed, this study included independent scale construction procedures to test the integrity of these scales. The strategy used in this study is similar to the rational-empirical approach discussed by Jackson (1970). Each of the "original" scales served as the starting point, or rational basis, for scale development. These rational and A... .1“ cu uad .1 .- A.‘ v. ‘9‘ -\ ‘5‘ 80 scales were then modified, if necessary, to create "revised" scales with maximal internal consistency and independence. First, individual items were examined for variance. Any item for which 90% of the valid scores fell in two adjacent categories (or in one category in the case of dichotomous items) was discarded. The next step involved the assessment of the internal consistency of the scales, using the Reliability program of SPSSX (SPSS Inc., 1986) which provides corrected item-total correlations and computes Chronbach’s alpha as the measure of internal consistency. When only two-items are in scale, Guttman split—half is used as the measure of internal consistency and when the data in the scale is dichotomous, Kuder- Richardson-20 (KR-20) is used. Items which had corrected item-total correlations that were not statistically significant were removed from the scale. The "Reliability" procedure was repeated. Additional adjustments were made using these same procedures until all scales were internally consistent. The corrected item-total correlations were compared to the items’ correlations with the other scales. Those items which had higher correlations with three or more other scales than with their own scale were discarded. The reliability of the scales was recalculated and the relationships of items to other scales were then reexamined. This procedure was repeated until no further adjustments 5.. . v and Fr ~-.. 3 ‘t- 1.. .: o: s 1. Q. IA .so 81 were required. For all scales, a high score represents a better score. Items which were reverse scored are specified in each of the tables demonstrating the reliability of the scales. Measures of Intervention Integrity One of the goals of the study was to learn how the youths perceived various criminal justice sanctions. As mentioned in the earlier review, it is more important to understand how the youth perceives his or her treatment, than it is to rely on professionals’ assumptions about these various sanctions. The interviews of the youths therefore included questions pertaining to their perceptions of the purpose of various sanctions, their feelings about the sentence they received, and feelings about their victim. Section A was designed specifically for this study and asked youths about the main purpose of four different criminal sanctions. This was used to answer research question 3. No attempt was made to scale these items as they were most valuable as individual items. Section B was based on the Juvenile Offender Instrument (JOI), developed for the national evaluation of juvenile restitution programs (Wilson, 1983), and asked youths how they felt about the disposition they received, the components of the sentence not pertaining to community service, and the community service components of the disposition. Each of these three questions presented the 82 same series of nine semantic differential scales for the youths to rate their feelings. However, only those items relating to the community service disposition were retained. The general disposition and other services received questions were dependent on the presence of other sanctions, which not all youths had. In other words, for youths not ordered to other sanctions, the other services questions were never asked and the general disposition questions were identical to the community service questions. On the other hand, for youths who were ordered to other sanctions, the disposition questions reflected a composite of all sanctions. Because the primary interest was the attitudes toward community service, only those items were retained. Results from this scale was used to answer research question 4. These items were scored from 1 to 7 as indicated on the card. By discarding two items from the scale, the coefficient alpha was increased to .86 and the remaining items were not strongly related to other scales. In addition, most item-total correlations increased slightly (see Table 4). At posttest, experimental youths were also asked about the tasks that they performed as part of their community service. They were asked about their perceptions of these tasks and their experience to further assess the extent to which the intervention met its goals of being tangible, measurable, meaningful, and rewarding (Keve, 1978). Three v a. p\._ 83 Table 4 Internal Consistency of Attitudes Toward Community Service Scale Item—total Correlations Community Services Items Scale A - fair/unfair .83 B - helpful/harmful .80 C - wrong/right .41 E - pleasant/painful .59 F — exciting/dull .51 H - interesting/boring .64 I - useful/worthless .71 Coefficient Alpha .86 NOTE: Items A, B, E, F, H, and I were reverse scored. NOTE: n = 82. np'. 41". mar b‘c'. 0-5 ‘Sa ~\~ .\~ 0_‘ b“ a-u CV 84 scales emerged from these items. The Understanding Scale is based on the youth’s knowledge of the community service requirements and the perception of his her ability to complete these requirements (88 and B10). This Guttman split-half coefficient was .92 and the item-total correlations of .87. The Effort Scale rated the amount of effort required by the youth (B5, B6, B7, B11). The coefficient alpha forthe scale was .72 and the item— totalcorrelations ranged from .39 to .66. The Benefit Scale assessed the extent to which the youth learned something or changed as a result of the community service (B12 and B13). The coefficient alpha for the scale was .50 and the item- total correlations were .33 (see Table 5). Because these three scales each assessed the community service experience, it was expected that there might be some relationship between these scales. However, interscale correlations presented in Table 6 show that only the Understanding and Effort Scales are related. The three scales were therefore retained. Section C, used to address research question 5, was also based on the JOI and used 14 semantic differential items to determine the youth’s perception of the victim. These items were also scored from 1 to 7. The item-total correlations for the Victim Scale ranged from .43 to .68 and the scale had a coefficient alpha of .89. No adjustments 85 Table 5 Internal Consistency of Community Service Outcome Scales Scale Item-total Reliability Items Correlations Coefficient' UNDERSTANDING .92 88 - Know what was needed to complete community service .87 B10 - Feel you could complete community service requirements .87 EFFORT .72 85 - How much effort was community service .39 86 — How much time did you give up .59 B7 - How inconvenient was it .47 811 — How easy was it to complete .66 BENEFIT .50 812 - Did you learn a skill .33 813 - Did community service make you a different person .33 NOTE: Item 811 was reverse scored. NOTE: n = 37. ' For the Understanding and Benefit Scales, Split-Half coefficient was used. For the Effort Scale, Coefficient Alpha was used. the Guttman Table 6 86 Community Service Outcome Scale Intercorrelations UNDERSTANDING EFFORT BENEFIT UNDERSTANDING 1.00 EFFORT -.28x 1.00 BENEFIT .08 .15 1.00 . g < .05. NOTE: n = 37. u. v14 87 were necessary and the scale was retained in its original form (see Table 7). Bachman, Kahn, Mednick, Davidson, and Johnston (1967) developed measures of reciprocity and independence which were included as measures of responsibility and accountability to answer research questions 5 and 6. Section D contained five items of the Reciprocity Scale (2,4,6-8,10,12) and six items of the Independence Scale (1,3,5,9,11). These items were scored from 1 (very bad) to 6 (very good). A number of items were discarded from both scales because of a lack of item variance and poor reliability. The revised reciprocity scale contained four items and the independence scale three items. Because the revised scales were significantly correlated (m = .39, E < .001), these items were combined into the Accountability Scale. The reliability of this scale exceeds that of the two subscales (see Table 8). Self-esteem The Rosenberg Self-esteem scale (Rosenberg, 1965), found in Section E, was used to answer research question 10. This was the ten-item measure of self-esteem which was used by Elliott, Ageton, Hunter, and Knowles (1975) with a coefficient alpha equal to .75. Items were scored from 1 to 4 and totalled as suggested by Rosenberg. The reliability of the original scale was adequate but a number of the items A“ . —— (4) 88 Table 7 Internal Consistency of Attitudes Toward Victim Scale Item-total Items Correlations Alpha VICTIM .89 03A - troublesome/cooperative .54 C38 - good/bad .61 C3C - breaks rules/obeys rules .57 C3D - rude/polite .67 C3E - helpful to others/harmful to others .55 C3F - cowardly/brave .45 C30 - dumb/smart .56 C3H - honest/dishonest .68 C31 - lazy/hardworking .56 C3J - tough/weak .47 C3K — not wild/wild .43 C3L - mean/nice .60 C3M - kind/cruel .67 C3N - enemy/friend .61 NOTE: Items C38, C3E, C3H, C3J, C3K, and C3M were reverse scored. NOTE: n = 91. 89 Table 8 Internal Consistency of Accountability Scale Scale Item-total Items Correlations Alpha ACCOUNTABILITY .65 D5 - Being outspoken and frank in expression one’s likes and dislikes .41 D6 - Sticking up for someone who once stuck up for you .37 D7 — Going out of your way to pay people back for being kind .32 D8 - People paying their debts no matter what .30 D9 - Thinking and acting freely, without social restraints, and encouraging others to do likewise .46 D10 - People returning favors you have done them .37 D11 - Being independent, original, non- conformist, different from other other people .38 NOTE: All items were reverse scored. NOTE: n = 94. 90 had poor variance and were discarded. The revised scale contained 6 items (see Table 9). Delinquent Associations Section F assessed differential associations as measured by Elliott and Voss (1974) who reported an alpha coefficient of .84. These items were used to answer research question 8 and were scored from 1 (none) to 4 (most). In this study, the coefficient alpha of the Delinquent Associations Scale was only .66. No adjustments to this scale were made (see Table 10). Youths were also asked at pretest about specific instructions they may have been given by the judge or referee about restricting their contact with certain friends. Additional questions were developed for this study to examine the extent to which youths friends may be different between pretest and posttest. These were retained as individual items. Normative pressure (Section G) was assessed by eight items developed by Elliott and associates (1975) and used by Simons and associates (1980). These items, which were used to answer research question 9, were reported to have an internal consistency coefficient of .64 (Elliott et al., 1975) and .74 (Simons et al., 1980). These items were scored 1 (no) and 3 (yes). Item 62 was discarded because it had zero variance and item 08 was also discarded. The 91 Table 9 Internal Consistency of the Self-Esteem Scale Scale Item-total Items Correlations KR-20 SELF-ESTEEM .54 Scale 2 .35 E4 - I am able to do things as well as most other people E5 - I feel I do not have much to be proud of Scale 4 .21 E7 - On the whole, I am satisfied with myself Scale 5 .47 E8 - I wish I could have more respect for myself Scale 6 .32 E9 - I certainly feel useless at times E10 - At times I think I am no good at all NOTE: Items ES, E8, E9, and E10 were reverse scored. NOTE: n = 94. 92 Table 10 Internal Consistency of Delinquent Bonding Scales Scale Item-total Reliability Items Correlations Coefficienta DELINQUENT ASSOCIATIONS .66 F1 - Current friends ever in trouble with the law .47 F2 — Friends known the longest ever in trouble with the law .44 F3 - Best friends in trouble with the law while best friends .53 NORMATIVE PRESSURE .74 G1 - Friends would think less of a person if he/she were to get in trouble with the law .63 G3 - Friends feel that laws are good and should be obeyed .59 G4 — Friends get into trouble .31 G5 - Kids that get into trouble a lot feel uncomfortable with my friends .62 G6 - I choose friends that are not afraid to have fun, even if it means breaking the law .24 G7 - Kids who get into trouble are put down in my group .51 NOTE: Items G4 and G6 were reverse scored. NOTE: n = 94. ' For the Delinquent Associations Scale, Coefficient Alpha was used. KR-20 was the coefficient for the Normative Pressures Scale. 93 item-total correlations and the KR-20 coefficient alpha increased (see Table 10). These items were asked a second time to ascertain the extent to which the youth agreed or disagreed with each statement as a measure of delinquent bonding as well as the extent to which the youth fit in with his or her peer group. The Normative Values Scale retained only three of the items (G9, G13, G15) that were in the Normative Pressure Scale (see Table 11). Conventional Bonding The final two sections of the interview assessed conventional bonding. Section H contained 22 items measuring normlessness (Elliott et al., 1975; Simons et al., 1980) with a previously reported coefficient alpha of .70. These items were scored from 1 (strongly disagree) to 4 (strongly agree). Two items were discarded because of poor variance, the remainder because of low item-total correlations or high correlations with other scales. The revised scale, General Bonding, contained 10 items with a coefficient alpha of .77 (see Table 12). Section I contained 13 items from the Family (3, 5, 10, 13), Peer (1, 6, 8, 12) and School (2, 4, 7, 9, 11) Normlessness Scales (Elliott et al., 1985), with reliability coefficients reported between .60 to .69. These items were scored from 1 (strongly agree) to 5 (strongly disagree). In this study, the coefficient alphas of the original scales 94 Table 11 Internal Consistency of the Normative Values Scale Scale Item-total Items Correlations KR-20 NORMATIVE VALUES .59 G9 - I would think less of a person if he/she were to get into trouble with the law .37 G13 - Kids that get into trouble a lot feel uncomfortable with me .37 G15 - Kids who get into trouble are put down by me .51 NOTE: n : 59. 95 Table 12 Internal Consistency of Conventional Bonding Scales Scale Item-total Items Correlations Alpha GENERAL BONDING .77 HI - Sometimes necessary to lie on job application to get job you want .38 HQ — I often feel lonely .44 Hll - It’s easier for other people to decide what is right than it is for me .32 812 - The chances for me and my friends making it in life are getting worse not better .42 813 - My friends don’t like me as much as they did in the past .34 814 - I often feel awkward and out of place .50 H15 — It’s not worth planning for the future because I don’t know what’s going to happen these days .55 816 - I sometimes feel like nobody cares about me anymore .54 HI? - I often feel like it’s not worth trying to change things in my life .52 821 - Everything changes so quickly I often have trouble deciding which are the right rules to follow .31 (table continues) NOTE: All items in the General Bonding Scale were reverse scored. NOTE: n = 93 for the General Bonding Scale. 96 Table 12 (continued) Internal Consistency of Conventional Bonding Scales Scale Item-total Items Correlations Alpha SPECIFIC BONDING .80 11 — Making a good impression is more important than telling the truth to friends .42 12 - To stay out of trouble, it’s sometimes necessary to lie to teachers .41 I3 - Making a good impression is more important than telling the truth to parents .48 16 - You have to be willing to break some rules if you want to be popular with your friends .46 18 - In order to gain the respect of your friends, it’s sometimes necessary to beat up on other people .44 19 — At school, it’s sometimes necessary to play dirty in order to win .40 110 - Sometimes it’s necessary to lie to your parents in order to keep , their trust .47 111 - Making a good impression is more important than telling the truth to teachers .66 112 — It’s okay to lie if it keeps your friends out of trouble .57 113 - It may be necessary to break some of your parents’ rules in order to keep their trust .50 NOTE: = 94 for the Specific Bonding Scale. 97 ranging from .56 to .64. However, Table 13 shows that the revised scales were highly correlated. The items in these three scales were therefore combined into a Specific Bonding Scale with a coefficient alpha of .80 (see Table 12). Social Desirability Finally, a portion of the Marlowe-Crowne Social Desirability Scale (Crowne & Marlowe, 1960) was included to determine the extent to which youths were providing honest, rather than socially desirable responses. While some view social desirability as a personality construct, it was used in this study as a measure of response set only. First, two items were excluded because they involved voting and driving activities which many of the subjects had never done. The remaining items were used as the item pool. Every other item scored "true" and every other item scored "false" were included, for a total of 15 items. By discarding a number of items, the reliability of the Honesty Scale was increased (see Table 14). Interscale correlations. Measurement error in dependent variables creates a systematic bias that produces lower correlations between variables than would be found if measurement were perfect. Correction for attenuation (unreliability) of the dependent variables produces a corrected correlation that would be found if measurement were less unreliable. This correction involves dividing the observed correlation of two dependent variables by the Table 13 98 Specific Normlessness (Revised) Scale Intercorrelations FAMILY PEER SCHOOL FAMILY 1.00 PEER .66‘ 1.00 SCHOOL .54‘ .57‘ 1.00 * E < .001. NOTE: n = 94. 99 Table 14 Internal Consistency of the Honesty Scale Scale Item-total Items Correlations KR—20 HONESTY .60 J1 - I have never intensely disliked anyone .25 J7 - I am always willing to admit it when I make a mistake .32 J9 - I sometimes try to get even, rather than forgive and forget .31 J10 - I am always courteous, even to people who are disagreeable .36 J11 - There have been occasions when I felt like smashing things .27 J12 - 1 never resent being asked to return a favor .23 J13 - I have almost never felt the urge to tell someone off .24 J14 — I am sometimes irritated by people who ask favors of me .33 J15 - I have never deliberately said something that hurt someone’s feelings .25 NOTE: Items J9, J11, and J14 were reverse scored. NOTE: n = 94. 100 product of the square roots of the reliability of the two variables. The corrected interscale correlations are presented in Table 15. While the two scale of delinquent bonding (Delinquent Associations and Normative Pressure) and the two scales of conventional bonding (General Bonding and Specific Bonding) were highly correlated, they were retained as separate scales to have multiple measures of these very important constructs. Further, it should be noted that all scales were negatively correlated with the Honesty Scale (7 of 9 were statistically significant) suggesting that youths who gave more socially desirable answers on the Honesty Scale (low scores) also gave more socially desirable responses on the other scales (high scores). Recidivism Court records were also examined to assess past offenses and recidivism rates (See Appendix R). This data was collected by staff blind to the treatment condition of the youths. Recidivism included all contacts with the court after referral to the program, excluding those instances where the charges were dismissed. The dates of the petition, the offense, and the sanctions ordered were coded for each contact. Subsequent offenses were divided into two types. First, petitions which were filed during the period the youth was performing his or her community service (or would have been if in the control group) were considered. This included all petitions filed between the date of the 101 Table»15 Scale Intercorrelations Scales 1 2 3 4 5 6 7 8 9 1 Community Service 2 Attitudes to Victim. .15 3 Account- ability -.12 .20 4 Self- esteem -.13 —.13 .17 5 Delinquent Associa- tions .12 .08 .20 .10 6 Normative Pressure -.19 .07 .20 .00 .41** 7 Normative 8 General Bonding .22 -.06 .16 .60** .29** .29** .15 9 Specific Bonding .37* .15 .17 .243 .44** .60** .39** .48** 10 Honesty -.10 -.16 -.24* -.25* -.44** -.50** -.49** -.35** -.43** * p < .05. *1 p < .01. NOTE: .All correlations have been corrected for attenuation. Number of subjects for correlations ranged from 82 to 94. 