The goal of the Pre-Trial Risk Assessment Pilot Study was to improve the pre-trial case processing system by using the developed risk assessment tool to provide the Judiciary, Prosecutors Office and the Defense Bar with accurate, validated information about a defendant’s risk of violating conditions of bail or failure to appear for court. By having this information available at the time of arraignment, the parties are better able to make informed decisions about bail conditions thereby reducing the number of higher-risk defendants released while awaiting trial, and reducing the number of low-risk defendants being held in jail while on pre-trial status.
The Study included a review of previous pre-trial studies and assessments, which were used to develop a structured interview questionnaire using elements from past assessments that have been shown to be most predictive of failing to appear (FTA) and being arrested under pre-trial supervision (recidivism.)
While the findings of our study are important to review and analyze, what is of equal importance to learn from is the process by which we conducted our study. There were benefits derived from our methodology that have implications on how we conduct research in the future that are important to highlight.
Our Process:
The following outlines the process we used to develop the pre-trial risk assessment instrument, each documented component was vital to the success of our research and should be considered necessary when considering replication of this study:
Volunteers of America Northern New England hired three full time pre-trial screeners to collect the data for this study. They were supervised by the Director of Corrections, an existing staff person from Volunteers of America.
Their duties were to collect specific information, identified by the Muskie School and Volunteers of America, on every new arrestee that was brought to the Two Bridges Regional Jail. The screeners were trained by Muskie School researchers in data collection methods; the accuracy and completeness of the data was paramount to accurate findings.
A screener was on duty at the Two Bridges Regional Jail 24 hours a day, seven days a week in order to capture all of the data required.
Screeners used a structured questionnaire to gather information on a series of factors that have been shown to be predictors of FTA and new crime recidivism, including criminal history, probation history, employment, mental and medical health, and residency. Information obtained was verified, when possible, through a review of the individual’s case file information.
The interview included information normally used by pre-trial agencies, as well as new factors. More than 100 questions were included in the interview.
Data Collection:
The data collection phase was conducted over a period of six months.
The defendants who were entered into the system were then tracked over the following six months to determine if they had either violated the conditions of their release or failed to appear for their court hearing.
The defendant was tracked using county court records, reports from pre-trial supervision officers, and the state criminal history records system for whether they failed to appear at a court hearing (FTA) and/or whether any new offense occurred while they were released.
The information was then entered into a database and collected by the Muskie School on a monthly basis for analysis.
It is important to note that this is a study not of a defendant’s risk as it relates to future violent acts, but of their risk of either violating the conditions of their release or failing to appeal for their court hearing.
Data Analysis:
The follow up data was then analyzed by the Muskie School to determine which variables were the best predictors of violation or failure to appear. The risk factors were then separated from the other information collected by the screeners and developed into a pre-trial risk assessment tool to be used on new arrestees prior to arraignment to provide information to the judges, district attorneys and defense attorneys.
Once data was collected, individual items from the data collection tool that have a significant relationship with the outcome measures were identified, using chi-square correlations statistics. Items that were identified as significant were combined into a draft pre-trial risk assessment instrument.
A series of cross-tabulation analyses was then be used to create cut-off values in the distribution of risk scores of the construction sample in order to classify defendants as low, moderate and high risk for receiving an FTA or to commit a new offense while under supervision.
This second round of analyses examined the ability of the draft pre-trial assessment to distinguish among risk groups by evaluating their respective failure rates.
