First Advisor

Nancy A. Perrin

Term of Graduation

Spring 1992

Date of Publication


Document Type


Degree Name

Master of Science (M.S.) in Psychology






Pre-trial release, Prediction of criminal behavior-- Mathematical models



Physical Description

1 online resource (3, vii, 105 pages)


The problem of jail overcrowding has forced corrections officials and jail administrators to examine ways in which to better manage available jail space. Pretrial release and detention policies have been a target of this examination as pretrial defendants typically account for 50% of a jail's population. Standards for pretrial release exist, but their administration varies by jurisdiction. The impact of jail overcrowding on pretrial release policies has been to decrease the time available to render a decision. Recent efforts to standardize pretrial release standards in Oregon have not addressed the issue of expediency. The current study examines pretrial misconduct (failure to appear in court and rearrest) with regard to information that is available to jail personnel and release office personnel at the time of arrest, with the specific intent to develop a predictive model of pretrial misconduct that will function as an initial risk assessment.

Six hundred defendants arrested in Washington County, Oregon during 1991 served as subjects. The results indicated that 90.9% of all defendants arrested are released pending trial, and that 22.7% of those released engaged in pretrial misconduct. The results of the loglinear model-building indicated that the variables prior failure-to-appears, employment, and age were the best predictors of pretrial misconduct. The construction sample (n = 395) accurately predicted 94.5% of the observed pretrial misconduct compared to 90.7% for the validation sample (n = 150). The loglinear analysis yielded 16 typologies (based on the variables included in the model) by which defendants could be ranked as to their risk of pretrial misconduct. Spearman Rank Order coefficents for the construction and validation samples were .847 and .626 respectively. Data were also collected on detained subjects. A Chi-Square test using detained with released subjects by typology indicated that the categories are not independent (p < .01). Further examination indicated that the detained subjects did represent higher risks of pretrial misconduct as estimated by the typologies. The results also indicated that defendants currently on probation or parole were more likely to detained than other defendants.

The results do not reject the assumptions by Sturz (1962), whose Manhattan Bail Project is the basis for pretrial release, that persons with strong ties to the community may pose the least risk of pretrial misconduct. The results also found sex and ethnic differences with regard to pretrial misconduct. The sex differences may have been confounded by age and crime type; however, the ethnic differences may reflect a systemic inability to communicate with Hispanic offenders.


In Copyright. URI: This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).


If you are the rightful copyright holder of this dissertation or thesis and wish to have it removed from the Open Access Collection, please submit a request to and include clear identification of the work, preferably with URL.

Persistent Identifier