34 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, vol. vii | 35 Program Description This evaluation focuses on the Legal Aid Society program called the Parole Revocation Defense Unit (PRDU). Specifically, the study focuses on the written mitigation drafted by social workers at PRDU and its impact on preventing parole revocation. Written mitigation is a powerful tool that details a client’s personal life and relevant history. It humanizes the client so judges can see the human being beyond the alleged crime (The Legal Aid Evaluating the Success of Written Mitigation in Reducing Prison Sentences and Achieving Alternatives to Incarceration for Parole Violations Nnenna Onyema, Jaime Buseman, and Carrie Maldovan It is estimated that 2.4 million individuals are incarcerated in federal, state, and county prisons and jails in the United States, the largest number seen in the developed world (Flatow, 2014). In addition, there are an estimated 4.75 million, or 1 in 51, adults under community supervision in the United States (Herberman & Bonczar, 2014). This includes individuals on probation, parole, and other forms of post-prison supervision. There are issues with the mass incarceration system in the United States that extend beyond the sheer number of individuals incarcerated. For instance, people of color make up 30% of the population but 60% of those imprisoned (Kerby, 2012). For the about 850,000 on parole, many face struggles adapting to society and attaining basic needs such as health care, housing, and employment. These factors, among many others, lead these individuals to violate parole and become re-imprisoned, creating a cycle of perpetual contact with the legal system. The Parole Revocation Defense Unit at the Legal Aid Society provides services for those who have had their parole revoked or face parole revocation. Social workers work together with lawyers to provide comprehensive plans to reduce sentences and recommend alternatives to incarceration. One tool, known to be powerful within PRDU at the Legal Aid Society, is written mitigation, a process used to advocate for and provide holistic presentations of clients. This paper evaluates the benefits of written mitigation by looking at the sentencing for 20 cases before and after written mitigation. The findings show that written mitigation was helpful in reducing sentences and resulting in alternatives to incarceration with statistical significance. This information learned contributes to the larger discussion of the relevance of social workers in legal settings to provide holistic services and broader conversations of criminal justice reform and elimination of institutions that produce outcomes that do more harm than good. stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text © 2016 Onyema, Buseman, and Maldovan. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits the user to copy, distribute, and transmit the work provided that the original author(s) and source are credited. stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text 36 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 37 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan Society, 2012). A major aim of written mitigation is to reduce sentences issued by judges compared with those they would have given without mitigation. The short- term objective is that clients are released from jail into the community to begin recommended treatment and programs. The mid-term objective is for clients to successfully complete parole supervision. The long-term objective of PRDU is for clients to adjust to society with more opportunities for self- sufficiency and fewer interactions with the criminal justice system. Study Aims Because written mitigation is one of the few tools PRDU social workers use to advocate for reduced sentences, it is important to explore its effectiveness. The original sentences judges intended to give are compared with their final decisions to identify any changes. The level of efficacy in reducing sentences through written mitigation can assist social workers in improving their work and achieving better outcomes for clients. Given the prison system’s extensive costs to society , it is necessary to investigate the role of prison in reducing crime and rehabilitating inmates. In a comparison of two studies on recidivism, Ostermann (2013) found that a cohort of released inmates from 1983 had a rearrest rate of 62.5%, a reconviction rate of 46.8%, and a return-to-custody rate of 41.4% compared with a cohort of released inmates from 1994, of whom 67.5% were rearrested, 46.95% were reconvicted, and 51.8% returned to custody. This study suggests recidivism and subsequent rearrest, reconviction, and return to prison appear to increase over time. Additionally, a study by Bales and Piquero (2011)—controlling for sex, race, age, current offense, and prior record—found that overall, imprisonment leads to higher recidivism when compared with a non-incarcerative alternative. Nagin (2009) also found that compared with noncustodial sanctions, incarceration