102 youth’s referral to the community service program and the date of completion. A second set of petitions considered were those filed between the date of community service completion and the date the posttest was completed. These were then added to the first set of petitions for a second method of counting offenses. There were no differences between the findings using the first and second method of counting recidivism. Only the results from this second set are presented. Because the data was collected for a varied time period, the subjects were divided into quarters by the date of the pretest. The time lag from pre-to posttest was longer for the first and third quarters (138 and 140 days, respectively) than for the second and fourth quarters (105 and 107 days, respectively), however this difference was not statistically significant. Further, the time lag for the first and second quarters combined was not different than that for the third and fourth quarters combined. Data Coding Interviews were coded as they were conducted. Two interviewers coded a random sample of 5% of the interviews to assess inter—rater reliability. The pretest version of the interview contained 161 items. Of these items, there were between one and three discrepancies between the raters. Therefore, at pretest, there was total agreement between two raters on 98% to 99% of the items. The posttest version of 103 the interview contained 171 items and there was disagreement on zero to two of the items. At posttest, therefore, there was total agreement between two raters of 98% to 100%. Test-retest reliability was used with the court record data collection. A random sample of 10% of the records were recoded two to four weeks later. For six of the eight records that were used in this check, there were no differences between any of the data on the original form and the reliability forms. For one case, the date of an offense was recorded differently the second time, using the data the petition was received, rather than the date it was filed. This did not affect categorization of the petition as a prior petition. In one other case, a second charge which was dismissed was recorded on the sheet during the original data collection phase, but was omitted during the reliability coding. Because of the large number of variables being coded on this form (ranging from 10 to 24 for these eight cases, depending on the number of prior and subsequent offenses), these differences suggest a small amount of error in the coding of court records (only 3 of a possible 107 for these eight cases). In other words, for six cases, there was 100% agreement, for one case there was 96% agreement, and for one case their was 90% agreement. RESULTS Intervention Description Intervention Process Youths who participated in this study received one of two types of treatment: traditional services including community service or traditional services without community service. In this section, the first research question "What was the intervention process for each of the two conditions?" will be answered by detailing the actual treatment provided to the youths. This description will focus on the types of services provided as well as the length of the intervention. The two groups were compared on the similarity of the original court order, determined by the number of community service hours ordered, and the other services or restrictions which were ordered. Community Service Hours Ordered The number of community service hours ordered by the judge or referee prior to program referral are presented in Table 16. While the number of community service hours ordered for experimentals ranged from 10 to 115, the majority of youths in the experimental group (61%) were required to complete 30 or fewer hours of community service, with a mean 34.89 hours. The distribution of community service hours ordered for controls ranges from 15 to 100 hours, however, the majority (53%) of youths in the control groups were originally ordered to complete 30 or fewer hours 104 105 Table 16 Distribution of Community Service Hours Ordered Experimentals Controls Hours Ordered n n 10 1 0 15 7 4 20 6 11 25 8 8 30 5 1 35 1 1 38 0 1 40 7 7 45 0 1 50 2 2 60 2 2 65 0 1 70 2 0 75 1 0 80 0 2 85 1 2 100 0 2 115 1 O MEAN (SD) 34.89(21.85) 38.29 (23.74) F. (1,88) = 0.49 NOTE: This table represents the number of community service hours originally ordered. Controls did not complete these hours. NOTE: Due to missing court records, this data was not available for all subjects. 106 of community service, with a mean 38.29 hours. However, youths in the control group did not have to complete these hours. The number of community service hours originally ordered for experimentals was not significantly different from that ordered for controls (£(1,88) = 0.49). The low number of community service hours ordered for both experimentals and controls reflect the fact that the offenses the youths were charged with were not of a very serious nature. There was a significant positive correlation (m = .44, p < .001) between hours of community service and seriousness of current offense. Additional Court Ordered Requirements The number of court ordered requirements which were imposed is listed in Table 17. The court orders of 16% of the experimentals included no requirements beyond community service. For these youths, therefore, community service was the only service they received. Nearly half (41%) of the youths in the experimental group, however, had one or two services or restrictions which applied throughout the course of this study, in addition to community service. The court orders of 22% of the controls included no requirement other than community service. For these youths, therefore, assignment to the control group meant that they received no services or restrictions. Nearly half (42%) of the youths in the control group, however, had one or two services or restrictions which applied throughout the course of this Table 17 107 Total Number of Requirements Ordereday Not Including the Community Service Order Experimentals Controls Number n (%) n (%) No others 4 (11) 7 (18) One 9 (26) 5 (13) Two 7 (20) 11 (29) Three 5 (14) 4 (11) Four 6 (17) 4 (11) Five 1 (3) 3 (8) Six 0 4 (11) Seven 1 (3) 0 Eight 1 (3) 0 Eleven 1 (3) 0 MEAN (SD) 2.66 (2.36) 2.47 (1.91) E (1,72) : 0.14 NOTE: Due to missing court records, data not available for all subjects. 8 Includes 8 experimentals and 4 controls ordered to write a letter of apology. This was part of community service and control youths did not complete this part of their order. 108 study. The mean number of requirements, not including community service, was 2.57 for experimentals and 2.24 for controls (F (1,88) = 0.52.). Table 18 displays the specific court order requirements for all youths. For those youths in the experimental group, this list details the services or restrictions which were required in addition to the community service. For those youths in the control group, this chart displays the only services or restrictions received. This list does not itemize those requirements which are part of every probation order, as these are common to experimentals and controls. The most frequently listed requirements for both experimentals and controls was probation. Court orders also often included completing psychological evaluations, participating in individual and/or family counseling, attending school regularly, and obeying their parents. Also frequently mentioned was avoiding contact with one or more specific individuals. Missed Interview Appointments Most subjects did not miss any appointment for their pretest (see Table 19) or their posttest interviews (see Table 20). One-quarter of the experimentals (23%) and controls (20%), however, did miss one pretest appointment and one-quarter of the experimentals (25%) and controls (27%) missed one posttest appointment. There were no significant differences between the groups for the number of 109 Table 18 Distribution of Court Order Requirements Exp.Con Requirements n n Probation 25 25 Have no contact with one specific person 3 2 Have no contact with two or more specific persons 1 6 Curfew 3 1 Obtain employment 2 0 Make restitution to the victim 4 3 Submit to urinalysis 4 3 Complete high school or earn GED 1 0 Cooperate with the Volunteer Probation Officer 3 3 Do not physically or verbally abuse "parents" 3 4 Do not operate and automobile 2 4 Complete examination directed by P0 4 2 Complete psychological evaluation as directed 6 4 Individual and/or family counseling with parents 15 11 Parents attend parenting sessions 1 1 Assigned to Intensive Probation Services 4 5 Attend adolescent group of IPS 5 4 Parents participate in IPS parenting groups 2 5 Detention 1 1 In-home Detention 2 0 Attend school regularly 6 5 (table continues) 110 Table 18 (continued) Distribution of Court Order Requirements Exp. Con Requirements n n Obey parents 11 9 Write a theme 0 1 Abide by behavioral contract 0 1 See high school counselor 0 1 Residential Substance Abuse Program 0 1 Return Merchandise 1 0 Attend Alanon 1 0 111 Table 19 Distribution of Missed Pretest Appointments Pretest Experimentals Controls Appointments Missed n (%) n (%) 0 28 (64) 32 (71) 1 10 (23) 9 (20) 2 4 (9) 3 (7) 3 2 (5) 1 (2) MEAN (SD) .55 (.85) .40 (.72) E (1,88) = 0.76 112 Table 20 Distribution of Missed Posttest Appointments Posttest Experimentals Controls Appointments Missed n (%) n (%) 0 20 (53) 23 (56) 1 8 (21) 11 (27) 2 4 (11) 1 (2) 3 4 (11) 2 (5) 4 0 3 (7) 5 2 (5) 0 6 0 1 (2) MEAN (SD) 1.00 (1.40) 0.90 (1.45) .1: (1,78) = 0.09 113 pretest (F (1,88) = 0.76) or posttest (F (1,88) : 0.03) appointments missed. Lag from Time of Offense The amount of time that elapsed from the date of the current offense to the date of the pretest exceeded 100 days (mean = 101.78, sd = 69.27). There was no significant difference between the length of this lag for experimentals and controls (F (1,81) = .07, p = .792). The lagtime from the time of the current offense to the first day of community service work was four months (mean : 128.55 days, sd : 74.51). Pre to Posttest Interval The number of days that youths participated in the study was dependent upon how long it took the youths in the experimental group to complete the community service requirements. The number of days between the pretest and the posttest is listed in Table 21. The average number of days between pretest and posttest was 130.70 (s.d = 66.42) days for experimentals and 122.59 (s.d. = 59.88) days for controls (E (1,80) = 0.33). Group Equivalence To test the equivalence of the experimental and control groups, the two groups were compared on each of the variables which have been used to describe the intervention process. There were no significant differences between experimentals and controls in the number of community 114 Table 21 Distribution of Days Participatingyin Studya Days Experimentals Controls Participating n (%) n (%) 39 to 60 5 (13) 5 (12) 63 to 91 10 (26) 10 (24) 95 to 121 9 (23) 10 (24) 133 to 181 5 (13) 10 (24) 185 to 240 7 (18) 3 (7) 242 to 252 2 (5) 3 (7) MEAN (SD) 125.03 (62.06) 122.59 (59.88) 3(1178) = 0.03 ' Number of days between pretest and posttest. 115 service hours originally ordered, the number of services or restrictions originally ordered, the number of pretest appointments missed, the number of posttest appointments missed, or the number of days from pre to posttest. Smmmary of Intervention Process Youths in the experimental were originally ordered to complete 35 hours of community service and control youths were originally ordered to complete 38 hours of community service. This difference was not statistically different. The number of additional court order requirements were similar for experimentals (2.57) and controls (2.24). The number of pretest and posttest appointments missed were similar for experimentals and controls. Further, the number of days both from offense to pretest and pretest to posttest was not different for experimentals than controls. The control and experimental groups were similar on all measures of the intervention process. Implementation of Community Service The extent to which the community service intervention was implemented as designed (research question #2) was assessed in terms of the nature of the contact with the community service program and the services provided. This information was obtained from court records and observational assessments. First, a random sample of 5% of the intake and placement interviews were observed by the researcher, who 116 indicated which components of the interview were included or excluded. The intake interview checklist contained 11 items which were to be covered. In all cases, no more than one item was ever omitted from the intake interviews. One intake interview did not adequately relate the sanction to the offense, one did not remind the youth to bring the work permit to the placement interview, and one did not remind the youth to bring the letter of apology to the placement interview. The placement interview checklist contained 8 items. Observations of placement interviews found that only collection of the letter of apology was ever omitted from the placement interviews. In all cases, the letter of apology was obtained prior to posttest. A total of 21 different work sites were used by youths in this study. These agencies represented eight different types of placement sites, listed in Table 22. There were 33 youths who completed all of their community service hours in one site and an additional 10 who divided their hours across two different settings. The types of tasks that youths performed are listed in Table 23. The most common tasks were cleaning and/or housekeeping, and maintenance. The number of days from the intake interview to the placement interview ranged from zero to 183 days. One-third (32%) of the youths were placed within one week and half (51%) were placed within 16 days (see Table 24). The average was 29.22 days from intake to placement. Many of 117 Table 22 Distribution of Placements Used by Experimentals who Completed Community Service Only Placement One of two Placements Types of Placement n n Parks work program 5 4 Museum 10 4 YMCA 4 1 Library 5 0 Church 2 1 Service Organization 7 3 Police Department 0 1 School District 0 5 County 0 1 118 Table 23 Distribution of Type of Community Service Tasks Performed Task n Outdoor park cleanup 5 Yardwork 8 Cleaning and/or housekeeping 23 Maintenance 10 Shelving books 4 Cleaning and answer phones/secretarial 4 Food bank 2 Work with people (e.g. sports) 9 119 Table 24 Distribution of Number of Days Between Intake Interview and Placement Interview Number of Days % 7 days or less 32% 16 days or less 51% 30 days or less 73% NOTE: n = 41. NOTE: Mean number of days between intake interview and placement interview : 31.03 (sd = 39.95). 120 those youths in the upper end of the distribution were intentionally delayed for placement, awaiting the start of a work crew program. Other youths missed earlier placement appointments and therefore delayed their own placement. The intensity of the community service intervention was examined in a number of different ways. First, the number of days that youths worked at their placements is listed in Table 25. The majority of youths (66%) attended five or fewer days, with an average of 6.17 days. For 14 youths, data on the number of days worked were not available because the work site supervisor did not return the schedule sheet which contained this information. However, information on completion status (success or failure) was available for all subjects. The intensity of the community service was also defined by the number of days that elapsed between the first day a youth went to the placement site for work and the last day of work. The number of days that youths were available for work ranged from 2 to 160 days. For one-quarter of the youths (27%), this period was seven days or less and for one-half (49%), it was 14 days or less (see Table 26). The average was 32.58 days at the placement site. Also computed was the number of community service hours divided by the number of days at the work site to get an average hours per day that the youths worked. The number of hours per day worked ranged from 2.73 to 8.50 hours. The 121 Table 25 Distribution of the Number of Days at Placement Number of Days at Placement n 2 days 1 3 days 3 4 days 4 5 days 9 6 days 1 7 days 3 9 days 1 10 days 2 11 days 1 13 days 1 18 days 1 Didn’t finish 4 Missing 14 Mean number of days at placement = 6.30 (sd : 3.55). 122 Table 26 Distribution of Number of Days Between First and Last Day of Placement Number of Days % 7 days or less 27% 14 days or less 49% 31 days or less 67% NOTE: n = 33. NOTE: Mean number of days between first and last day of placement = 32.58 (sd : 36.69). 