horuschan-blog1 - Tumblr blog

[horuschan-blog1]

Results

Following table shows descriptive statistics for Number of times reincarcerated and the quantitative predictors. The average times reincarcerated was 0.455 times (sd=0.70), with a minimum of 0 time and a maximum of 5 times.

In addition, Arrested for Illegal Drugs occurred in interviews occur for only 30.

Bivariate Analyses

Scatter plots for the association between number of times reincarcerated variable and quantitative predictors revealed that the more times illegal drugs used, the less number of times reincarcerated. But for the people with lower times illegal drugs used, there is now obvious correlation. For other three variables, there are no obvious trends between the number of times reincarcerated and themselves.

ANOVA results indicated that number of times reincarcerated variable differ significantly as a function of Whether Arrested for Illegal Drugs.

[horuschan-blog1]

Methods

Sample The sample included data of participants were recently released parolees with a minimum of 3 months of parole (n = 476) who were required as part of their parole to participate in a 12 weeks drug addiction treatment program. The capstone data set includes extensive interview data from screening and intake (baseline) interviews (600+ variables), and multiple primary substance use, crime, re-arrest, and re-incarceration outcomes assessed nine months after the intake interview. Measures The criminal justice recidivism risk was measured for Number of times reincarcerated from interview data. Predictors included 1) Total Number of Illegal Drugs Used (excludes tobacco and alcohol) in past 9 months, 2) Number Of Days Used Tobacco in past 9 months, 3) Number of Days Drank Alcohol in past 9 months, 4) Number Of Days Drank 5 Or More Alcohol Drinks in past 9 months and 6) Whether Arrested for Illegal Drugs(Yes or No).

[horuschan-blog1]

Data Analysis and Interpretation: Week 1 - Title, Research Question & Intro

For my capstone project, I will be participating the The Connection’s CBM program . The goal is to identify actors associated with recidivism. 

The Connection is a non-profit human service and community development agency.  One of their goals is to apply behavioral controls to develop effective treatment programs for inmates on parole, with the ultimate goal of reducing recidivism.

The title of my project is: An Algorithm for Identifying Major factors Associated with Recidivism Risk for People on Parole.

The purpose of this study is to identify the best predictors of recidivism risk. Potential predictors include Total Number of Illegal Drugs Used, Number Of Days Used Tobacco, Number of Days Drank Alcohol. I will consider which interview variables are the best predictors among others.

As a graduate student in the financial engineering area, its important to get better understanding correlations and predictors and to utilize statistics knowledge to deal with real world problems.