appears to have a null or mildly criminogenic effect on future criminal behavior. Therefore, findings suggest that prison actually contributes to the likelihood of future criminal behavior for previously incarcerated people. Given the findings on the ineffectiveness of prisons to reduce recidivism, parole supervision has been used as another avenue to reduce future criminal behavior. Ostermann (2013) found that the predicted likelihood of recidivating differed by about 5% between supervised (parole) and unsupervised (non- parole) groups in the 6-month and 1-year periods; however, parolees who were actively supervised for at least 3 years were estimated to recidivate at an 8% lower rate. While this finding is significant, the incidence of recidivism among parolees continues to be high. stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text stujourn Typewritten Text 36 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 37 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan Based on previously incarcerated people’s high incidence of recidivism, the limited effects of parole alone, and the evidence that prison contributes to future criminal behavior, the effectiveness of the traditional criminal justice system has been called into question (Zhang, Roberts, & Callanan, 2006). Alternatives to incarceration (ATI) are one promising solution. ATI have grown in recent years in response to failures in the traditional criminal justice system. These programs, which generally include aspects such as community service, rehabilitative services, and restitution, are less costly, less disruptive to families and communities, and helpful in maintaining stability and encouraging accountability among offenders (Pew Center on the States, 2009). According to the Legal Action Center (2012), alternative programs such as drug courts reduce future crime by as much as 35%, while intense community supervision combined with rehabilitation services reduces recidivism by up to 20%. Mitigation provided by PRDU is the gatekeeper between clients and either prison sentences or alternative programs. Therefore, the aim of the evaluation study is to investigate the effectiveness of mitigation in contributing to a reduced prison sentence or the offer of alternative-program sentences. Methodology The sample consisted of 20 individuals who allegedly violated parole during the period from July 2013 to January 2014. The age and ethnicity of these individuals vary, but the individuals are mainly middle-aged men who are either African American or Latino. Their alleged crimes include both violent and nonviolent offenses. Since this evaluation used the existing agency data, there was no recruitment process and the sample was convenient. Because agency records were used and no new data were collected, the study design of this program evaluation was secondary data analysis. Additionally, this is a retrospective study that examines and further analyzes client records based on the differences in sentences administrative law judges gave clients before and after social workers provided mitigation. The methodology used in this evaluation is quantitative analysis based on the existing data about the individuals who have allegedly violated parole. The study compared the judges’ sentencing of the 20 clients before and after mitigation to determine its effect in decreasing prison/jail-time sentences or offers of ATI, such as completion of a program. In addition, this study includes a brief analysis of the potential effects of other variables—including type of offense, gender, ethnicity, mental health status, chemical addiction, or “Mitigation provided by PRDU is the gatekeeper between clients and either prison sentences or alternative programs.” 38 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 39 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan educational levels—on sentencing outcomes. In terms of data collection procedures, the program evaluation is based on client records from a 6-month period from July 2013 to January 2014, including written mitigation provided by the Legal Aid Society. Since all the client records used in the study already existed, there were neither interviews nor surveys conducted, and clients had no incentives to participate in this study. Consent was obtained from the Legal Aid Society. No personally identifiable information, including names, was used. For the measurement and data analysis plan, a paired-sample T-test was used to compare sentencing decisions given to the clients before and after written mitigations to see if there was any improvement. Using 3 months as the unit of time, there were 14 values of the dependent variables (coded as “Original Offer” and “Final Offer”) ranging from a community program sentence (best outcome) to a 36-month jail sentence (worst outcome). A P value less than .05 showed the change in sentence was statistically significant, meaning the written mitigation improved offers for the clients. Another data analysis method used was chi-square, which was used