123 average was 5.36 hours per day at the placement site. Over one-third (35%) of the youths worked an average of 4 hours or less per day, another third (41%) between 5 and 7 hours per day, and one-quarter (24%) more than 7 hours per day. In addition, the number of community service hours ordered was divided by the time interval worked to get the number of hours per day "at risk for working". The number of hours per day that youths were available to work ranged from .16 to 8.00 hours per day. One-quarter (27%) of the youths worked less than one hour a day that they were available, another quarter (27%) worked from one to two hours a day that they were available, another quarter (21%) worked from just over two hours to less than four hours a day available, and another quarter (24%) worked between five and eight hours a day that they were available for work. The average number of hours worked per day that youths were available for work was 2.98 hours per day. There were 47 youths assigned to the experimental group. Of these, 9% did not complete their community service obligation and were returned to their caseworker for additional sanctions. Therefore 91% did complete the community service requirement as ordered. One component of the community service program required that youths write a letter of apology to their victim. If there was no victim (e.g., drug possession), youths were instructed to write a letter describing what they would do, 124 if anything, to avoid repeating the incident. These apologies were received from all of those youths who completed their community service and were posttested. Summary of Implementation of Community Service Observations of the intake and placement interviews showed that these included all of the required elements. Measures of intervention intensity showed that youths spent an average of 6 days at their placement, took one month to complete their community service hours and averaged just over 5 hours per day at the work site. Summaronf Intervention Description The experimental and control groups did not differ on any of the intervention process variables, thus ensuring group equivalence. Further, the community service program was implemented as designed, in that the intake and placement interviews were administered properly and participants completed the necessary community service requirements. Outcome There are three ways that the data from this study could have been examined. The presence of pretest data allows for the use of repeated measures analysis of variance to test for time main and interaction effects. Change scores (the difference between a score at pretest and posttest) could also be examined with analysis of variance. The use of change scores, however, compounds the error which 125 is included in both the pretest and posttest scores. This was therefore not a desirable method. A third alternative was to use one—tailed tests of significance, such as a one— tailed t-test, because the research hypotheses suggested the direction of expected findings. The most conservative and informative test is the repeated measures analysis of variance and was therefore chosen in this study. One-tailed t—tests were also performed, however, these results were not different from the repeated measures analysis of variance and are therefore not reported. When a data set contains a number of variables that measure similar constructs, and these variables are also correlated, multivariate analyses are suggested. Multivariate analysis of variance takes into consideration the correlation between dependent variables and controls for them. It allows one to examine relationships between variables rather than focusing on each in isolation. However, univariate analysis of variance assumes that there is a zero correlation between the variables, or that the relationship between the variables is of no interest (Bray & Maxwell, 1985). Where logical cluster of variables existed, analyses were therefore performed in a multivariate, repeated measures fashion using the MANOVA procedure of SPSSx (SPSS Inc., 1986). Post-hoc Scheffe tests were performed to identify group differences. Univariate repeated measures analyses of variance were 126 performed on the remaining variables. Both treatment condition (experimental and control) and referral status (formal and informal) were used as independent variables in these analyses. Intervention Integrity This section examines the extent to which the goals of accountability were achieved. This was assessed through the youths’ perception of the purpose of various criminal sanctions, attitudes toward community service, attitudes toward the victim, and feelings of accountability. In addition, there was data describing the community service experiences of youths in the experimental group. Purpose of Different Sanctions The third research question asked how community service participants differed from controls in their understanding of the purpose of community service relative to traditional sentencing aims. Youths were asked what they thought the main purpose was of four different criminal sanctions: diversion, probation, community service, and detention. They were allowed to select punishment, deterrence, rehabilitation, revenge, or reparation as their responses. The majority of youths required an explanation for diversion prior to responding. The categorical nature of the response categories for these items required a Chi-square for the analysis of this data. 127 The responses to these questions suggest that youths recognized that these sanctioning options served different purposes. At pretest, the majority of youths indicated that the purpose of diversion was deterrence, the purpose of probation was rehabilitation, and the purpose of detention was punishment. The responses for community service were less clear, with one-third suggesting reparation and another one—third or more divided between punishment and deterrence (see Table 27). However, there were no significant differences between experimentals and controls for any of these sanctions. It was expected that, at posttest, youths in the experimental group would believe the main purpose of community service was to repair the damage. No changes in the other responses were expected. At posttest, the majority of youths indicated that the purpose of diversion was deterrence, the purpose of probation was rehabilitationv or deterrence, and the purpose of detention was punishment. This is fairly similar to the pretest responses. The majority of youths also felt that the purpose of community service was reparation (see Table 28). However, again, there were no significant differences between experimentals and controls. There was a 20% increase from pretest to posttest in the number of youths in the experimental group who believed the purpose of community service was to repair the damage. 128 Table 27 Purpose of Criminal Sanctions - at Pretest Experimentals Controls Cramer’s Sentence n(%) n(%) V Diversion .10 Punishment 4 (9) 3 (6) Deterrence 29 (62) 28 (60) Rehabilitation 9 (19) 8 (17) Revenge 0 0 Repair the damage 5 (11) 8 (9) Probation .18 Punishment 13 (28) 9 (19) Deterrence 11 (23) 11 (23) Rehabilitation 21 (45) 21 (45) Revenge 0 1 (2) Repair the damage 2 (4) 5 (11) Community Service .27 Punishment 11 (23) 10 (21) Deterrence 7 (15) 12 (26) Rehabilitation 6 (13) 10 (21) Revenge 4 (9) 0 Repair the damage 19 (40) 15 (32) Detention .20 Punishment 27 (57) 28 (60) Deterrence 5 (11) 7 (15) Rehabilitation 10 (21) 8 (17) Revenge 0 2 (4) Repair the damage 5 (11) 2 (4) NOTE: Cramer’s V is equivalent to the Phi Coefficient. 129 Table 28 Purpose of Criminal Sanctions - at Posttest Experimentals Controls Cramer’s Sentence n(%) n(%) V Diversion .28 Punishment 2 (5) 4 (10) Deterrence 23 (61) 18 (44) Rehabilitation 7 (18) 15 (37) Revenge 0 1 (2) Repair the damage 6 (16) 3 (7) Probation .32 Punishment 6 (16) 12 (29) Deterrence 17 (45) 11 (27) Rehabilitation 11 (29) 11 (27) Revenge 2 (5) 0 Repair the damage 2 (5) 7 (17) Community Service .25 Punishment 5 (13) 13 (32) Deterrence 4 (10) 5 (12) Rehabilitation 5 (13) 6 (15) Revenge 1 (3) 1 (2) Repair the damage 23 (61) 16 (39) Detention .26 Punishment 21 (55) 17 (42) Deterrence 9 (24) 9 (22) Rehabilitation 5 (13) 10 (24) Revenge 3 (8) 2 (5) Repair the damage 0 3 (7) NOTE: Cramer’s V is equivalent to the Phi Coefficient. 130 The responses were therefore recoded such that the item was scored as either yes or no for reparation. Comparisons between experimentals and controls still failed to yield a significant difference (X2 (1) = 3.65, p < .06). The pretest responses from experimentals were then compared to their posttest responses. Again, there were no significant differences. Attitudes Toward Community Service In order to answer the fourth research question, "How did community service participants differ from controls in their perceptions of the fairness of their community service order?", the fairness of the disposition was rated by the Community Service Scale, the score of which ranged from 0 to 7. This variable was analyzed with a repeated measures analysis of variance using condition and status as independent variables. The analyses (see Table 29) found a significant time (fi(1,57) = 4.98, m < .05) and a significant condition by time interaction effect (F(1,57) = 7.53, p < .01). Post—hoc Scheffe comparisons show that youths in the experimental group showed a more favorable attitude toward the community service component of the disposition at posttest than at pretest (E(1,29) = 11.28, p < .005). Attitudes Toward Victim Attitude toward the victim was rated by the Victim Scale which ranged from 0 to 7 to answer research question 131 Table 29 ANOVA of Attitude Toward Community Service Scale Test time Condition Status (n) Pretest Posttest mean (sd) mean (sd) 1 I 1 Formal (14) I 4.34 (1.26) I 5.25 (1.27) 1 Experimental 1 I 1 I I I Informal (16) I 4.83 (1.01) I 5.40 (1.04) I I I I I 1 I Formal (12) I 4.88 (1.67) I 4.82 (1.30) 1 Control 1 I I 1 1 I Informal (19) I 5.17 (1.03) I 5.08 (1.01) I I I 1 Analysis of Variance Source DF MS F Prob. w2 Condition (C) 1 0.04 0.02 .896 Status (S) 1 2.67 1.25 .268 C X S 1 0.02 0.01 .926 Subjects 57 2.14 Time (T) 1 3.26 4.98 .030 .05 C X T 1 4.93 7.53 .008 .10 S X T 1 0.25 0.39 .536 C X S X T 1 0.17 0.26 .612 Error 57 0.66 132 #5 (How did community service participants differ from controls in how they felt about their victim?). This scale was analyzed with a univariate repeated measures analysis of variance which did not reveal any significant effects (see Appendix U). Accountability The sixth research question asked how community service participants differed from controls in their feelings of accountability. The Accountability Scale ranged from 0 to 6. This scale was analyzed with a univariate repeated measures analysis of variance which showed a significant condition by status interaction effect (E(1,75) = 5.28, E < .05). Post-hoc Scheffe tests show that youths in the control group who were informally processed had a greater sense of accountability (E(1,39) = 5.57, E < .05) than did youths who were processed formally (see Table 30). Response to Community Service Youths in the experimental group described their experience on three different scales. While no statistical comparisons were possible since these questions were asked only at posttest for experimentals, the results are useful in describing the quality of the experience for the youths (see Table 31). The mean score for the Understanding Scale was 3.82 on a 4 point scale suggesting that youths understood and felt able to complete the requirements of 133 Table 30 ANOVA of the Accountability Scale Test time Condition Status (n) Pretest Posttest mean (sd) mean (sd) I I I Formal (20) I 4.84 (.73) I 5.01 (.42) 1 Experimental I I I 1 I I Informal (18) I 4.76 (.53) I 4.87 (.52) I 1 1 I I I 1 Formal (21) I 4.68 (.53) I 4.68 (.62) 1 Control 1 I I I I I Informal (20) I 5.04 (.58) I 5.09 (.49) I I I I Analysis of Variance Source DF MS F Prob. w2 Condition (C) 1 0.00 0.00 .918 Status (S) 1 0.73 1.60 .210 C X S 1 2.42 5.28 .024 .04 Subjects 75 0.46 Time (T) 1 0.26 1.54 .218 C X T 1 0.11 0.65 .423 S X T 1 0.00 0.00 .988 C X S X T 1 0.03 0.20 .652 Error 75 0.17 134 Table 31 Feelings About Community Service Experience Scale Mean SD UNDERSTANDING 3.82 .56 EFFORT 2.15 .81 BENEFIT 2.26 .90 NOTE: Based on 38 subjects in the experimental group. 135 community service. Eighty—seven percent of the youths had a mean score of 4.00 on that scale. The mean score on the Effort Scale was 2.15 on a 4—point scale, suggesting only a little effort was required to complete community service. The average score on the Benefit Scale was 2.26. Half of the youths felt like they received little or no benefit from the program, while half felt like they had benefited somewhat or a great deal. Exploratory Analyses All analyses were repeated, excluding those youths who indicated at the end of their posttest, that the reason they thought they were selected to receive community service was because they drew the wrong envelope. It was thought that these youths would have been less likely to see the relationship between their offense and their punishment, simply because of something in the research process. This did not, however, change the findings. Summary of Intervention Integrity This section was designed to assess the extent to which accountability was achieved. While 20% more of the youths in the experimental group at posttest than at pretest reported the purpose of community service was to repair the damage, this was not significantly different from controls. Experimental youths at posttest, did show a more favorable attitude to community service than at pretest. There were no 136 differences between experimentals and controls on the Accountability Scale. Intermediate Outcomes The scales which assessed delinquent and conventional bonding were analyzed using a multivariate repeated measures analysis of variance. These four scales (Delinquent Associations, Normative Pressure, General Bonding, and Specific Bonding) were not only highly correlated, but also theoretically linked. A univariate approach was used with the Normative Values and Self-esteem Scales. For all comparisons, condition (experimental and control) and referral status (formal and informal) were used as independent variables. Additional analyses examined the relationship between these outcome variables and several process variables. Delinquent and Conventional Bondimg Research question #7 asked, "How did community service participants differ from controls in the types of friends they spent time with?" The Delinquent Association Scale assessed the extent to which the youths friends "were in trouble with the law". The values of the youths friends was rated by the Normative Pressure Scale, which was used to answer research question #8, "How did community service participants differ from controls in the pressure they felt from their friends?". The General Bonding and Specific Bonding scales rated attachment to conventional norms, in 137 order to answer the ninth research question which asked how community service participants differed from controls in their attitudes towards conventional beliefs. A multivariate repeated measures analysis of variance did not yield a significant condition effect, nor condition by status, condition by time, status by time, or condition by status by time interaction effects. However, there was a significant status effect for the Delinquent Associations Scale (F(4,72) = 10.84, p < .05). Youths who were formally processed had stronger bonds with other delinquents than did youths informally processed (see Table 32). Post-hoc Scheffe tests showed that there was also a significant time effect for the General Bonding Scale (F(4,72) = 6.20, m < .05), such that youths expressed greater attachment to conventional norms at posttest than at pretest (see Table 33). There were no significant effects for the Normative Pressure and Specific Bonding Scales (see Appendix V). Normative Values A univariate repeated measures analysis of variance found a significant time effect for the Normative Values Scale (E(1,72) = 7.02, p < .01.). Youths had more attachment to delinquent values and norms at posttest than at pretest (see Table 34). Self-Esteem Research question # 10 was "How did community service participants differ from controls in their self—esteem?" A Table 32 138 ANOVA of the Delinquent Associations Scale Test time Condition Status (n) Pretest Posttest mean (sd) mean (sd) I I 1 Formal (20) I 2.78 (.78) I 2.68 (.77) 1 Experimental I I I I I I Informal (18) I 3.32 (.65) I 3.24 (.52) I 1 I I I I 1 Formal (21) I 3.02 (.71) I 2.76 (.52) 1 Control I I I 1 I I Informal (20) I 3.22 (.58) I 3.10 (.51) I I I I Analysis of Variance Source DF MS F Prob. w2 Condition (C) 4 0.01 0.02 .884 Status (S) 4 6.52 10.84 .002 .001 C X S 4 0.74 1.24 .270 Subjects 72 0.60 Time (T) 4 0.73 3.38 .070 C X T 4 0.95 0.44 .509 S X T 4 0.66 0.30 .583 C X S X T 4 0.03 0.14 .708 Error 75 0.22 Table 33 ANOVA of the General Bonding Scale 139 Test time Condition Status (n) Pretest Posttest mean (sd) mean (sd) I I I Formal (20) 1 2.95 I 3.07 (.46) I Experimental 1 I I I I I Informal (18) I 2.85 I 2.93 (.52) I I I I I I I Formal (21) I 2.72 (.41) I 2.90 (.34) I Control I I I I I I Informal (20) I 3.04 (.40) I 3.03 (.41) I I I 1 Analysis of Variance Source DF MS F Prob. w2 Condition (C) 4 0.27 0.90 .765 Status (S) 4 0.10 0.33 .565 C X S 4 1.13 3.73 .057 Subjects 72 0.30 Time (T) 4 0.33 6.20 .015 .07 C X T 4 0.00 0.08 .783 S X T 4 0.14 2.62 .110 C X S X T 4 0.50 0.95 .333 Error 72 0.05 Table 34 140 ANOVA of Normative Values Scale Test time Condition Status (n) Pretest Posttest mean (sd) mean (sd) 1 I I Formal (19) I 1.33 (.57) I 1.11 (.33) 1 Experimental I I I I I I Informal (17) I 1.35 (.63) I 1.16 (.50) I I I I I I I Formal (21) I 1.29 (.53) I 1.21 (.45) I Control I I I I 1 I Informal (19) I 1.42 (.71) I 1.21 (.50) I I I 1 Analysis of Variance Source DF MS F Prob. w2 Condition (C) 1 0.07 0.18 .673 Status (S) 1 0.10 0.26 .612 C X S 1 0.01 0.03 .869 Subjects 72 0.40 Time (T) 1 1.20 7.02 .010 .06 C X T 1 0.04 0.25 .620 S X T 1 0.02 0.14 .714 C X S X T 1 0.06 0.37 .547 Error 72 0.17 141 univariate repeated measures analysis of variance found that there was a significant time effect for the Self-Esteem Scale (E(1,75) = 7.30, p < .01). Youths showed a greater sense of self-esteem at posttest than at pretest (see Table 35). ExploratoryiAnalyses First, the relationship between these intermediate outcomes and a number of process variables was examined. The process variables considered were the number of community service hours ordered, the number of prior petitions, the seriousness of the current offense, and the amount of time from pretest to posttest. The resulting correlation matrix is presented in Table 36. It should be noted that only three of the 24 correlations were statistically significant, and all involved the Delinquent Associations Scale. There were significant negative correlations between the Delinquent Association Scale and the number of community service hours, the number of prior petitions, and the seriousness of the offense. Thus, the greater the number of community service hours required, the greater the number of prior petitions, or the more serious the current offense, the more friends the youth had who "were in trouble with the law". It was suspected that the number of community service hours ordered, the number of prior petitions and the seriousness of the offense would be related, and correlations confirmed this. Seriousness of the current offense was significantly related to the number of community 142 Table 35 ANOVA of the Self-Esteem Scale Test time Condition Status (n) Pretest Posttest mean (sd) mean (sd) 1 I 1 Formal (20) I 0.74 (.24) 1 0.73 (.29) 1 Experimental 1 I I I I I Informal (18) I 0.63 (.30) I 0.68 (.22) I I I I I I 1 Formal (21) 1 0.58 (.29) I 0.70 (.26) I Control I I I I I I Informal (20) I 0.69 (.26) I 0.76 (.19) I I I 1 Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.00 0.02 .881 Status (8) 1 0.00 0.00 .973 C X S 1 0.25 2.23 .140 Subjects 75 0.11 Time (T) 1 0.14 7.30 .009 C X T 1 0.06 2.96 .089 S X T 1 0.00 0.07 .785 C X S X T 1 0.03 1.63 .205 Error 75 0.02 143 Table 36 Correlations Between Process and Intermediate Outcome Variables Intermediate Outcomes Process Delin. Norm. Gen. Specific Norm. Self Variables Assoc. Bonding Bond. Bonding Values Esteem Community Service -.42¥*** -.12 .04 -.17 -.12 .17 Hours Number of Prior -.26* -.04 .06 -.02 -.08 .09 Petitions Seriousness of Current -.36*** -.11 .13 .02 .01 .14 Offense Days from Pre- to -.04 .08 -.07 -.05 -.18 .07 Posttest Days from Current Offense to Pretest .37**t* -.26* -.14 -.28** -.04 -.06 t E < .05. *it E < .005. #8:: E < .001. NOTE: Correlations based on 73 to 81 subjects. 144 service hours ordered (m = .44, p< .001) and the number of prior petitions (m = .46, p < .001). The number of prior petitions was not significantly related to the number of community service hours for the current offense (3 = .15). Next, it was clear from the original interscale correlations that there were also significant negative correlations between the Honesty Scale and many of the intermediate outcomes. The subjects were therefore divided into two groups based on their posttest Honesty Scale score. Those who had responded to four or more items in a socially desirable way were placed in one group, "lie", with the remaining in another group, "honest". This variable was then entered into a series of univariate analysis of variance tests as an independent variable, along with treatment condition. All scales were analyzed for differences between youths who responded in a socially desirable way versus those who did not. The only significant main effects were for the Normative Pressure (F(1,71) = 9.50, p < .005) and Specific Bonding Scales (£(1,71) = 13.46, p < .001). Youths who gave more socially desirable responses reported that their peers had less delinquent values and that they had greater attachment to conventional values. In addition, on the Self-esteem Scale, there was a significant interaction effect for treatment condition and honesty (E(1,71) = 5.31, p < .05). Youths in the control group who gave socially desirable responses also reported greater self-esteem than 145 did youths in the control group who were more honest in their responses. Finally, there was a significant condition, status and honesty interaction for the Delinquent Associations Scale (F(1,71) = 4.24, m < .05). For formally processed experimentals, the "dishonest" youths scored higher than the "honest" youths while for informally processed experimentals, the honest youths score higher than the dishonest youths. The results were opposite for controls. There were no significant differences for any of the other variables (see Appendix W). As with the intervention integrity variables, all analyses were repeated, excluding those who thought that they were required to perform community service because of the draw of the envelope. The only change in findings was that this resulted in a significant time effect for Delinquent Associations, but there was no longer a significant time effect for Normative Values. Smmmary of Intermediate Outcomes There were no differences between experimentals and controls in terms of delinquent and conventional bonding, normative values, or self-esteem. Final Outcome Recidivism Research question #11 asked "Did community service participants have lower recidivism rates than controls?" The experimental group had a 27% recidivism rate versus 12% for 146 controls. A repeated measures analysis of variance for previous offenses and recidivism, found significant effects for condition, status, time and the status by time interaction (see Table 37). Experimentals had more offenses than did controls. Formals had more offenses than did informals, almost by definition. That there were more previous offenses than there were subsequent offenses probably reflects the fact that the time period covered by previous offenses was much greater than that for subsequent offenses. The status by time interaction reflects the fact that formals had more previous offense than did informals but the difference between the groups on subsequent offenses was Ismaller. Again, this is most likely due to the short time 13eriod by which subsequent offenses were counted. However, an ANOVA found no significant differences Iaetween the experimental groups for the number of subsequent c>ffenses (F (1,72) = 3.43, p =.07). Further, a repeated nueasures ANOVA performed after dichotomizing prior and ssubsequent offenses as either no offenses or one or more (>ffenses found no sigfificant condition effect or interaction (effects involving condition. The subsequent petitions were based on new offenses as \vell as violations of probation. If those youths whose ‘petitions included only a violation of probation, with no new offenses, were excluded, the recidivism rate would be 24% for experimentals and 9% for controls. Table 37 ANOVA of Recidivism 147 Test time Condition Status (n) Previous Subsequent mean (sd) mean (sd) I I I (19) I 1.23 (1.20) I 0.58 (0.90) 1 Experimental I I 1 1 I I Informal (16) I 0.18 (0.54) I 0.25 (0.58) I I I I I I I (20) I 0.85 (0.93) I 0.30 (0.57) 1 Control I 1 I I I I Informal (18) I 0.11 (0.32) I 0.00 (0.00) I I I 1 Analysis of Variance Source DF MS F Prob . w 2 (Dondition (C) 1 2.35 4.01 .049 .04 Status (S) 1 13.53 23.11 .000 .24 C X S 1 0.30 0.52 .474 Subjects 69 0.59 'Time (T) 1 3.73 7.59 .007 .08 C X T 1 0.00 0.01 .923 S X T 1 3.18 6.48 .013 .11 C X S X T 1 0.21 0.44 .511 Error 69 0.49 148 Exploratory Analyses The number of subsequent offenses was examined in relation to the same process variables as were the intermediate variables: the number of community service hours ordered, the number of petitions, the seriousness of the current offense, and the number of days from pretest to posttest, (see Table 38). There was a significant correlation between the number of prior petitions and recidivism (m = .22, p < .005). There was also a significant correlation between the number of days from pre- to posttest and recidivism (; = .31, p < .001). The correlation between the number of prior petitions 23nd recidivism were examined separately for controls and Texperimentals. For those youths in the control group, there 1was a significant positive relationship between the number of jprior petitions and recidivism (m = .39, p < .005). For experimentals, however, there was not a significant relationship between the number of prior petitions and recidivism (g = .21). A Fischer’s g to m transformation (Cohen & Cohen, 1975) showed that the difference between the correlations of experimentals and controls was not statistically significant. §mmm§ry of Fingl Outcomg The experimental group had a significantly higher recidivism rate than did controls. Recidivism was 149 Table 38 Correlations Between Process Variables and Subsequent Petitions Subsequent Petitions Process Variables Recidivism Community Service .05 Hours Number of Prior .22* Petitions Seriousness of Current .18 Offense .Days from IPre— to .31** Posttest Days from Current (foense ‘to Pretest .00 8 E ( .005. XX 2 < .001. INOTE: Correlations based on 73 to 81 subjects. 150 significantly related to the number of prior petitions and the number of days from pretest to posttest. Additional Tests of the Intervention Two different processes were employed to further explore the extent to which the intervention and intermediate variables were related to recidivism. First, those who received a clear message of accountability were separated from those who did not. This allowed a test of the importance of accountability in producing results. Second, typologies of youths were developed to see if there were different youths for whom community service might be most effective. These typologies were based on the variables of delinquent and conventional bonding. Accountability Research question #12 asked "Did subjects who had high :scores on accountability measures at posttest differ from ‘those with low scores in terms of other outcome measures?" ESubjects were therefore divided into groups based on the eextent to which the message of accountability was received Eand these groups were then compared to examine the relative (effectiveness of the intervention. This was done in two Ways . First, youths were divided into two groups based on their posttest perception of the purpose of community service (A3). Those youths who saw community service as reparation were in one group and those who saw it as punishment, 151 deterrence, rehabilitation, or revenge were in the other group. There were no significant effects involving the purpose of community service on any of the outcome variables (Attitude Toward the Victim, Self-esteem, Delinquent Association, Normative Pressure, Normative Values, General Bonding, Specific Bonding, Understanding, Effort, Benefit, and Recidivism) (see Appendix X). Similar comparisons were performed on just the subjects in the experimental group, where the notion of reparation was thought to be strongest. Youths in the experimental group who thought the purpose of community service was reparation had a significantly lower sense of accountability according to the Accountability Scale, than did youths who thought the jpurpose of community service was something other than :reparation (£(1,36 = 5.23, E < .05). There were no :significant differences for any other variables (see Appendix 1?). Second, youths were divided into two groups based on ‘their scores on the Accountability Scale. Youths with posttest mean scores less than 5.00 were placed in one group (n = 41) while those with scores of 5.00 and up were in the second group (n = 38). This divided the subjects into high- and low-intervention integrity groups to test this model. The only significant difference involving accountability was found for the Specific Bonding Scale. Those youths with higher feelings of accountability at posttest, had more 152 attachment to specific conventional norms (E(1,75) = 3.98, p < .05) than did those with low feelings of accountability. When the experimental group was examined separately, there was a significant effect for accountability on the Delinquent Associations Scale (F(1,36) = 4.81, p < .05). Youths who felt more accountable also felt less attached to delinquent peers at posttest, than did youths who felt less accountable. There were no significant differences for any other variables (see Appendix 2). Typolqgies Research question #13 asked: "Can youths be classified into two or more groups based on the extent of conventional and delinquent bonding at pretest? If so, are there differences between these groups/types of subjects on other toutcome variables?" Cluster analysis empirically forms groups of similar entities. The first step in cluster zinalysis is determining which variables should be used to (create these typologies. Ideally, theory should drive their Eselection (Aldenderfer & Blashfield, 1984). However, one Eshould also limit the number of variables used for (classification. The more variables that are used, the less zany one variable will contribute to the classification. In addition, interpretation of the solution is more difficult 'with large numbers of input variables. Ideally one should also attempt to avoid the use of highly correlated variables. 153 And finally, in selecting input variables, one should minimize the amount of missing data (Amdur & Herman, 1988). This cluster analysis used the Delinquent Association, Normative Pressure, General Bonding and Specific Bonding Scale scores from pretest as input variables. While these scales were correlated, the multi-dimensional causal theory that drove this research identified these as important variables. The next decision involves the standardization of these input variables. While most studies do standardize the variables, there are some disadvantages of doing so (Aldenderfer & Blashfield, 1984). While standardization equally weights all variables, it is not an equivalent transformation if the distributions of the items differ zacross subjects. This study was concerned with ensuring the (equal weighting of variables and these scales were therefore (converted into z-scores prior to the cluster analysis. The third step in preparing a cluster analysis is sselection of a similarity measure which is computed for each ]pair of subjects. Correlation and distance measures are the Inost common in the social sciences (Aldenderfer & Blashfield, 1984). Similarity coefficients assess the shape, elevation, and scatter of a distribution. However, correlation is sensitive only to shape. A distance measure which is sensitive to all three dimensions is the Euclidian Distance. 154 This was selected as the similarity measure for this analysis. The next step is deciding on the clustering method. Different methods of clustering produce different solutions (Aldenderfer & Blashfield, 1984). Hierarchical agglomerative methods are the most popular. Hierarchical methods work in a tree structure format where the most similar subjects are combined into a cluster. Then the next two most similar clusters are brought together in a higher order cluster. One disadvantage of this procedure is that subjects are not removed from a cluster if they no longer fit. Ward’s method is a common hierarchical method which optimizes the minimum variance within each cluster. For each cluster, the squared Euclidian distance to the cluster means is summed. The (combination of two clusters that results in the smallest increase in the overall sum of squared-within cluster (distances is selected as the next higher-order cluster (Norusis, 1985). A second common method of clustering is iterative Jpartitioning. This process begins by partitioning subjects :into clusters by some method (selection of clusters is left ‘to the researcher), and the centroids of these clusters are (calculated. Then each data point is moved to the nearest cluster and the centroids are recalculated. These last two steps are repeated until there is no movement of subjects from one cluster to another. Often, the initial seeds can be 155 determined from a hierarchical analysis (Aldenderfer & IBlashfield, 1984). The number of clusters must be specified in these procedures. For this study, Ward’s method was used 'to generate initial seed points, and the iterative jpartitioning procedures were used until clusters stabilized. The next decision concerns the number of clusters to (create. This may be determined by theory or by examining the :resulting clusters. Due to the sample size, two, three, and :four cluster solutions were considered; anything greater tdould have resulted in small sample sizes within the (clusters. After completing the hierarchical analysis with ‘the CLUSTER program of SPSSx (SPSS Inc., 1986), the means of ‘the clusters for the two, three, and four cluster solutions tdere entered into an iterative partitioning method using the (QUICK CLUSTER program of SPSSx (SPSS Inc., 1986). Each of ‘the clusters stabilized after five iterations, and the :resulting clusters were examined. The two cluster solution generated more conceptually .+ 00a>how " hfiduslioo 0050: u . uncapnaooonc unauaom 000308 00w>00m nonnommo lau>dcaoom ucozucqmon Hanoaov noczvupp< hpaaqnupasooo< haunsaloo 00mpm QN.+ 163 Further, a cluster analysis grouped participants into a high scoring and a low scoring group, based on delinquent and conventional bonding scores at pretest. There were no differences between these clusters on posttest scores except on the Attitude Toward Community Service Scale. Path analyses showed relationships somewhat similar to ‘what would be predicted, with some notable exceptions. Summary of Outcome There were no differences between experimentals and controls in terms of delinquent and conventional bonding, normative values, or self-esteem. Further, the level of accountability at posttest was not related to outcome. Finally, the level of conventional and delinquent bonding at pretest was not related to outcome. DISCUSSION Several research questions were proposed by this research. The questions fell into three main areas: implementation of both the research, per se, and the community service intervention; the relative impact of community service on outcomes of delinquent bonding, conventional bonding, self-esteem, and recidivism; and the relationship between community service and a theory of delinquency causation, as well as the philosophy of accountability. In the first section a summary and interpretation of the findings is provided. Then, possible explanations for these findings are discussed. Finally, recommendations for future community service programming and research are also be discussed. First, however, two methodological problems must be discussed. Prior studies suggested that a sample size of 82 was adequate for 70% power. This is, however, most likely an underestimate given the nature of the prior research findings on which this estimate was based. As a result, the current sample may well be too small to detect any real changes or differences between the groups. Further, the intermediate variables relied on self- report. Seven of the nine intermediate outcome variables were significantly negatively correlated with the Honesty Scale. This means that youths who responded in socially desirable ways on the Honesty Scale also responded in 16“ 165 socially desirable wyas on the other scales. While social desirability may be considered by some to reflect a response set bias, it could also reflect general positive presentation of self. Summarygand Interpretation of Findings Implementation Successful implementation of the research design and intervention is one of the most crucial issues in evaluation. Research is often not fully implemented as designed and it is therefore necessary to systematically assess implementation variables. Tornatsky and Johnson (1982) have suggested that measurement of the degree of implementation is as important as the measurement of its effects. Therefore, both the research design and the community service intervention were examined. implementation of the research. In this study, the number of community service hours ordered, the number of court ordered requirements, the type of court ordered requirements, the number of missed pretest and posttest appointments, and the number of days from pretest to posttest were used to describe the implementation of the research as well as to document the equivalence of the experimental and control groups. Random assignment was maintained for all subjects and these comparisons documented that there were no significant differences between youths in the groups based on these or demographic variables. 166 Implementation of community service. This study zassessed the extent to which the experimental intervention ()f community service was implemented. One complaint of lresearch in restitution has been inadequate program (iescriptions (Gendreau & Ross, 1987). In addition, to cietermine if community service is a viable sentencing silternative, one must know what that alternative consists (sf, in detail. Sechrest and Yeaton (1981) have emphasized ‘the importance of examining both the strength and integrity (of treatment. Strength is defined as the planned intensity (of the treatment and integrity as the degree to which the intervention was implemented. Overall, the results of 'the implementation assessment suggest that the community service intervention was implemented as designed. The lobservations of intake and placement interviews demonstrated that all youths receiving community service had similar experiences during these interviews. Further, those items ‘which were to be covered during these meetings were covered. While there was a broad range of community service hours ordered and completed, the number of hours reflected the seriousness of the offense, which was consistent with the use of the matrix to determine these hours. The average lTumber of community service hours ordered was 35 hours. The JI‘esearch was designed to accommodate variation in the number <>f community service hours so that the punishment would be 167 proportional to the offense (Rubin, 1985-1986, Schneider, 1985). The amount of time that elapsed between the offense and the pretest was just over 100 days. This is a fairly long time, and possibly enough time for youths to lose sight of the connection between the offense and the penalty. While speedier processing by the legal and judicial systems may be difficult, a long lag time may serve to undermine one of the fundamental principals of this sanction: that of an evident relationship between crime and penalty. The number of days that it took to place youths in a work site also varied from 0 to 183 days. This is most likely due to the fact that youths with violent histories were assigned to the work crew programs which operated in spring and summer months. Therefore, if a youth was initially interviewed in the fall, there may have been a six month wait for the next work program. Alternatively, some youths consistently did not show up at their placement appointments, thereby delaying the start of their community service work. The majority of youths (91%) successfully completed their community service responsibilities. This completion rate is consistent with that of other studies of restitution (Cooper & West, 1981a, 1981b; "Expansion", 1981; Keldgord, 1978; Maori, 1978; Nelson, 1978; Schneider, 1983). 