to determine any associations or correlations between the following other variables and the changing outcomes: instant offense (valued as nonviolent or violent crimes), educational levels (less than high school, high school/GED, some college, unknown) , and mental health status (presence of self-reported diagnosis or not), all of which were categorical. The changing outcomes were valued as no change of outcome, less jail time, and zero jail time. Results The findings based on the quantitative data of this program evaluation included the demographic information, the results of the paired sample T-test result explaining whether mitigation improved the outcome of the 20 clients, and the association between other variables and the final outcome. The sample contained 20 clients—17 male and 3 female. Ages ranged from 19 to 71. Half the sample—10—identified as African American, while 3 individuals identified as Latino/a, 2 as white, 3 as two or more races, and 2 individuals’ races were unknown. In terms of educational level, 2 of the clients had attended some college, while 6 had a high school diploma or GED, 9 did not graduate from high school, and 3 had an unknown education level. This information indicated that the sample had a low level of educational attainment compared with the general population. Mental illness was prevalent within the sample, with 18 of 20 clients reporting a diagnosis. Additionally, 15 out of the 20 individuals reported substance abuse or dependence. Of the sample, 11 individuals committed a violent crime. 38 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 39 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan The number of parole violations ranged from 1 to 14; however, only 2 of the individuals violated parole more than 2 times. To determine whether the mean differences were statistically significant, which would indicate that written mitigation was effective in improving a client’s offer, this evaluation used a paired sample T-test. The results of the test are shown in Appendix A. Figure 1 indicated a significant decrease in the means of time offered by the judges after the written mitigations, which decreased from 5.25 to 2.75. This reduction means that sentences decreased from an average of 10 months to an average of less than 3 months or a program. Additionally, the difference between “Original Offer” and “Final Offer” was statistically significant at a P value of .005, shown in Figure 2 and 3 in Appendix A. As a result, there was a statistically significant decrease in offer scores from “Original Offer” (M=5.25, SD=3.23) to “Final Offer” (M=2.75, SD=2.61), t(19)= 4.80, p<.0005 (two tailed). The mean decrease in offer scores was 2.50 with a 95% confidence interval ranging from 1.41 to 3.59, shown in Figure 3. Therefore, the results from the paired sample T-test revealed that the written mitigation from social workers at PRDU improved the final sentences of the clients, either lessening their jail time or getting them another program. Additionally, we performed the Chi-square test to see if there was an association between different variables and the final outcome. The variables we highlighted were educational level and instant offense. More specifically, the Chi-square test was used to determine whether violent or nonviolent offenses would be associated with an improved offer, and whether educational level of the clients would be associated with a change in the final offer. However, because the sample size was very small, the lowest frequency in some cells was less than 5, which means that an association cannot be proven using the Chi-square test. The results of the Chi-square analysis are shown in Appendix B. Discussion Statistical, Practical, and Clinical Significance The results show a statistical significance between written mitigation and improved sentences for individuals accused of violating parole. More practically, this illustrates the importance and effectiveness of advocating for vulnerable populations, such as individuals with a prior offense. Humanizing individuals involved with the criminal justice system is helpful in reducing or avoiding prison sentences and increases referrals to potentially effective programs. This is one aspect that highlights how social workers are invaluable within a legal setting and provide such unique services to clients involved 40 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 41 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan with the criminal justice system. Furthermore, the results suggest that social workers should complete mitigation for legal matters more often, as it has an effect on outcomes for clients. The results suggest many individuals who are on parole may need mental health and addiction services, as the majority of the sample was diagnosed with a mental illness and many had chemical addictions . This is an opportunity for social workers to intervene and help decrease recidivism at the clinical level. This study is also a reminder that social workers must provide holistic services