168 The intensity of the community service intervention was assessed in four ways. Both the number of days at the work site and the number of days elapsing between the first and the last day of work were calculated. Youths worked an average of six days during the course of just over one month. The number of community service hours ordered was the divided by these two numbers so as to create the number of hours worked per day at the work site and the number of hours worked per day available for work. The average number of hours worked per day at the work site was just under 5 1/2 and the average number of hours worked per day available to work was just over 3 1/2 hours. Had this interval been based on the date of referral and the last day of work, the level of intensity would have been even less. Collectively, these measures do not describe a very intense intervention. Youths were not ordered to complete very many hours, they were not at the work site for very many days, and they did not take very long to complete their required community service hours. One could therefore question whether this was sufficiently intense to produce the desired results. One of the goals of this community service program was to produce a sense of accountability in its participants. Earlier studies have shown that restitution programs based on this philosophy are successful (Gilbeau et al., 1980; Hofford, 1983; Remington, 1979). The extent to which this 169 message got across to youths was assessed with a number of measures. First, youths were asked what they thought was the main purpose of different criminal sentences. Two- thirds of the subjects in the experimental group recognized, at posttest, that the purpose of community service was reparation. This was consistent with previous research (Thorvaldson, 1978, 1980a; Van Voorhis, 1985). While 20% more experimental at posttest than at pretest stated that the purpose of community service was to repair the damage a person did, this change was not statistically significant. There were significantly more positive attitudes at posttest than at pretest toward the community service aspects of the sentence for experimentals than for controls (Community Service Scale). However, there were no differences between experimentals and controls on their sense of accountability or their attitudes toward their victim. Heide (1983) suggested that a successful restitution program requires both an appreciation of the payback notion and an awareness of victim needs. Neither of these messages were strongly achieved, limiting the potential of the program. It has been suggested that completion of community service must be an effort or sacrifice of time; the assignment should be clear, measurable, and achievable; and it should produce some rewards (Keve, 1978). Each of these variables was assessed at posttest for those youths in the experimental group. The Effort Scale assessed the extent to 170 which the youths felt that completion of community service was easy or was inconvenient and required a sacrifice of time on their part. The average score was 2.15 on a 4 point scale suggesting that youths did not perceive community service to be much of an effort. This is consistent with the fact that the number of community service hours ordered and the intensity measures were low. The Understanding Scale assessed the extent to which youths understood the requirements of community service and thought they could complete it, or clear and achievable as described above. The average score was 3.82 on a 4 point scale suggesting that the community service was in fact clear and achievable. The Benefit Scale assessed the extent to which youths felt that they had learned a new skill or changed as a result of community service. The average score was 2.26 on a 4 point scale. Half of the youths had responded that they had benefitted somewhat or a great deal as a result of the community service. It should also be noted that half of the youths felt little or no benefits. Based on the Keve’s (1978) suggested components of a promising community service program, this intervention successfully met one (measurable and achievable), failed to meet one (a sacrifice of time, an effort), and had mixed results on the third (produce a benefit).' Overall, therefore, this program may not be expected to have strong 171 positive impacts because of the low intensity and little effort that youths reported. Outcome Both intermediate and ultimate outcome variables were examined. Intermediate variables were those related to self-esteem, and delinquent and conventional bonding suggested by a multi-causal delinquency theory (Elliott et al., 1985). The ultimate outcome was recidivism. This study found no differences between experimentals and controls on the extent to which youths associated with delinquents (Delinquent Associations Scale), their own values related to delinquent associations (Normative Pressure Scale), conventional bonding (General Bonding and Specific Bonding Scales), or self-esteem. While early studies showed that good self-esteem could insulate against delinquency (Dinitz et al., 1962; Reckless et al., 1956; Scarpitti et al., 1960), this study found no significant differences for self-esteem. The overall recidivism rates suggested that approximately one-quarter of the experimentals and 10% of controls had new petitions filed between pretest and posttest. The difference in these rates for experimentals and controls was statistically significant, but the number of offenses was small, consistent with prior research (Klein, 19 ; Schneider et al., 1982). This is not inconsistent with the causal model which guided selection of 172 the intermediate variables. This model suggested that because the intervention was not associated with an impact on delinquent or conventional bonding, the recidivism rates of experimentals should be no better than controls. In fact, this study showed that they were worse. Relationship to Restitution and Delinquency Theory The community service program which served as the experimental intervention in this study was based on a principle of accountability which suggests that community service will teach youths to feel accountable and that this accountability will affect subsequent delinquent behavior (Armstrong, 1983; Gilbeau et al., 1980; Hofford, 1983; Schneider, 1985). The examination of the extent to which the message of accountability was perceived by the youths showed no significant differences between youths who completed community service and those who were in the control group. Further analyses were then designed to examine differences between those who felt accountable and those who did not. It was found that for youths in the experimental group who believed the purpose of community service was reparation had significantly lower feelings of accountability. This is directly in contrast to the expectations of many authors (Armstrong, 1983; Schneider, 1985). However, in support of these authors, youths who felt a sense of accountability at posttest, also felt attachment to conventional norms. Experimentals who felt a 173 sense of accountability at posttest were also significantly less attached to delinquent peers. While these results offer partial support for this theory, the inconsistency of findings is not encouraging. A second test was designed to see if typologies of youths could be formed that were related to future delinquency. While there was clearly a group of youths in this study who felt strong attachment to conventional norms and little attachment to delinquent others, and another group who had strong associations with delinquent peers and weak attachments to conventional values, consistent with Elliott and associates (Elliott et al., 1985), this classification was unrelated to almost every other variable. Further, this classification does not represent two distinct typologies, but rather divides a group of subjects into high and low performers. A final test was designed to test a model of delinquency causation based on theory (Elliott et al., 1985) with the addition of certain process variables. The results were mixed. The model showed that the background and process variables were intervening as expected for all subjects at pretest. There was a positive relationship between prior offenses and hours of community service, a negative relationship between both hours of community service and time from offense to pretest and Accountability, a positive 174 relationship between Accountability and Attitude Toward Community Service and Attitude Toward Community Service and both General Bonding and Specific Bonding. The posttest results for experimental subjects on this model showed that the relationship between number of community service hours and Accountability was positive. This suggests that the community service increased the sense of Accountability such that there was no longer a negative relationship between these variables, but a positive one. Further, the relationship between prior offenses and recidivism, thought to be one of the most enduring, was small and negative for this group, while it was larger and positive without intervention at pretest. Again, this suggests that participation in community service may disrupt some of these relationships. However, it must be noted that other relationships changed in unexpected and undesirable directions. The path analysis results show that community service may not be a solution to recidivism. Where Did the Study Go qumg1_ Rossi, Freeman, and Wright (1982) described three kinds of implementation failure: no treatment, or not enough treatment was delivered; the wrong treatment was delivered; or the treatment was unstandardized. The examination of the process data clearly demonstrated that treatment was received. However, it was also shown that relatively few 175 community service hours were completed over a short period of time, with little intrusion into the lives of the participants. It could easily be argued that not enough treatment was delivered to demonstrate an effect. The next question concerns whether the wrong treatment was delivered. It was clear from the process review that the treatment was delivered as designed, so in that sense, it was not the "wrong" treatment. The next issue, then, is whether or not this was an appropriate treatment. There is considerable literature that suggests that community service is at least as effective at reducing subsequent criminal behavior as are other sanctions (Schneider et al., 1982). The recidivism rates found in this study are consistent with others (Cannon & Stanford, 1981; Lajeunesse, 1979; Schneider et al., 1982; Schneider, 1983). Further, the intervention directly addresses components found in a delinquency causal model that has high explanatory power. Theory suggests that the intervention should work. However, the path analysis results produced mixed results about the extent to which the intervention worked in practice in this setting. It is most likely that the amount of community service completed was not enough to have an impact in the expected way. Further, the amount of time that it took to introduce youths to the idea of community service may also contribute to this failure. Since the intervention did not show increased feelings of accountability, it is not surprising 176 that experimental youths performed no better than controls in terms of conventional or delinquent bonding, or recidivism. Methodologically, statistical power was low and, for intermediate variables which relied on self-report, social desirability may have interfered. The possibility, however, does remain that it simply was the wrong intervention. Further, it is true that the amount of community service varied across participants. The intervention was designed this way so that punishment would be proportion to the offense (Rubin, 1985—1986; Schneider, 1985). It is unrealistic to expect that a specified amount of community service would be ordered regardless of the offense (it would equivalent to sentencing all adults to 5 years in prison regardless of whether they shoplifted a loaf of bread or raped and murdered 20 women). Because the literature suggests that it is best to set the amount of community service individually, it is unlikely that this accounted for the lack of positive findings. Finally, there has not been sufficient controlled research on community service to suggest what the ideal intensity should be for programs to be effective. Earlier studies which have also evaluated programs with few hours of community service, such as Wax (1977) with 20 hours and Koch (1985) and Davidson and Johnson (in press) with 32 to 48 hours over 3 months, have also failed to show significant 177 effects on recidivism or intermediate outcome variables. This patterns of findings therefore suggests that these low orders are not sufficient. Summary This study showed that there was complete implementation of the research design and the community service intervention. The random assignment was completely maintained and the experimental group did in fact perform community service while the control group did not. Further, the intake and placement interviews included the components designed to emphasize the accountability philosophy of the program, and most youths completed the required amount of community service. However, despite complete implementation of the intervention, the data suggest that the intervention may not have been sufficiently intense to produce the hypothesized effects. There was consistent lack of support for the idea that this community service program would increase the participants’ sense of accountability or appreciation of the victim. Further, there was little impact observed on conventional and delinquent bonding and recidivism, consistent with some recent community service research (Davidson & Johnson, in press; Koch, 1985). The most two plausible explanation for these findings are that of implementation failure and theory failure. While it is possible the wrong intervention was selected (however, this 178 intervention was selected because of its relationship to theory), it is also likely that the amount of community service performed was not adequate enough to produce the desired effects. The literature suggests that community service should be an effort for the youths. Participants’ responses suggested that community service was not an effort, and examination of the intensity of the intervention would also lead one to conclude that it was not. It is also possible the time lag from pre to posttest was not sufficient to show differences in either intermediate or ultimate outcome variables. Conclusions and Recommendations This study has reported the results of an experimental test of the effectiveness of community service. On the basis of these findings, a number of conclusions can be drawn about the adequacy of the intervention and recommendations for future programming and research can be offered. It was proposed that, given successful implementation, recidivism would be reduced for participants in community service, through increased conventional bonding and decreased delinquent bonding. While the program was successfully implemented, there were no findings to support the impact on either recidivism or the intermediate variables. However, the low recidivism rates overall are 179 consistent with other studies (Heinz et al., 1976; Hudson & Chesney, 1978; Koch, 1985; Schneider, 1986; Schneider et al., 1982; Wax, 1977). An examination of threats to internal and external validity, as well as implementation failure, suggest that this research failed to show the expected results for a number of possible reasons. First, the intervention described was rather weak. While the community service was implemented as designed, the intervention was not very intense. The number of community service hours and the amount of time it took to complete these hours suggest an intervention which the youths did not see as an effort or sacrifice. While others (Rubin, 1985-1986; Schneider, 1982) suggest that the amount of community service should be proportional to the amount of damage, perhaps a more. successful alternative would be to make the intensity of the community service proportional. Instead of specifying the number of hours of community service to be performed, orders could, for example, specify the number of community service hours to be completed weekly and the number of weeks to be worked. This can still be influenced by outside factors such as jobs, extra-curricular commitments, family commitments, and the like. Nevertheless, such a procedure could make the community experience more intensive, while keeping the intensity dependent on the amount of damage as well as extraneous variables. 180 Second, the posttest and recidivism assessment was performed shortly after completion of the intervention. It is possible that some of the interpersonal changes that community service is thought to create don’t become immediately obvious. A longer follow-up period, therefore, may have revealed some delayed outcomes. Third, social desirability influenced the scores of the intermediate variables, making outcome conclusions impossible. Fourth, the sample size may have been too small to show real differences. Finally, community service may not be an appropriate way of reducing recidivism of youthful offenders. Keve (1978) has argued that the economic and political benefits of community service are adequate justification for continuation of this sanctioning option. Judges, line- workers, victims, and community members have consistently shown support for community service (Chesney, 1976; Hudson et al., 1980; Novack et al., 1980). These factors alone may be justification for continuing investigations into this alternative. Given the findings of this study, further research is necessary to evaluate more intensive models of community service. New studies should also have larger sample sizes, more refined measurement tools, and longer follow-up periods. APPENDICES 181 APPENDIX A Distribution of Subjects by Variable Source Pretest Posttest Variable Exp. Control Exp. Control Interview Demographics 47 47 na na Purpose of Community Service 47 47 38 41 Attitude Toward Community Service 41 41 33 35 Attitude Toward Victim 46 45 37 39 Accountability 47 47 38 41 Self-Esteem 47 47 38 41 Delinquent Association 47 47 38 41 Normative Pressure 47 47 38 41 Normative Values 46 46 37 40 General Bonding 47 47 38 41 Specific Bonding 47 47 38 41 Honesty 47 47 38 41 Understanding na na 38 na Effort na na 38 na Benefit na na 38 na Court Records - Successful completion and posttested Recidivism na na 35 38 Current Offense na na 35 38 Total Sanctions na na 38 41 Community Service Records - Successful completion and posttested Community Service Hours na na 38 41 Pretests Missed 44 45 na na Posttests Missed na na 38 41 Days in study na na 38 41 Days to placement na na 37 na Days working na na 27 na Days from first to last day of work na na 33 na 1232 APPENDIX B Date of Referral Caseworker COMMUNITY SERVICE REFERRAL [JIndividual Placement EJStaff Supervised Work Program Referral (non-assaultive offenders) (CSC, Arson, Violent Offenders) Hinor's Name 008 Phone Parent/Guardian Address School Full-time Part—time Not enrolled Extra-curricular Activities Presently employed No Yes - Hours worked Program Recommendations: Total number of volunteer hours recommended/ordered Date volunteer work is to be completed Names of any co-offenders also referred to Community Service Program - Intake Only - Participation in this program is: An Intake Diversionary measure Offense Information: Date of Offense Number of victims Type of Offense Previous Police or Court History. if known Victim Information: Type of Victim: person household school or public preperty store or business Victim loss Actual documented loss 3 Total amount recovered (if any) 3 Amount paid back by way of restitution to this date (if any) 5 Other: (placement consideration, specific concerns, potential problems) ‘Please attach copy of first sheet of Report of Investigation for all formal cases of legal status. Maximum Full-time school Job Extra-curricular school Extra—curricular court No priors Minimum 183 APPENDIX C Matrix of Community Service Hours Class A (15-40) 40 Class B (40-70) 70 Class C (70-130) 130 -10 ~10 -10 -1O -20 7O 184 CLASS A Assault Assault and Battery Conspiracy to Commit Offense Curfew Violation Disorderly Conduct Disturbing the Peace Fraud (Attempted) Illegal Possession of Credit Card Improper Registration of Motor Vehicle Indecent Exposure .Larceny under $100 Littering Malicious Destruction of Property under $100 Minor in Possession (Alcohol) No Operator's License No Proof of Insurance Police Officer. Failure to Obey Police Officer, Obstructing by Disguise-- Possession of Burglar Tools . Possession of Imitation Controlled Substance Prostitution ' Prowlinq Receiving and Concealing Stolen Property under $100 Trespassing Possession of Brass Knuckles Driving, Suspended License 185 CLASS B Ambulance Call, False Assault, Aggravated Assault with Intent to Rob (Unarmed) Breaking and Entering Coin Machine Delivery of a Controlled Substance Dog Fight, Causing A Dog, Keeping For the Purpose of Fighting Embezzlement Entry Without Breaking. Failure to Report Property Damage Accident Fire Alarm. False Forgery, Uttering, and Publishing Fraud Larceny Over $100 Larceny from Motor Vehicle Malicious Destruction of Property (Attempted). Operating Motor Vehicle Under Influence (Alcohol) Police Officer, Fleeing and Bluding Police Officer, Resisting and Obstructing Police Report. False Possession.of Controlled Substance Possession of Open Alcohol in Motor Vehicle Possession of Switchblade Knife Receiving and Concealing Stolen.Property Over $100 Robbery, Unarmed Safebreaking Unlawful Use of Motor Vehicle Possession of a Controlled Substance, Intent to Deliver Carrying a Dangerous Weapon CLASS 8 OR C Arson (Attempted) Violation of Court Order/Probation 186 CLASS C Arson Assault with a Deadly Weapon Assault with Intent to do Great Bodily Harm, Less than Murder Assault with Intent to Murder Assault with Intent to Rob (Armed) Breaking and Entering ' Breaking and Entering (Attempted) Carrying a Concealed Weapon Criminal Sexual Conduct Extortion Firearm, Careless Discharge Firearm, Illegal Discharge Negligent Homocide Possession of Firearm During Commission of Felony Robbery. Armed Robbery, Armed (Attempted) Unlawful Driving Away Auto Unlawful Driving Away Auto (Attempted) Assault with Intent to Commit Criminal Sexual Conduct Driving. Felonious Bomb Threat f_+;EE—M ~ ~ +** 1837 APPENDIX D Appointment Letter DONALD 5. OWENS :00 Minute CCUNW 8mm R GEORGE ECONOMY '30) W. KAWAZOO Stan-r JUDGES Lansmc. Wows»: 43933 Tmcmlfla$|73| DOUGLAS W. SUD! COUNTV COURTHOUSE Count Wanton . Mason Macaw 48854 noun: um 51:11! of Michigan Tush-noun (5|?) 6760238 INGHAM COUNTY PROBATE AND JUVENILE COURT Dear Participation in the Ingham County Probate Court's Community Service Program by was recommended or ordered at his/her court hearing. As the Community Service Program Advisor. I have set an appointment for this minor and at least one parent for at - The purpose of this appointment will be to discuss your involvement in the Community Service Program. By the time of this scheduled appointment. the minor is expected to have his/her request for a work permit, which can be obtained from the home school or from the Lansing School District Administration Building, Room 305, 5l9 u. Kalamazoo, Lansing. Either a birth certificate or a driver's license is necessary in order to pick up the request for work permit. If you have any questions or are unable to attend this appointment. please contact the Community Service Program Office. Our meeting will take place at: Ingham County Probate Court 300 Ingham County Building 303-N..Kalamazoo Street ‘Lansing, Michigan‘ 48933 Juvenile Probation Office on Third Floor Sincerely. Gail Moore Community Service Program Advisor 485-l751, ext. 555 188 APPENDIX E Participation Agreanem: 'miscmrthasagreedtoparticipateinasuflydesigmdtofirdbettermys ofdealingwithyamgpeoplevtnaresenttojuvenilecourtforcertain offenses. 'niisproject isdesignedto findoutwhat optionsarebest for what kindsofpeople. ‘Itneevmoagreetoparticipateintheprojectwillbe helping court officials to make decisions about serving other young people in thefuture. ‘Ihisprojectisgoingtolookattmodifferentwaysofhelpingyouthsstaywt of trouble. One-half of the youths who participate will be required to carplete the order/reocImendatim handed down by the judge or the referee, which includes cmpletion of a specified umber of camunity service hours. 'meotherhalfofflieymuiswfllberequiredtocamleteaflyuiosepartsof the original order/recannerriatim which do not include camamity service. In othermrch, aymthordered/reccunerdedtosixmrthsprobationandmhmrs of cammity service would cmplete aflythesixmxthsprobatim ifheorshe wereinthissecmdgroup. 'nedetaflsoftrecammityservicerequirenentswfllbemrkedwtwithflie OammityServiceAdvisormefllmrkwiththeyamgpeopletodetermine appropriate placanartsandnegotiateanagreeableworksdaechle. Ifymagreetoparticipate,ymwillbeputintomeofthesetwogmipeby lottery. Ymhaveaneqzaldiarneofgettingintoeachgmip. Youdonot have toparticipate in theproject. Ifymdorntagreetoparticipate, you wfllbereqairedtocmpleteflieorigimlorder/recmnerflatimfrmfliejuige or referee. There will be no additional penalty for refusing participation. Ifymdoumrttoparticipateinthepreject,youwillbeaslcedtosignthis fem stating thatyouagreetothe following: 1. Wevolmitarilyagreetoparticipateinthissbny. Wearefreeto withdrawatanytim. Ifnewithdrawfrunthestudy,therewillheno paultyinadditimtothatalreadyprescribedhythecourt. 2. Weagreetobeassignedbylotterytoaieoftwogrcups: a. Grurplwillreceivetheoriginalorder b.6rup2willreceivetheoriginalordermixusthecmmmity servicereqriralent. 3. Wemrierstandthatymthsmoreceiveccnmmityservicevmldhave receivedcaunmityservioeanywayifthissmdywasmtbeirqcorducted. 4. Weunderstandthat ifanewoffenseiscmmittedduring-the study, orthe orders/recamexidations are not fulfilled, the youth will be treated by the police and the count officials in the usual mamer. Participation will inmmyaffectanymrrerrtorfuturecourtactiais. 10. 11. 189 Wemderetanithattheyungpeopleinbothgroupswillbeintervieded twice: once right now and once again a short time later. These interviewsshculdlastllSminrtestoonehair. Wandaowrate answersareneeded. We understand that the following kinds of information will be gathered during the interview: a. Background information sud: as family history, age, gender, and so (11. b. Attitmdestheyarthholdstowardsdifferentoptiaisthecairthas available, their victim, themselves, their friends, and social rules and expectatiais. Wegiveourpermissim forthereseardistafftoexaminethecmmt recordsoftheymthforthepastyearandthenexttmoyears. Wemiderstardfliatalloftheinformatimfrunflieinterviadsanicmrt records will be handled confidentially by the research staff. Infomaticn will only be released in group form and aralymisly (without names). Wemrierstarrithattheyardmwillreceivefoodcwpaisatcmpletimof fliesecaidinterviw. 'Iheinterviewsarecoleideredtobepartofthe shriyardcmpensatimisofferedforthetinespentincmpletingthan. Weurrierstanithattheremaybenodirectbenefitsasaresultof participation. However, otters my benefit in the long run because of the information whichisgathered. Atmyreqrest,asmmeryoftheresultswi11begiventonewhen available. 12. Wehavebeengivenadiancetotalkaboutthereseardistuiyandask questiais. If I have additimal greetions, I may contact Carolyn Feis, Department of Psychology, Michigan State University (517) 353-5015. We agreetoparticipateinflreshflydeecribedabovewiththemflerstarfling thatmarefreetowithdrawatanytinewittmtpenalty. Youth print signamre Parent print signature Staff print signature [bite 190 APPENDIX F UCRIHS Approval MICHIGAN STATE UNIVERSITY UNIVERSITY COMMITTEE ON RESEAICM INVOLVING EAST LANSING 0 MICHIGAN 0 “.14-IO“ HUMAN SUBJECTS (L‘CIIHS) In ADMINISTRATION IUILDING (SI?) 355-2186 May 12, 1988 Carolyn I... Feis Dept. of Psychology Psychology Research Building Dear Ms. Feis: Subject: "EXPERIMENTAL EVALUATION OF A COMMUNITY PROGRAM W" UCRIHS' review of the above referenced pro'ect has now been completed. I am pleased to advise that since the reviewer‘s comment has en satisfactorily addressed, the conditional approval given by the Committee at its May 2, 1988 meeting has been now cha'fi'ged to full approve . You are reminded that UCRIHS approval is valid for one calendar car. If you plan to continue this project beyond one ear, please make provisions for o taining appropriate UCRIHS approval prior to May 198 . Any changes in procedures involvin human subjects must be reviewed by the UCRIHS prior to initiation of the change. U RIHS must also be notified promptly of any problems (Knexpefited side effects, complaints, etc.) involving human subjects during the course of t e wor Thank you for bringing this project to our attention. If we can be of any future help, please do not esitate to let us know. Sincerely, J hn K. udzik, PhD. air, UCRIHS JKH/sr cc: W. Davidson 191 APPENDIX C COMMUNITY SERVICE PROGRAM APPLICATION Name: Birth Date: Address: Phone: School: School Hours: that type of volunteer work are you Later-ester". in doing? (make at least tiara choices and circle your choices.) 9.. Cleric..la : sl.ort..a.ad, typing, adding, ansrrering phones, filling out forts, calling people, filing, other 3. Indoor Maintenance: painting; vasl“ .i:.g wires-‘5. walls or floors: Sweeping: dust: ; emptying wasta'oaskets: repairing other C. tdoor Maintenance: lawn care, trash/litter pick-up, painting, snow remval, animal care, gard...‘ =iair.g, otia D. Htrran Services: cnild care, nursing hone or seaior citizen activities, handicapped ac .ivities, currenionsbip, reading to at .,ers other 8. Recreation: arts 5 crafts, spo.ts progrems, other P. Other: Do you have any special interests, skills or mbies? I agree to participate in this program to the best of my ability. I understand that if I an dismissed frcn this program due to poor pe_r‘or::~ance, I ray have to appear in Court for a preliminary hearing or a violation of my probation orders. Volunteer Date I, as parent or guardian. mfiersteni the philosp'ny and goals of the Comm-dry Service Program and agree that my child he a participant. , I will assist my child, to the best of my abilities, to see that these cornitrents are carried out, including arranging for transport: tion I authorize the Calamity Service Advisor to release information concerning my child's offense to the agency being considered for my child's volunteer plateau. ant. Parent“ 3 Sig-eat are Date Commit! Service Rinse: Dote 192 APPENDIX H Cooperating Agencies Arthritis Foundation Boys & Girls Club - North Boys 5 Girls Club - South Burcham Hills Retirement Center Capitol Alternative Education Central Y.M.C.A. Cristo Rey Church Cristo Rey Community Center Eaton County Parks Department East Lansing High School East Lansing Public Library First Missionary Church Gerber' 8 Children's Center Good Samaritan Family Center Haslett Public Library Holt-Delhi Public Library Holt United Methodist Church Holy Cross Church School Impression 5 Museum Indian Ingham Ingham Ingham Ingham Ingham Center County Building County, Hilliar Bldg. Co., Human Services Bldg County Humane Society County Library Ingham County Parks Department Lansing Area Safety Council Lansing Civic Players Guild Leslie Ambulance Service Leslie High School Mary Avenue Care Center Meridian Twn Fire Station #1 Michigan School for the Blind Oak Park Village Y.M.C.A. Okemos High School Parkwood Y.M.C.A. Potter Park Methodist Church R.E. Olds Museum Resurrection Day Care Center St. Casmir Catholic Church St. Vincent Home for Children Salvation Army Church South Church of the Nazarene V.F.W. Post 48 Waverly High School Webberville High School Whitehills Health Care Center Woldumar Nature Center Y.W.C.A. Youth Development Corporation 193 APPENDIX I Placement Interview Instructions Remember: You are to bring a letter of apology to your placement interview. This letter should take some thought on your part and be honest and sincere. The letter should include the following: 1. Address the letter to "Whom it may concern" 2. Explain that you are apologizing, and what you are apologizing for. 3. Explain that you accept responsiblity for what you did. 4. Expalin that you will be performing community service as a way to repay the community for the offense. This letter should be in your own words and not just a restatement of the four points above. Your' placement interview is scheduled for on at . You will be meeting with me and your work site supervisor , ' ' who can be reached at . **** c on o i 191+ APPENDIX J INGHAM COUNTY COMMUNITY SERVICE NORKSITE CONTRACT Minor's Name: Assigned Hours: Placement: 'Address: Contact Person: Phone: Days to Report: Time to Report: Beginning Date: Ending Date: WORK REQQIREHENTS l. You must arrive at the worksite on time. Transportation is not provided, so please make arrangements for a ride to and from the worksite. Remember, you have an obligation to fulfill your required hours and any absence must be made up. 2. If you are assigned to a full day of work, a lunch should be brought from home. A lunch break will be arranged. 3. Your duties will be explained to you at the worksite. It is your responsibility to complete the assigned tasks on a daily basis. 4. The following violations may lead to termination from the program: Excessive tardiness Excessive absence Refusing to do assigned work Not following instructions of work supervisors or others Leaving the worksite without permission Disruptive or destructive behavior Poor performance on the job could result in either reprimand or dismissal from the program. Dismissal may require you to appear for a preliminary hearing or be considered a violation of your probation orders. If you cannot make it to the worksite when scheduled, contact the worksite SUpervisor ggg_either your caseworker or me at the Probate Court, 485-l75l. Leave a message if you are unable to make contact. If problems arise, you are to contact the Community Service Advisor and your caseworker. I have reviewed this contract with the worksite supervisor and we are in agreement with it. Minor’s Signature Horksite Supervisor's Signature Community Service Advisor cc: Caseworker. Worksite Supervisor, Minor Date Person Method worker 195 APPENDIX K .KESUE/WXMNEER CONTACT SHEET Date Person Method Worker Date Person Method worker Date Person Method Worker l96 JCS ST: FDI— 13.2.2.3 CC\T .aCT SnZZT Date @072sz Zea/at 5/ flit/Lug W dad/Lola Person .&tAE2;L/ 4- a A”, .21/SW ax . 39:21ch ”7% W ..P 3614/ VW 'n‘or‘xer {SZ/jl ‘2/ ,,L¢‘,.é/.¢‘M_ #7534 ' XMA—c.’ E f 9% 7L ' Parsonw , 6W /fl/Q}0 C L? JO cm W “7 Date .r’/. Y an“, M, .7 W. ..e ezhcd n24: Aorker i'fl J—/7 ldfi MJMWL‘ZL-VWEC M’ZZMMW “elm-cam, S" 4 $7 3.9 3447/ _rvfl) , 197 CASE CONTACTS . . ' I MW:5‘ W. 3-9? at: 2212‘}; MW °. 7 /'00 WMWW , arson—W gag/1 WM wM WXW Zethod all. ! Z 'or'xar M .. 3.1.21! “90W it“ WM sai Lalo M We fin. . ’ oi IQ. l W. IJV .etfdw M be a) to (1.913 W MAW 0n ““3, m 1L4 cam Mg Ave vim Math! are» A may has to work H‘I'L Instead-oi l‘l. 3 . — ..Caflz) (W Two:- 4) inform ‘(3 W was gal 2%ch W O!) “& M“ $3 4» (0.16230 W ham CGLQMQ, arm-JILL , “53“ {9 (and, W5 WW1..1’E[C’) figggmaefiw “2 fl ’— Qm‘wé hm' wfiime. ileum» arson m U? W“ fathod | | Min (N3 =ate 'erson iethod 'orker ‘11:: | 'GI’SOII “' "in! 198 APPENDIX L WORK-SITE ASSIGNMENT Minor's Name 0.0.8. Address Phone Placement Site Hours Assigned Supervisor In Case of Emergency Contact Phone DATE HOURS WORKED VERIFICATION/COMMENTS * Include in comments rating of performance (i.e. inadequate, adequate, excellent). Please do not give the minor credit for work unless he/she came to work on time, worked hard and diSplayed a good attitude. 199 APPENDIX M Completion Letters R. GEORGE ECONOMY 303 w KALAMAIOO Snug-r JUDGES Lansmc. MICHIGAN 48933 TELEPHONE. is I 7) 465 l 751 DOUGLAS w SLADE Counv Ammnsmnon Paco": Rec-sun Comm" Counmoust MASON. MICHOGAN 48854 TELEPHONE (5|736760288 INGHAM COUNTY PROBATE AND JUVENILE COURT Dear You have successfully completed your comunity serVice hours as required by the Juvenile court. I hope that you now understand that there are consequences for your behavior, and that this has been a useful and positive experience for you. I understand that it may have been difficult for you to fulfill this obligation at times, and your persistence and dedication is recognized. It is now time for your follow-up interview. as explained when you agreed to participate in the research project. You will receive coupons from local merchants for this interview which is scheduled from on . If you cannont make this appointment. please call Carolyn Feis at 485-1751. ' Again congratualtions on your completion and don’t forget about your final interview. Our meeting will take place at: Ingham County Probate Court 300 Ingham County Building 303 West Kalamazoo Street Lansing, Michigan 48933 Juvenile Probation Office on Third Floor Sincerely. Gail Moore 200 300 INGHAH COUNTY BUILDING 303 w Kauwazoo STREET unsma. woman: 48933 Tun-Howe. (517) 4851751 DONALD S. OWENS R. GEORGE ECONOMY JUDGES DOUGLAS W SLAM Coon" Caunrnousc Count Aommsmnoa MASON. Memo“: 48854 pno.Ayg nggggftg fitate of withigan TELEPHONE (517) 676-0288 lNGHAMHflMMWFYPROBATEflwHDJUVENflJECOURT Dear It is now time for your follow—up community service interview, as explained when you agreed to participate in the community serVice research prOJect. You will recieve coupons from local merchants for this interview which is scheduled for on If you cannont make this appointment, please call me at 485-1751. Our meeting will take place at: Ingham County Probate Court 300 Ingham County Building 303 West Kalamazoo Street Lansing. Michigan 48933 Juvenile Probation Office on Third Floor Sincerely, Carolyn Feis DONALD S. OWENS R. GEORGE ECONOMY JUDGES DOUGLAS W. SLADE Count soc-«mumm- recurs Recast" 201 APPENDIX N Certificate of Recognition 300 nus-um Comm: DUILOPOG 303 w MWAIOO Stun LAN‘ONG. MtCmGAN 48933 Truman-c; (517) 48517:: Cm Cow-noon": MASON. NewcAN 48854 §taie nf gflichignn "WW“: '5 l 7! 6760288 INGHAM COUNTY PROBATE AND JUVENILE COURT January l7. l989 Dear Nathan He would like to congratulate you on the outstanding job you did while performing your Community Service.- Attached is a special certificate signed by both of the Probate Court Judges. Only seven other youth have received this honor in the past six months. we wish to thank you on behalf of the agency you served and know that if you ever need to use them as as an employ- ment reference, they will give you an excellent report. Good luck! Sincerely, Gail Moore Community Service Advisor 202 Uh<00¢& $0 HODDQ i=0 2had2wae for today and.canflt.worry'about what might happen to‘him.or’her tomorrow. 2811. It is easier for other people toldecide what is right than it is fer me. m2.fl)edm'oesformearximyfrielfisnekingitinlifeare getting‘worse, not.better; 1813. in) friends than“: seemitxa likeanralas ntxiiias tixgyldid.irliflwe m. ’ 1814. I«often feel awkward.and out of place. m5.It'snotworthp1armingforanythinginthefuturebeauseI really donflt know what is going to»happen these days. 1816. I sometimes feel like nobody cares about me anymore. (25) (25) (27) (23) (29) (30) (31) (32) (33) 229 14 Participant ID - Strongly Disagree - Disgree - Agree = Strunglypgree ****** 819. Ioftenfeel like it'smtmrtheventryingtodiangethings inmylife. 818. (me prcblan with the world today is that most pecple don't believe in anything. 819.1tseetsthatitishardertolcm1'mtoacttodaythanit usedtobe. 820.8yfriendsseentodiangetheirmirosaba1tthingsnoreoften thaninthepast. 821.8verythingd1angessoq1icklythesedaysthatIoftenhave trouble decidirg which are the right rules to follow. 822. PeCpIewerebetteroffintheolddayswheneveryonekrewjust howheorshewasexpectedtoact. ii # #IhrchuthCardIfi # # Iwouldliketohaveyuxramwerstoafwmorestatanentsabcut yourself. ReubenIaminterestedinycurcpinia'isarrlthereare norightorwruganswers. # # #1a-StronglyAgree # #2=Agree # #3=NeitterAgreeNorDisagree # #4=Disagree # :5=StronglyDisagree : Il. Hakingagocdinpressionisnoreinportantthantellingthe mmfrierds. 