and take into account environmental factors and other macro-level issues that may lead an individual to commit a crime. Limitations While the study did show statistically significant results in terms of the effectiveness of mitigation in decreasing a client’s sentence, there were significant limitations. Because of restraints on data collection, a sample of only 20 clients was obtained. This is a very small sample, especially for the use of Paired Sample T-tests which typically require a sample of 30 or more individuals . Furthermore, the sample was non-random, as clients whose information was accessible were used for the sample. Additionally, all clients in the sample received written mitigation, so there was no control group. If the study were to be replicated, a larger, random sample of the client population would be preferable. The study also did not address variations in the mitigation that may have affected the outcome, such as the proposed alternative and the particular details about the client’s life as presented by the social worker. Further analysis should examine different types of written mitigation to see which were more or less effective. Looking at variations between social workers in order to identify the most effective aspects to include in written mitigation could also prove beneficial. The greatest challenge faced was collecting the data about pre-mitigation offers from the judges, as this information was not readily available in client files. It was difficult to identify more individuals for the sample. A larger random sample would have been preferable in order to get more valid results in the study. “Humanizing individuals involved with the criminal justice system is helpful in reducing or avoiding prison sentences and increases referrals to potentially effective programs. This is one aspect that highlights how social workers are invaluable within a legal setting and provide such unique services to clients involved with the criminal justice system.” 40 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 41 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan Implications The implications of this study for practice and research are very significant. The findings show that written mitigation provided by social workers is effective in improving client outcomes, leading to either reduced jail time or an alternative to incarceration. It may be beneficial to include social workers in other departments at the Legal Aid Society and potentially in other organizations involved with the criminal justice system that do not employ social workers. As social workers, it is important for us to evaluate the programs in which clients receive treatment to confirm that they are effective. When clients are released from programs, there is rarely any follow up. The results of this study suggest that social workers should provide services not only during clients’ exposure to the criminal justice system but after program completion as well, as the unique problems that clients face do not end once a case is closed. Social workers should communicate with parole officers, family, medical staff, and any other individuals in clients’ lives to ensure the best possible outcomes are consistently achieved. While written mitigation is helpful, it does not solve the macro-level issues. Recommendations With the current statistics and literature revealing that incarceration is not cost-effective and can actually contribute to recidivism (Vera Institute of Justice, 1996, 2012), and other research showing that ATI are increasingly effective (Legal Action Center, 2012; Ostermann, 2013; Zhang, Roberts & Callanan, 2006), it is important to continue investigating the effectiveness of mitigation, especially for individuals who have violated parole. Therefore, further research should be conducted using random sampling and larger samples to further investigate the effectiveness of mitigation in increasing the use of alternatives to incarceration. Because written mitigation utilized by PRDU is an effective intervention to promote ATI for individuals who violate their parole, a suggestion is to replicate this program in the hope of increasing the use of ATI. Additionally, replication of this study within other legal settings may demonstrate a need to increase the number of programs such as PRDU that seek ATI, which will limit the number of individuals given prison sentences for violating parole or probation, as well as prevent or reduce initial prison sentences. Further studies could also provide valuable information needed to create and implement more successful programs such as PRDU and improve the effectiveness of written mitigation. While written mitigation is not unique to PRDU, studying its efficacy in other organizations can further increase awareness regarding the importance of social workers and written mitigation 42 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 43 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan in legal settings. References Bales, W. and Piquero, A. (2011). Assessing