12. To stay out of trouble, it's sanetimes necessary to lie to teachers. I3. Makingagocdinprossimisnoreinportantthantellingthe tnithtoparents. (34) (35) (36) (37) (38) (39) (40) (41') (42) (43) (44) (45) 230 15 Participant ID # ii #1=StruiglyAgree # #2BPgree # #3=NeitherAgreeNorDisagree # #4=Disagree it :5=St2:onglyDisagree : I4. Ywmnnekeitinsdioolwitl'nithavirgtocheatmexamsor tests. IS. It's inportanttobehcnestwithyourparents, even if they becaleupsetoryougetpmushed. I6. Youhavetobewillingtobreaksaneniles ifyouwanttobe popular with your friends. I7. It’s inportanttodoycnramworkatsdioolevenifitneans salestudentsvm'tlikeyou. 18. In order togain the wt of your frierris, it's sanetimes necessarytobeatuponotherpecple. I9. At school, it's sanetimes necessary to play dirty in order to win. IlO.Sanetinesit'snecessarytolietoymrparentsinorderto keeptheirtrust. Ill.Makingagccdinprcssimisnoreinportantthantellingthe truthtoteachers. 112. It's okay to lie if it keeps your friends out of trouble. I13.Itmybemcessarytobreaksmeofydumpara1ts' rulesin ordertokeepsaneofyourfriends. I'mgoh'gtoreadywammberofstatelents personal attitudes and traits. Listen to sea) itan and decide whether the statane'rtistEorFAISEasitpertaimtoympersonally. # # tl=True # #z-False # # # J1. I have never intaeely disliked anyone. J2. moccasithavehaddcubtsabcutmyabilitytosucceedin life. (46) (47) (48) (49) (50) (51) (52) (53) (54) (55) (56) (57) (.58) (59) (59) .3 J3. J4. J5. J6. J8. J9. J10. J11. J12. J14. J15. .2 1 3 l6 Participant ID HytablenemiersatrxmeareasgocdaswhenIeatmtina restaurant. IfIcaildgetintoamviewithcutpayingandbesureIwas notseen, Iwaildprobablydoit. I like to gossip at times. I can renedaer "playing sick" to get out of sanething. I'm always willing to admit it when I make a mistake. I don't find it particularly difficult to get along with loud mouthed, obnoxious pecple. Isatetinestrytogetevenratherthanforgiveandforget. Iamalwaysccurtems,eventopecplewhoaredisagreeable. Beretavebeenoccasionswhenlfeltlikesneshingthirgs. Ineverresentbeingaskedtoremrnafavor. Ihavealmcstneverfeltthemgetotellsareoreoff. Imnsaletinesirritatedbypewlevmoaskfavorsofne. I have never deliberately said salething that hurt sareone's feelings. #mw_fi Finally, pleasetellmehoworwhyyouthimcymwereselected toreceivedcomunityservice? 1=8ecnusetheydeservedit 2-‘1hedrawoftheenvelcpe 3=Ot11er 8=N/A # # mte: #Czrd3# Interviewer: (I # B'IvelopeNunber: 2 32 17 maps 'llopmishsaneaae 'lbdetersareonefruncaunittirgfumrecrines 'Ibrehabilitatesaneone 'Ibgetrevenge 'Ibrepairthedamagethatthepersondid QRDB EXAMPLE: Foreadisetofwords, fickthenmberthatbsstslwwswhatkindof persmycntmmcyware: QJIET 1 2 3 4 5 6 7 1018‘! SAD 1 2 3 4 5 6 7 HAPPY fair 1 2 3 4 5 6 7 unfair helpful 1 2 3 4 5 6 7 harmful my 1 2 3 4 5 6 7 right tough 1 2 3 4 5 6 7 easy Pleasant 1 2 3 4 5 6 7 painful awaiting 1 2 3 4 5 6 7 dull frigitenirg 1 2 3 4 5 6 7 not frightening interesting 1 2 3 4 5 6 7 boring useful 1 2 3 4 5 6 7 worthless 233 UI UIUIUIUIUIUI 18 cccperative obeys rules polite harmfulto brave friend rude I 2 3 helpful to others 1 2 3 others cowardly 1 2 3 dtmb 1 2 3 honest I 2 3 lazy 1 2 3 tough 1 2 3 not wild 1 2 3 mean 1 2 3 kini 1 2 3 many 1 2 3 Card D aVERYGOODthingforpeopletodo amoothingforpeopletodo ammmODthingforpeopletodo ammflflmforpeopletodo IIBADthingforpeopletodo anYBADthingforpeqaletodo 2 34 QEDE Strongly agree Disagree Strongly disagree (ERDF Wwere SEVERALwere VERYFEWwere MEwere CARDG Dcn't know Yes a... n Strongly disagree Disagree Strongly agree 19 CARDI Straiglyagree Neitheragreenordisagree Strongly disagree 235 20 236 APPENDIX U Table U-l ANOVA of the Attitudes Toward Victim Scale Test time Condition Status Pretest Posttest I I I Formal I 4.66 I 4.56 I Experimental I I I I I I Informal I 4.57 I 4.63 I I I I I I I Formal I 3.98 I 4.40 I Control I I I I I I Informal I 4.65 I 3.84 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.74 0.32 .571 Status (S) 1 2.86 1.26 .266 C X S 1 3.03 1.33 .253 Subjects 72 2.27 Time (T) 1 0.76 1.49 .227 C X T 1 0.98 1.92 .171 S X T 1 0.02 0.04 .852 C X S X T 1 0.38 0.75 .390 Error 72 0.51 237 APPENDIX V ANOVAs of Bonding Scales Table V-1 ANOVA of the Normative Pressure Scale Test time Condition Status Pretest Posttest I I Formal I 1.96 I 2.29 Experimental I I I I Informal I 2.12 I 2.16 I I I I Formal I 2.04 I 1.95 Control I I I I Informal I 2.17 I 2.17 I I HHHHHHHHHHHH Analysis of Variance Source DF MS F Prob. Condition (C) 4 0.11 0.20 .655 Status (S) 4 0.37 0.72 .400 C X S 4 0.28 0.53 .466 Subjects 72 0.52 Time (T) 4 0.20 0.99 .324 C X T 4 0.51 2.59 .112 S X T 4 0.09 0.47 .497 C X S X T 4 0.35 1.74 .191 Error 72 0.20 Table V-2 ANOVA of the Specific Bonding Scale 238 Test time Condition Status Pretest Posttest I I 1 Formal I 3.77 I 3.84 I Experimental I I I I I I Informal I 3.72 I 3.63 I I I I I I I Formal I 3.48 I 3.62 I Control I I I I I I Informal I 3.86 I 3.90 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 4 0.06 0.10 .754 Status (S) 4 0.51 0.85 .361 C X S 4 1.84 3.04 .085 Subjects 72 0.60 Time (T) 4 0.12 1.36 .247 C X T 4 0.06 0.70 .406 S X T 4 0.12 1.33 252 C X S X T 4 0.00 0.00 .995 Error 72 0.09 Table W-l 239 APPENDIX W ANOVAs with Honesty Scale ANOVA of the Accountability Scale Honesty Condition Status Honest Dishonest I I I Formal I 4.96 I 5.03 I Experimental I I I I I I Informal I 4.79 I 5.02 I I I I I I I Formal I 4.79 I 4.62 I Control I I I I I I Informal I 5.21 I 5.06 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.05 0.17 .680 Status (S) 1 0.71 2.54 .116 Honesty (H) 1 0.03 0.12 .734 C X S 1 1.16 4.19 .045 C X H 1 0.40 1.46 .232 S X H 1 0.03 0.11 .746 C X S X H 1 0.02 0.08 .779 Error 67 0.28 Table W-2 240 ANOVA of the Attitudes Toward Community,Service Scale Honesty Condition Status Honest Dishonest I I I Formal I 4.95 I 5.64 I Experimental I l I I I I Informal I 5.41 I 5.40 I I I I I I I Formal I 4.26 I 4.90 I Control I I I I I I Informal I 4.74 I 5.24 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 4.24 3.08 .085 Status (S) 1 1.11 0.81 .373 Honesty (H) 1 3.08 2.23 .140 C X S 1 0.44 0.32 .576 C X H 1 0.22 0.16 .694 S X H l 0.70 0.51 .480 C X S X H 1 0.32 0.23 .644 Error 60 1.38 ‘ Table W-3 241 ANOVA of the Attitudes Toward Victims Service Scale Honesty Condition Status Honest Dishonest I I I Formal I 4.36 I 5.13 I Experimental I I I I I I Informal I 4.64 I 4.62 I I I I I I I Formal I 4.37 I 4.42 I Control I I I I I I Informal I 4.69 I 4.90 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.02 0.01 .912 Status (8) 1 0.88 0.62 .432 Honesty (H) 1 0.72 0.51 .478 C X S 1 1.17 0.83 .367 C X H 1 0.24 0.17 .685 S X H 1 0.37 0.26 .609 C X S X H 1 0.95 0.67 .416 Error 68 1.41 242 Table W-4 ANOVA of the Normative Values Scale Honesty Condition Status Honest Dishonest I I I Formal I 1.05 I 1.27 I Experimental I I I I I I Informal I 1.00 I 1.30 I I I I I I I Formal I 1.00 I 1.33 I Control I I I I I I Informal I 1.33 I 1.17 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.03 0.13 .719 Status (S) 1 0.00 0.00 .958 Honesty (H) 1 0.56 2.78 .100 C X H 1 0.12 0.61 .439 S X H 1 0.22 1.09 .300 C X S X H 1 0.35 1.74 .192 Error 67 0.20 Table W-5 243 ANOVA of the General Bonding Scale Honesty Condition Status Honest Dishonest I I I Formal I 2.99 I 3.25 I Experimental I I I I I I Informal I 2.88 I 2.97 I I I I I I I Formal I 2.77 I 2.98 I Control I I I I I I Informal I 2.99 I 3.05 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.09 0.48 .489 Status (S) 1 0.01 0.05 .825 Honesty (H) 1 0.34 1.75 .190 C X S 1 0.52 2.73 .103 C X H 1 0.01 0.05 .828 S X H 1 0.10 0.54 .463 C X S X H 1 0.00 0.00 .960 Error 71 0.19 Table W-6 244 ANOVA of the Understanding Scale Honesty Status Honest Dishonest I I I Formal I 3.61 I 4.00 I I I I I I I Informal I 3.94 I 3.90 I I I I Analysis of Variance Source DF MS F Prob. Status (S) 1 0 19 0 61 .439 Honesty (H) 1 0 26 0 81 .375 S X H 1 0.40 1.27 .268 Error 34 0.32 Table W-7 245 ANOVA of the Effort Scale Honesty Status Honest Dishonest I I Formal I 2.13 I 2.29 I I I I Informal I 2.31 I 1.97 I I Analysis of Variance HHHHHH Source DF MS F Prob. Status (8) 1 0.06 0 01 .926 Honesty (H) 1 0.07 0.11 .747 S X H 1 0.55 0.78 .382 Error 34 0.70 Table W-8 ANOVA of the Benefit Scale 246 Honesty Status Honest Dishonest I I I Formal I 1.93 I 2.50 I I I I I I I Informal I 2.19 I 2.65 I I I I Analysis of Variance Source DF MS F Prob. Status (8) 1 0.40 0.52 .478 Honesty (H) l 2.30 2.96 .095 S X H 1 0.03 0.03 .857 Error 34 0.78 247 APPENDIX X ANOVAs with Purpose of Community Service Scale Table X-l ANOVA of the Accountability Scale Purpose of Community Service Condition Reparation Other I I Experimental I 4.79 I 5.18 I I I I Control I 4.87 I 4.89 I I HHHHHH Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.22 0.78 .379 Purpose (P) 1 0.70 2.54 .115 C X P 1 0.61 2.22 .141 Error 71 0.28 248 Table X-2 ANOVA of the Attitudes Toward Community Service Scale Purpose of Community Service Condition Reparation Other I I Experimental I 5.46 I 5.19 I I I I Control I 4.90 I 4.86 I I HHHl—(HH Analysis of Variance Source DF MS F Prob. Condition (C) 1 3.18 2.30 .134 Purpose (P) 1 0.41 0.30 .587 C X P 1 0.21 0.15 .701 Error 64 1.38 Table X-3 ANOVA of the Attitudes Toward 249 the Victim Scale Purpose of Community Service Condition Reparation Other I I I Experimental I 4.51 I 4.65 I I I I I I I Control I 4.57 I 4.68 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.04 0.03 .868 Purpose (P) 1 0.28 0.20 .658 C X P 1 0.01 0.01 .944 Error 72 1.39 Table X-4 250 ANOVA of the Self-Esteem Scale Purpose of Community Service Condition Reparation Other I I I Experimental I 0.67 I 0.75 I I I I I I I Control I 0.69 I 0.76 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.00 0.04 .834 Purpose (P) 1 0.10 1.76 .189 C X P 1 0.00 0.01 .975 Error 75 0.06 Table X-5 251 ANOVA of the Delinquent Association Scale Purpose of Community Service Condition Reparation Other I I I Experimental I 2.78 I 3.20 I I I I I I I Control I 3.00 I 2.88 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.05 0.13 .724 Purpose (P) 1 0.36 0.96 .331 C X P 1 1.36 3.57 .063 Error 75 0.38 252 Table X-6 ANOVA of the Normative Pressure Scale Purpose of Community Service Condition Reparation Other I I Experimental I 2.25 I 2.19 I I I I Control I 1.97 I 2.11 I I Analysis of Variance HHHHHH Source DF MS F Prob. Condition (C) 1 0.62 1.79 .186 Purpose (P) 1 0.04 0.11 .743 C X P 1 0.19 0.54 .464 Error 75 0.35 253 Table X-7 ANOVA of the Normative Values Scale Purpose of Community Service Condition Reparation Other I I I Experimental I 1.18 I 1.05 I I I I I I I Control I 1.27 I 1.18 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.21 1.04 .312 Purpose (P) 1 0.21 1.05 .310 C X P 1 0.01 0.05 .823 Error 71 0.20 254 Table X-8 ANOVA of the General Bondinngcale Purpose of Community Service Condition Reparation Other I 1 Experimental I 2.98 I 3.03 I I I I Control I 3.03 I 2.92 I I Analysis of Variance HHHHHH Source DF MS F Prob. Condition (C) 1 0.02 0.12 .736 Purpose (P) 1 0.02 0.12 .727 C X P 1 0.13 0.68 .412 Error 75 0.19 Table X-9 255 ANOVA of the Specific Normlessness Scale Purpose of Community Service Condition Reparation Other I I Experimental 1 3.71 I 3.85 I I I I Control I 3.55 I 3.89 I I Analysis of Variance HHHHHH Source DF MS F Prob. Condition (C) 1 0.07 0.17 .680 Purpose (P) 1 1.10 2.86 .095 C X P 1 0.20 0.53 .468 Error 75 0.38 256 Table X—10 ANOVA of the Understanding7Scale Purpose of Community Service Condition Reparation Other I I Experimental I 3.74 I 3.93 I I Analysis of Variance HHH Source DF MS F Prob. Purpose 1 0.34 1.08 .305 Error 36 0.32 257 Table X-ll ANOVA of the Effort Scale Purpose of Community Service Condition Reparation Other I I I Experimental I 2.12 I 2.20 I I I I Analysis of Variance Source DF MS F Prob. Purpose 1 0.06 0.09 .770 Error 36 0.68 Table X-lZ 258 ANOVA of the Benefit Scale Condition Purpose of Community Service Reparation Other I I I Experimental I 2.35 I 2.13 I I I I Analysis of Variance Source DF MS F Prob. Purpose 1 0.42 0 51 .479 Error 36 0.82 259 Table X-13 ANOVA of Recidivism Purpose of Community Service Condition Reparation Other I I 1 Experimental I 0.36 I 0.54 I I I I I l I Control I 0.13 I 0.17 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.75 2.65 .108 Purpose (P) 1 0.21 0.73 .395 C X P 1 0.24 0.86 .357 Error 69 0.28 260 APPENDIX Y ANOVAs with Purpose of Community Service Scale - Experimentals Table Y-1 ANOVA of the Self-esteem Scale Purpose of Community Service Condition Reparation Other H H H Experimental I 0.67 I 0.75 I I I I Analysis of Variance Source DF MS F Prob. Purpose 1 0.05 0.78 .383 Error 36 0.07 Table Y-2 ANOVA of the Delinquent Associations Scale Purpose of Community Service Condition Reparation Other I I I Experimental I 2.78 I 3.20 I I I I Analysis of Variance Source DF MS F Prob. Purpose 1 1.58 3.31 .077 Error 36 0.48 261 Table Y-3 ANOVA of the Normative Pressure Scale Purpose of Community Service Condition Reparation Other I I I Experimental I 2.25 I 2.19 I I I I Analysis of Variance Source DF MS F Prob. Purpose 1 0.03 0.10 .756 Error 36 0.32 Table Y-4 ANOVA of the Normative Values Scale Purpose of Community Service Condition Reparation Other H H H Experimental I 1.18 I 1.04 H Analysis of Variance Source DF MS F Prob. Purpose 1 0.17 1.00 .325 Error 35 0.17 262 Table Y-5 ANOVA of the General BondigggScale Purpose of Community Service Condition Reparation Other I I 1 Experimental I 2.98 I 3.03 I I I I Analysis of Variance Source DF MS F Prob. Purpose 1 0.02 0.10 .759 Error 36 0.24 Table Y-6 ANOVA of the Specific Bonding Scale Purpose of Community Service Condition Reparation Other I l 1 Experimental I 3.71 I 3.85 I I I 1 Analysis of Variance Source DF MS F Prob. Purpose 1 0.16 0.48 .494 Error 36 0.34 ‘ 263 Table Y-7 ANOVA of the Recidivism Scale Purpose of Community Service Condition Reparation Other I I 1 Experimental I 0.36 I 0.52 I I I I Analysis of Variance Source DF MS F Prob. Purpose 1 0.25 0.41 .529 Error 33 0.62 264 APPENDIX 2 ANOVAs with Accountability Scale Table Z-l ANOVA of the Attitude Toward Community Service Scale Accountability Condition Low High I I Experimental I 5.11 I 5.55 I I I I Control I 4.69 I 5.12 I I Analysis of Variance HHHHHH Source DF MS F Prob. Condition (C) 1 3.00 2.24 .139 Accountability (A) 1 3.25 2.43 .124 C X A 1 0.00 0.00 .986 Error 64 1.34 265 Table 2-2 ANOVA of the Attitude Toward Victim Scale Accountability Condition Low High I I I Experimental I 4.61 I 4.57 I I I I I I I Control I 4.41 I 4.82 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.00 0.00 .980 Accountability (A) 1 0.68 0.50 .484 C X A 1 0.94 0.68 .412 Error 72 1.37 Table 2-3 266 ANOVA of the Self-esteem Scale Accountability Condition Low High I I I Experimental I 0.71 I 0.69 I I I I I I I Control I 0.70 I 0.76 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.01 0.23 .634 Accountability (A) 1 0.01 0.15 .701 C X A 1 0.04 0.15 .444 Error 75 0.06 267 Table Z-4 ANOVA of the Delinquent Associations Scale Accountability Condition Low High I I I Experimental I 2.73 I 3.22 I I I I I I I Control I 2.90 I 2.95 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.03 0.07 .787 Accountability (A) 1 1.32 3.55 .064 C X A 1 0.92 2.46 .121 Error 75 0.37 Table Z-5 268 ANOVA of the Normative Pressure Scale Accountability Condition Low High I I Experimental I 2.10 I 2.39 I I I I Control I 2.02 I 2.09 I I Analysis of Variance HHHHHH Source DF MS F Prob. Condition (C) 1 0.66 1.95 .166 Accountability (A) 1 0.63 1.85 .178 C X A 1 0 24 0.72 .399 Error 75 0.34 Table Z-6 ANOVA of the Normative Values Scale 269 Accountability Condition Low High I I 1 Experimental I 1.03 I 1.25 I I I I I I I Control I 1.16 I 1.27 I I I I Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.10 0.50 .481 Accountability (A) 1 0.48 2.41 .125 C X A 1 0.05 0.27 .603 Error 71 0.19 270 Table Z-7 ANOVA of the General Bonding_Scale Accountability Condition Low High I I Experimental I 2.99 I 3.02 I I I I Control I 2.94 I 2.98 I I HHHHHH Analysis of Variance Source DF MS F Prob. Condition (C) 1 0.04 0.19 .661 Accountability (A) 1 0.02 0.09 .768 C X A 1 0.00 0.00 .984 Error 75 0.19 Table 2-8 ANOVA of the Recidivism Scale 271 Accountability Condition Low High 1 I Experimental I 0.30 I 0.60 I I I I Control I 0.28 I 0.05 I I Analysis of Variance HHHHHH Source DF MS F Prob. Condition (C) 1 1.34 3.52 .065 Accountability (A) 1 0.01 0.03 .875 C X A 1 1.25 3.28 .074 Error 69 0.38 272 Table Z-9 ANOVA of the Understanding Scale Accountability Condition Low High I I I Experimental I 3.71 I 3.94 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 0.48 1.55 .221 Error 36 0.31 Table Z-10 ANOVA of the Effort Scale Accountability Condition Low High I I I Experimental I 2.05 I 2.28 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 0.50 0.76 .389 Error 36 0.66 k 273 Table Z-Il ANOVA of the Benefit Scale Accountability Condition Low High 1 I I Experimental I 2.33 I 2.18 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 0.23 0.28 .599 Error 36 0.82 Table Z-12 ANOVA of the Attitude Toward Community Service Scale Accountability Condition Low High I I I Experimental I 5.11 I 5.55 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 1.60 1.26 .270 Error 31 1.27 _ 274 Table Z-13 ANOVA of the Attitude Toward Victim Scale Accountability Condition Low High I I I Experimental I 4.61 I 4.57 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 0.02 0.01 .912 Error 35 1.21 Table Z-14 ANOVA of the Self-esteem Scale Accountability Condition Low High I I I Experimental I 0.71 I 0.69 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 0.01 0.07 .789 Error 36 0.07 {m '31 I 275 Table Z-15 ANOVA of the Normative Pressure Scale Accountability Condition Low High I I I Experimental I 2.10 I 2.39 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 0.82 2.75 .106 Error 36 0.30 Table Z-16 ANOVA of the Normative Values Scale Accountability Condition Low High I I I Experimental I 1.03 I 1.24 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 0.37 2.30 .138 Error 35 0.16 Table Z-17 276 ANOVA of the General Bonding;§cale Accountability Condition Low High I I I Experimental I 2.99 I 3.02 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 0 01 0 03 .867 Error 36 0.25 Table 2-18 ANOVA of the Specific Bonding Scale Accountability Condition Low High I I I Experimental I 3.67 I 3.89 I I I I Analysis of Variance Source DF MS F Prob. Accountability 1 0.46 1 39 .246 Error 36 0.33 ¥ g=eeefigi LIST OF REFERENCES LIST OF REFERENCES Accountability board stresses restitution. (1977). Criminal Justice Newsletter, 8(7), 5. Akers, R. L. (1977). Deviant behaviors: A social learning perspective. Belmont, CA: Wadsworth. Akers, R. L., Krohn, M. D., Lanza-Kaduce, L., & Radosevich, M. (1979). Social learning and deviant behavior: A specific test of a general theory. American Sociological Review, 41, 636-655. Aldenderfer, M. S., & Blashfield, R. K. (1984). Cluster analysis. Beverly Hills: Sage Publications. Altschuler, D. M. & Armstrong, T. L. (1989, March). Intensive community based aftercare program for chronic serious juvenile offenders. Unpublished manuscript. Amdur, R. & Herman, S. E. (1988). Cluster analysis. In S.E. Herman, R. Amdur, K. Hazel, S. Cohen, P. Blondin, & C. T. Mowbray (Eds.). Clients with serious mental illness: Characteristics and typology (pp. 6-11 to 6— 30). Lansing: Michigan Department of Mental Health. Armstrong, T. (1983). An overview of practices and approaches in reparative justice. In T. Armstrong, M. Hofford, D. Maloney, C. Remington, & D. Steenson (Eds.), Restitution: Agguidebook for juvenile justice practitioners (p. 1-15). Reno, NV: National Council of Juvenile and Family Court Judges. Armstrong, T., Hofford, M., Maloney, D., Remington, C., & Steenson, D. (Eds.). (1983). Restitution: A guidebook formjuvenile justice prmgtitioners. Reno, NV: National Council of Juvenile and Family Court Judges. Bachman, J. G., Kahn, R. L., Mednick, M. T., Davidson, T. N., & Johnston, L. (1967). Youth in transition. Volume 1: Blueprint for a longitudinal study of adolescent boys. Ann Arbor, MI: Institute for Social Research. 277 278 Balivet, T., Briggs, C., Chadwick, C., Conti, S., Helms, J., Hoffman, R., Moore, L., Stathakis, J., & Steelman, D. (1975). Recommendations for improving the use of 3 restitution as a dispositional alternative, as administered by the Connecticut adult probation division. Williamsburg, VA: National Center for State Courts. Barnett, R. E. (1979). The justice of restitution. The American Journal of Jurisprudence, 21, 117-132. Barnett, R. E., & Hagel, J. III (1977). Assessing the criminal: Restitution, retribution, and the legal process. In R. E. Barnett & J. Hagel, III (Eds.), Assessing the criminal: Restitution, retribution, and the legal process (p. 1-31). Cambridge, MA: Ballinger Publishing Company. Beck-Zierdt, N., Shattuck, S., Ascher, J., Chesney, S., & Jaede, A. (1982). Repairingmthe dammge: A juvenile restitution guide. St. Paul: Criminal Justice Program, Office of Local Government, DEPD. Bonta, J. L., Boyle, J. Motiuk L. L., & Sonnichsen, P. (1983). Restitution in correctional half-way houses: Victim satisfaction, attitudes, and recidivism. Canadian Journal of Criminology, 25, 277-293. Bray, J. H. & Maxwell, 8. E. (1985). Multivariate analysis of variance. Beverly Hills: Sage Publications. Briar, S., & Piliavin, I. (1965). Delinquency, situational inducements, and commitment to conformity. Social Problems, l3, 35-45. Brown, B. (1977). Community service as a condition of probation. Federal Probation, 41(4), 7-9. Brown E. J. (1983). The correlates and consequences of the payment of regtitution. Unpublished doctoral dissertation. State University of New York at Albany, Albany. Cannon, A., & Stanford, R. M. (1981). Evaluation of the juvenile alternative services project. Florida Department of Health and Rehabilitative Services. Casson, J. W., III (1983). Restitution: An economically and socially desirable approach to sentencing. New England Journal on Criminal and Civil Confinement, g, 349-385. 279 Challeen, D. A., & Heinlen, J. H. (1978). The Win-onus restitution program. InaB. Galaway & J. Hudson (Eds.), Offender restitution in theory and action (p. 151-159). Lexington, MA: Lexington Books. Chesney, S. L. (1976). Summary report: The assessment of restitution in the Minnesota Probation Services. Minnesota Department of Corrections. Chesney, S., Hudson, J., & McLagen, J. (1978). A new look at restitution: Recent legislation, programs, and research. Judicature, 6;, 348-357. Cloward, R., & Ohlin, L. E. (1960). Delinquency and opportunity: A theory of delinquent gangs. Glencoe, IL: Free Press. Cook, T. J., & Campbell, D. T. (1979). Quasi- experimentation. Chicago: Rand-McNally. Cooper, G., & West, A. S. (1981a). An evaluation of the community service restitution program: A cluster analysis. Denver: University of Denver. Cooper, G., & West, A. S. (1981b). An evaluation of the community service restitution program: A cluster analysi§;,Executive Smmmary. Denver: University of Denver. ' Criminal Justice ReSearch Center (1982). Expansion of criminal restitution plans possible, BJS says. Trial, .1_8.(1)a 14° Crotty, J. & Meier, R. D. (1980). Evaluation ofgjuvenile restitution progmam project: Detour. Final report. East Lyme, CT: Behavioral Systems Associates, Inc. Crowne, D. P. & Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 21, 349-354. Davidson, W. S., & Johnson, C. D. (in press). Diversion in Michigan. In L. Olsen-Fulero & S. Fulero (Eds.). Advancegmin law and child development. Greenwich, CT; JAI Press. Dinitz, S., Scarpitti, F. R., & Reckless, W. C. (1962). Delinquency vulnerability: Across group and longitudinal analyses. American Sociological Review, 2.7.. 515-517. 280 Eglash, A. (1977). Beyond restitution - creative restitution. In J. Hudson & B. Galaway (Eds), Restitution in criminal justice: A critical assessment of sanctions (pp.91-99). Lexington, MA: Lexington Books. Elliott, D. S., Ageton, S. S., & Canter, R. J. (1979). An integrated theoretical perspective on delinquent behavior. Journal of Research in Crime and Delinquency, m, 3-27. Elliott, D. S., Ageton, S. S., Hunter, M., & Knowles, B. (1975). Research handbook for community plmnning:and feedback instruments: Volume 1. Boulder: Behavioral Research and Evaluation Corporation. Elliott, D. S., Huizinga, D., & Ageton, S. S. (1985). Explaining7delinquency and drug use. Beverly Hills: Sage Publications. Elliott, D. S., & Voss, H. (1974). Delinquency and dropout. Lexington, MA: D. C. Heath. - Evans, R. C., & Koederitz, G. D. (1983). The requirement of restitution for juvenile offenders: An alternative disposition. Journal of Offender Counseling! Services & Rehabilitation, 1(3/4),1-20. Expansion of victim restitution programs recommended by BJS. (1981). Juvenile Justice Digest, 9(13), 1,6-7. Fagan, J. (1988, November). Correctional treatment and reintegration of violent offenders: Experimental results and theoretical implications. Paper presented at the meeting of the American Society of Criminology, Chicago, IL. Galaway, B. (1977a). Toward the rational development of restitution programming. In J. Hudson & B. Galaway (Eds.), Restitution in criminal justice: A critical assessment of sanctions (pp. 77-89). Lexington, MA: Lexington Books. Galaway, B. (1977b). The use of restitution. Crimg_and Delinquency, (Jan) 57-67. Galaway, B. (1979). Making the offender pay: Revitalizing an ancient idea. In R. G. Iacovetta & D. H. Chang (Eds.), Critical issues in crimmnal justice (pp. 352-369). Durham, NC: Carolina Academic Press 281 Geis, G. (1977). Restitution by criminal offenders: A summary and overview. In J. Hudson & B. Galaway (Eds.), Restitution in criminal justice: A critical assessment of sanctions (pp. 147-164). Lexington, MA: Lexington Books. Gendreau, P. & Ross, R. R. (1987). Revivification of rehabilitation: Evidence from the 1980’s. Justice Quarterly, 1, 349-407. Gilbeau, D. Hofford, M., Maloney, D., Remington, C., & Steenson, D. (1980). Accountability: The real issue in juvenile restitution prggrammimg. The Jameson Group, Inc. Growing pains cited in juvenile restitution expansion (1985). Criminal Justice Newsletter, 16(19), 4-5. Harding, J. (1982). Victims and offenders: Needs and responsibilities. London: Bedford Square Press of the National Council for Voluntary Organizations. Harland, A. T. (1978). Theoretical and programmatic concerns in restitution: An integration. In B. Galaway & J. Hudson (Eds.), Offender restitution in theory and action (pp. 193-202). Lexington, MA: Lexington Books. Harland, A. T., Warren, M. Q., & Brown, E. J. (1979). Evaluation objectives, evaluation methodology and action research report. Washington, DC: Criminal Justice Research Center. Harris, M. K., Carleton, G. A., & Siebens, R. A. (1979). Community service by offenders. Washington, DC: United States Department of Justice. Heide, K. M. (1983). An empirical assessment of the value of utilizing personality data in restitution outcome prediction. In W. S. Laufer & J. M. Day (Eds.), Personality theory, mpral development, and criminal behavior. Lexington, MA: Lexington Books. Heinz, J., Galaway, B., & Hudson, J. (1976). Restitution or parole: A follow-up study of adult offenders. Social Service Review, March, 148-156. Hepburn, J. R. (1977). Testing alternative models of delinquency causation. The Journal of Criminal Law and Criminology, 61, 450-460. Hirschi, T. (1969). Causes of delinquency. Berkeley, CA: University of California Press. 282 Hofford, M. (1981). Juvenile restitution program - Final report. Charleston, SC: United Way. Hofford, M. (1983). Essential principles. In T. Armstrong, M. Hofford, D. Maloney, C. Remington, & D. Steenson (Eds.), Restitution: A guidebook for juvenile justice practitioners (p. 42-52). Reno, NV: National Council of Juvenile and Family Court Judges. Hudson, J. & Chesney, S. (1978). Research on restitution: A review and assessment. In B. Galaway & J. Hudson (Eds.), Offender restitution in theory and action (pp. 131-148). Lexington, MA: Lexington Books. Hudson, J. & Galaway, B. (1977). Introduction. In J. Hudson, & B. Galaway (Eds.), Restitution in criminal justice: A critical assessment of sanctions (p. 1-17). Lexington, MA: Lexington Books. Hudson, J. & Galaway, B. (1978). Introduction. In B. Galaway & J. Hudson (Eds.), Offender restitution in theory and action (pp. 1-11). Lexington, MA: Lexington Books. Hudson, J. & Galaway, B. (1980). A review of restitution and community-service sanctioning research. In J. Hudson, & B. Galaway (Eds.), Victims, offenderg, and alternative sanctions (p. 173-194). Lexington, MA: Lexington Books. Hudson, J. Galaway, B., & Chesney, S. (1977). When criminals repay their victims: A survey of restitution programs. Judicature, pp, 312-321. Hudson, J., Galaway, B., & Novack, S. (1980). National assessment of adult restitution progpams: Final report. Duluth, MN: University of Minnesota. Hunter, J. E., & Hamilton, M. A. (1986). PATH: A program in BASICA. East Lansing: Michigan State University. Jackson, D. N. (1970). A sequential system for personality scale development. In C. Spielberger (Ed.), Current topics in clinical and community psychology_(Vol. 2) (pp. 62-96). New York: Academic Press. Jacob, B. (1977). The concept of restitution: An historical overview. In J. Hudson & B. Galaway (Eds.). Restitution in criminal justice: A critical assessment of sanctions (pp. 45-62). Lexington, MA: Lexington Books. 283 Juvenile crime rate down sharply, stiff sentences and restitution are seen as major factors. (1978). Juvenile Justice Digest, 6(20), 7-8. Kansas restitution program termed ’A character builder’. (1975). Juvenile Justice Digest, 3(16), 4. Keldgord, R. (1978). Community restitution comes to Arizona. In B. Galaway & J. Hudson (Eds.), Offender restitution in theory,and action (pp. 161-166). Lexington, MA: Lexington Books. Keve, P. W. (1978). The therapeutic uses of restitution. In B. Galaway & J. Hudson (Eds.), Offender regtitution in theory and action (pp. 59-64). Lexington, MA: Lexington Books. Klein, A. R. (19 ). Koch, J. R. (1985). Community service and omtright release as alternatives to juvenile court: An experimental evaluation. Unpublished doctoral dissertation, Michigan State University, East Lansing. Kraemer, H. C., & Thiemann, S. (1987). Howfimany subjects? Statistical power analysis in research. Newbury Park: Sage Publications. Lajeunesse, T. (1979). The Essex County pilot diversion project: Executive summary. In Innovative approaches to jpvenile justice: Executive summaries of four juvenile diversiongproject reportg. Ottawa: Solicitor General Canada. LEAA’s JD office to promote $30 million to support juvenile restitution programs. (1978). Juvenile Justice Digest, §(5), 2-30 Maori, A. (1978). Off days sentencing program. In B. Galaway & J. Hudson (Eds.), Offender restitution in theory and action (pp. 167-170). Lexington, MA: Lexington Books. Maloney, D., Gilbeau, D., Hofford, M., Remington, C., & Steenson, D. (1982). Juvenile restitution: Combining common sense and solid research to build an effective program (Part I). New Designs for Youth Development, §_(3), 3-80 Matthews, W. G. (1981). Restitution: The chameleon of corrections. Journal of Offender Counselimg, Services and Rehabilitation, 5(3/4), 77-92. 284 McAnany, P. D. (1978). Restitution as idea and practice: The retributive process. In B. Galaway & J. Hudson (Eds.), Offender restitution in theory and action (pp. 15-31). Lexington, MA: Lexington Books. McDonald, W. F. (1978). Expanding the victim’s role in the disposition decision: Reform in search of a rationale. In B. Galaway & J. Hudson (Eds.), Offender restitution in theory and action (Pp. 101-109). Lexington, MA: Lexington Books. McGregor, J. (1978). Restitution shows success. Corrections Perspective, 1(1), 8. Miller, T. I. (1981). Consequences of restitution. Law and Human Behavior, g, 1-17. Nelson, T. (1978). Postincarceration restitution. In B. Galaway & J. Hudson (Eds.), Offender restitution in theory and action (pp. 185-189). Lexington, MA: Lexington Books. Newton, A. (1979). Sentencing to community service and restitution. Criminal Justice Abstracts, ll, 435-468. Norusis, M. J. (1985). SPSSx advanced statisticsgguide. Chicago: SPSS Inc. Novack, S., Galaway, B., & Hudson, J. (1980). Victim and offender perceptions of the fairness of restitution and community service sanctions. In J. Hudson & B. Galaway (Eds.), Victims, offenders, and alternative sanctions (pp. 63-70). Lexington, MA: Lexington Books. OJJDP initiates comprehensive program to promote use of restitution program for juvenile courts. (1985). Juvenile Justice Digest, l§(7), 1-2. OJJDP program to promote use of restitution; Bids are sought. (1984). Criminal Justice Newsletter, l§(11), 4- 5. Pease, K. (1981). Community service orders: A first decade of promise. Scotland: Howard League for Penal Reform. Pease, K., Billingham, S., & Earnshaw, I. (1977). Community service assessed in 1976 (Home office research study No. 39). London: Her Majesty’s Stationery Office. 285 Raue, C. H. (1978). Victims’ assistance program. In B. Galaway & J. Hudson (Eds.), Offender restitution in theory and action (pp. 173-177). Lexington, MA: Lexington Books. Read, B. (1977). How restitution works in Georgia. Judicature, 6Q, 322-331. Reckless, W. C., Dinitz, S., & Murray, E. (1956). Self concept as an insulator against delinquency. American Sociological Review, 2;, 744-746. Remington, C. (1979). A new slant on restitution. Youth Authority Quarterly, §g(4), 14-18. Restitution evaluation project. (1983). Criminal Justice Newsletter, 14(12), 8. Restitution sentences are gaining favor but they may "widen the net" of punishment. (1977). Criminal Justice Newsletter, §(24), 1-2. Rosenberg, M. (1965). Society and the adolescent image. Princeton: Princeton University Press. Rossi P. H., Freeman, H. E., & Wright, S. R. (1982). Evaluation: A systematic approach. Beverly Hills: Sage Publications. 1. Rubin, H. T. (1985-1986). Community service restitution by juveniles. Also in need of guidance. Juvenile and Family Court Journal, §l(1), 1-8. Scarpitti, F. R., Murray, E., Dinitz, S., & Reckless, W. C. (1960). The "good" boy in a high delinquency area: Four years later. American Sociologicgl Review, 25, 555-558. Schafer, S. (1960). Restitution to victims of crimes. London: Stevens & Sons Limited. Schafer, S. (1970). Compensation and restitution to victims of crimes. Montclair, NJ: Patterson Smith. Schafer S. (1975). The restitutive concept of punishment. In J. Hudson & B. Galaway (Eds.), Considering_the victim: Readipgs in restitution and victim compensation (p. 102-115). Springfield, IL: Charles C. Thomas Publisher. Schmitt, G. R. (1985). Alternative sentencing: A proposed state model. Journal of Legislation, 12, 225-242. 286 Schneider, A. L. (1985). Guide to juvenile restitution. Washington, DC: Office of Juvenile Justice and Delinquency Prevention. Schneider, A. L. (1986). Restitution and recidivism rates of juvenile offenders: Results from four experimental studies. Criminology, 21, 533-552. Schneider, A. L., & Schneider, P. R. (1985). The impact of restitution on recidivism of juvenile offenders: An experiment in Clayton County, Georgia. Criminal Justice Review, 10(1), 1-10. Schneider, P. R. (1983). Impact of organizational characteristics of restitution progpams on short-term performance indicators. Eugene, OR: Institute of Policy Analysis. Schneider, P. R., Schneider, A. L., Griffith, W. R., & Wilson, M. J. (1982). Two-year report on the national evaluation of the juvenile restitution initiative: An overview of program performance. Eugene, OR: Institute of Policy Analysis. Schneider, P. R., Schneider, A. L., Reiter, P. D., & Cleary, C. M. (1977). Restitution requirements for juvenile offenders: A survey of the practices in American juvenile courts. Juvenile Justice, 28(4), 43-56. Sechrest, L. & Yeaton, W. E. (1981). Assessing the effectiveness of social programs: Methodological and conceptual issues. New Directions for Program Evaluation, g, 41-56. Shichor, D., & Binder, A. (1982). Community restitution for juveniles: An approach and preliminary evaluation. Criminal Justice Review, 1(2), 46-50. Siegel, L. (1979). Court ordered victim-restitution: An overview of theory and action. New Eggland Journal of Prison Law, 5, 135-150. Simons, R. L., Miller, M. G., & Aigner, S. M. (1980). Contemporary theories of deviance and female delinquency: An empirical test. Journal of Research in Crime and Delinquency, 11, 42-57. Smith, K. D. (1977). Implementing restitution within a penal setting: The case for the self-determinate sentence. In J. Hudson & B. Galaway (Eds.). Restitution in criminal justice: A critical assessment of sanctions (pp. 131- 146). Lexington, MA: Lexington Books. 287 SPSS Inc. (1986). SPSSx user’s guide (2nd. ed.). New York: McGraw-Hill Book Company. Staples, W. G. (1986). Restitution as a sanction in juvenile court. Crime and Delinquency, 22, 177-185. Steenson, D. (1983). Why reparative justice - It just makes good sense. In T. Armstrong, M. Hofford, D. Maloney, C. Remington, & D. Steenson (Eds.), Restitution: A gpide forpjuvenile justicegpractitioners (p. 16-29), Reno, NV: National Council of Juvenile and Family Court Judges. Steggerda, R. O., & Dolphin, S. P. (1975). Victim restitution: An assessment of the restitution in probation experiment operated by the Fifth Judicial District Department of Court Services - Polk Countyp Iowa. Polk County, IA: Polk County Department of Program Evaluation. Stookey, J. A. (1977). The victim’s perspective on American criminal justice. In J. Hudson & B. Galaway (Eds.). Restitution in criminal justice: A critical assessment of sanctions (pp. 19-25). Lexington, MA: Lexington Books. Sutherland, E. H., & Cressey, D. R. (1978). Criminology (4th ed.). Philadelphia: J. B. Lippincott Company. Thornberry, T. P. (1987). Toward an interactional theory of delinquency. Criminolqu, 25, 863-891. Thorvaldson, S. A. (1978). The effects of community service on the attitudes of offenders. Unpublished doctoral dissertation, University of Cambridge. Thorvaldson, S. A. (1980a). Does community service affect offender’s attitudes? In J. Hudson & B. Galaway (Eds.), Victims! offendergj and alternative sanctions (p. 71- 80). Lexington, MA: Lexington Books. Thorvaldson, S. A. (1980b). Toward the definition of the reparative aim. In J. Hudson & B. Galaway (Eds.), Victimspgoffenders, andpglternative sanctions (p. 15- 28). Lexington, MA: Lexington Books. Tittle, C. R. (1978). Restitution and deterrence: An evaluation of compatibility. In B. Galaway & J. Hudson (Eds.). Offender restitution in theorygand action (pp. 33-58). Lexington, MA: Lexington Books. 288 Tornatsky, L. G. & Johnson, E. C. (1982). Research on implementation: Implications for evaluation practice and evaluation policy. Evaluation and Program Planning, 5, 193-198. Utne, M. K., & Hatfield, E. (1978). Equity theory and restitution programming. In B. Galaway & J. Hudson (Eds.). Offender restitution in theory and action (pp. 73-87). Lexington, MA: Lexington Books. Van Voorhis, P. (1983). Theoretical perspectives on moral development and restitution. In W. S. Laufer & J. M. Day (Eds.), Personality_theory, moral developmentipand criminal behavior. Lexington, MA: Lexington Books. Van Voorhis, P. (1985). Restitution outcome and probationers’ assessments of restitution: The effects of moral development. Criminal Justice and Behavior, 12, 259-287. Viano, E. C. (1978). Victims, offenders, and the criminal justice system: Is restitution an answer? In B. Galaway & J. Hudson (Eds.). Offender restitution in theorypand action (PP- 91-99). Lexington, MA: Lexington Books. Wax, M. (1977). The effects of symbolic restitution and presence of victim on delinquent shoplifters. Ph.D. dissertations. Washington State University. Wilson, M. J. (1982). Restitution compliance and in-program reoffense rates: A comparison of experimental and control group performance in Ventura CountyL California. Unpublished report. Wilson, M. J. (1983). The juvenile offender instrument: Administration and a description of finding . Eugene, OR: Institute of Policy Analysis. Yeaton, W. H. & Sechrest, L. (1986). Use and misuse of no- difference findings in eliminating threats to validity. Evaluation Review, lg, 836-852. “11111111111111