the impact of imprisonment on recidivism. Journal of Experimental Criminology, 8(1), 71-101. Flatow, N. (2014, September 17). The United States has the largest prison population in the world - And it’s growing. Retrieved from http://thinkprogress.org/justice/2014/09/17/3568232/the-united- states-had-even-more-prisoners-in-2013/ Herberman, E. J., & Bonczar, T. P. (2014, October). Probation and parole in the United States, 2013. Retrieved from http://www.bjs.gov/content/pub/pdf/ppus13.pdf Kerby, S. (2012, March 13). The top 10 most startling facts about people of color and criminal justice in the United States. Retrieved from https://www.americanprogress.org/issues/race/ news/2012/03/13/11351/the-top-10-most-startling-facts-about-people-of-color-and-criminal- justice-in-the-united-states/ Legal Action Center. (2012). Alternatives to incarceration factsheet. Retrieved from http://lac.org/wp- content/uploads/2014/07/ATI_Factsheet.pdf Legal Aid Society. (2012). The Parole Revocation Defense Unit. Retrieved from: https://www.legal-aid.org/en/ criminal/criminalpractice/parolerevocationsdefenseunit.aspx Nagin, D. (2009). Imprisonment and reoffending. Crime and Justice. 38(1), 115-200. Ostermann, M. (2013). Active supervision and its impact upon parolee recidivism rates. Crime & Delinquency. 59(4), 487-509. Pew Center on the States. (2009, March). One in 31: The long reach of American corrections. h t tp ://www.pewt r us t s.o rg/~/med i a/ l eg acy/up loaded f i l e s/pcs_a s se t s/2009/ pspp1in31reportfinalweb32609pdf Zhang, S. X., Roberts, R. E., & Callanan, V. J. (2006). Preventing parolees from returning to prison through community-based reintegration. Crime & Delinquency, 52(4), 551-571. doi:10.1177/0011128705282594 Vera Institute of Justice. (2012). The Price of Prisons: What incarceration costs taxpayers. Retrieved from http://www.vera.org/sites/default/files/resources/downloads/price-of-prisons- updated- version-021914.pdf. Vera Institute of Justice. (1996). The Unintended Consequences of Incarceration. Retrieved from http:// www.vera.org/sites/default/files/resources/downloads/uci.pdf Nnenna Onyema is a 2015 graduate of the Columbia University School of Social Work from New Jersey. Her concentration was Policy Practice, and during her time at Columbia, she was selected to be a law minor and Washington DC Fisher-Cummings fellow. After graduating from Columbia, Nnenna served as a Graduate Research Assistant at Columbia University’s School of Social Work where she examined the mental health effects that police brutality and the media exposure of such abuses have on people of color. In her current role as a Project Manager with the National Initiative for Building Community Trust and Justice, Nnenna works with executive- level leadership to oversee project implementation and provide strategic advice to National Initiative partners and pilot sites. Prior to joining the National Initiative, Nnenna worked at the Legal Aid Society connecting incarcerated individuals to alternatives to incarceration at Rikers Island. In addition, at the Vera Institute, she co-managed survey research efforts for a NIJ grant looking at the relationship and challenges between public defenders and their clients with mental illness diagnoses. Nnenna graduated from Rutgers University magna cum laude with a BA in Sociology 42 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 43 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan and minor in Business Administration. Jaime Buseman is a 2015 graduate of the Columbia University School of Social Work. Her concentration was Social Enterprise Administration with a focus in health, mental health and disabilities. Prior to attending Columbia, Jaime volunteered as a Pennsylvania Certified Sexual Assault Counselor for Victim Services Center and as a hospice volunteer for Caring Hospice Services while attending Gwynedd-Mercy University and graduating cum laude with her BA in English Literature and a minor in Human Services. During her years at Columbia, Jaime worked as a School Social Worker at Democracy Prep Charter Middle School and alongside the Community Social Work Manager at Metropolitan Jewish Health System - Hospice (MJHS). Currently, Jaime is back in her hometown in Pennsylvania and is working on the HR team at Teva Pharmaceuticals that provides human resources support to over 8,000 US and Canadian employees. Carrie Maldovan is a 2015 graduate of Columbia University School of Social Work, where she concentrated in international social enterprise administration. Prior to obtaining her Master’s degree she lived and worked in Tanzania, consulting with nonprofit organizations and leading the social impact strategy at a tourism company. Currently, she serves as the Manager of Programs and Operations at The Adventure Project, an organization working to end extreme poverty by creating jobs in developing countries. Carrie is the co-founder of a responsible and sustainable tourism company, Above Safaris, that creates jobs, supports local businesses, and channels profits into sustainable community development projects in Tanzania. 44 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 45 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan Appendix A : Paired Sample T-test Figure1: Paired Samples Statistics Mean N Std. Deviation Std. Error Mean Pair 1 Original Offer 5.25 20 3.226 .721 Final Offer 2.75 20 2.613 .584 Figure2: Paired Samples Correlations N Correlation Sig. Pair 1 Original Offer & Final Offer 20 .701 .001 Figure3: Paired Samples Test Paired Differences t df Sig. (2-tailed)Mean Std. Devia- tion Std. Error Mean 95% Confi- dence Interval of the Differ- ence Lower Upper Pair 1 Original Offer - Final Offer 2.500 2.328 .521 1.410 3.590 4.802 19 .000 44 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 45 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan Appendix B: Chi-square Tests MENTAL ILLNESS & FINAL OFFER Mental Illness * The final improvement of the outcome Cross tabulation The final improvement of the outcome Total less jail time no jail time no im- provement Mental Illness Yes Count 3 10 5 18 % within Mental Illness 16.7% 55.6% 27.8% 100.0% % within The fi- nal improvement of the outcome 100.0% 100.0% 71.4% 90.0% % of Total 15.0% 50.0% 25.0% 90.0% No Count 0 0 2 2 % within Mental Illness 0.0% 0.0% 100.0% 100.0% % within The fi- nal improvement of the outcome 0.0% 0.0% 28.6% 10.0% % of Total 0.0% 0.0% 10.0% 10.0% Total Count 3 10 7 20 % within Mental Illness 15.0% 50.0% 35.0% 100.0% % within The fi- nal improvement of the outcome 100.0% 100.0% 100.0% 100.0% % of Total 15.0% 50.0% 35.0% 100.0% 46 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 47 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan Chi-Square Tests Value df Asymp. Sig. (2-sided) Pearson Chi-Square 4.127a 2 .127 Likelihood Ratio 4.628 2 .099 Linear-by-Linear Association 2.937 1 .087 N of Valid Cases 20 a. 4 cells (66.7%) have expected count less than 5. The minimum expected count is .30. Symmetric Measures Value Approx. Sig. Nominal by Nominal Phi .454 .127 Cramer's V .454 .127 N of Valid Cases 20 46 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 47 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan EDUCATIONAL LEVEL & FINAL OFFER Education Level * The final improvement of the outcome Cross tabulation The final improvement of the outcome Total less jail time no jail time no improve- ment Edu- cation Level Less than high school Count 1 6 2 9 % within Education Level 11.1% 66.7% 22.2% 100.0% % within The final improvement of the outcome 33.3% 60.0% 28.6% 45.0% % of Total 5.0% 30.0% 10.0% 45.0% High school diploma or GED Count 0 2 4 6 % within Education Level 0.0% 33.3% 66.7% 100.0% % within The final improvement of the outcome 0.0% 20.0% 57.1% 30.0% % of Total 0.0% 10.0% 20.0% 30.0% Some college Count 0 1 1 2 % within Education Level 0.0% 50.0% 50.0% 100.0% % within The final improvement of the outcome 0.0% 10.0% 14.3% 10.0% % of Total 0.0% 5.0% 5.0% 10.0% unknown Count 2 1 0 3 % within Education Level 66.7% 33.3% 0.0% 100.0% % within The final improvement of the outcome 66.7% 10.0% 0.0% 15.0% % of Total 10.0% 5.0% 0.0% 15.0% Total Count 3 10 7 20 % within Education Level 15.0% 50.0% 35.0% 100.0% % within The final improvement of the outcome 100.0% 100.0% 100.0% 100.0% % of Total 15.0% 50.0% 35.0% 100.0% 48 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 49 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan Chi-Square Tests Value df Asymp. Sig. (2-sided) Pearson Chi-Square 10.947a 6 .090 Likelihood Ratio 10.437 6 .107 Linear-by-Linear Association 2.897 1 .089 N of Valid Cases 20 a. 12 cells (100.0%) have expected count less than 5. The minimum expected count is .30. Symmetric Measures Value Approx. Sig. Nominal by Nominal Phi .740 .090 Cramer's V .523 .090 N of Valid Cases 20 INSTANT OFFENSE & FINAL OFFER Original Felony * The final improvement of the outcome Cross tabulation The final improvement of the outcome Total less jail time no jail time no improve- ment Original Felony Violent Count 2 6 3 11 % within Original Felony 18.2% 54.5% 27.3% 100.0% % within The final improvement of the outcome 66.7% 60.0% 42.9% 55.0% % of Total 10.0% 30.0% 15.0% 55.0% Non- Violent Count 1 4 4 9 % within Original Felony 11.1% 44.4% 44.4% 100.0% % within The final improvement of the outcome 33.3% 40.0% 57.1% 45.0% % of Total 5.0% 20.0% 20.0% 45.0% Total Count 3 10 7 20 % within Original Felony 15.0% 50.0% 35.0% 100.0% % within The final improvement of the outcome 100.0% 100.0% 100.0% 100.0% % of Total 15.0% 50.0% 35.0% 100.0% 48 | Columbia SoCial Work revieW, vol. vii Columbia SoCial Work revieW, volume vii | 49 evaluaTing THe SuCCeSS of WriTTen miTigaTion onyema, buSeman, and maldovan Chi-Square Tests Value df Asymp. Sig. (2-sided) Pearson Chi-Square .683a 2 .711 Likelihood Ratio .686 2 .710 Linear-by-Linear Association .601 1 .438 N of Valid Cases 20 a. 5 cells (83.3%) have expected count less than 5. The minimum expected count is 1.35. Symmetric Measures Value Approx. Sig. Nominal by Nominal Phi .185 .711 Cramer's V .185 .711 N of Valid Cases 20