Interorganizational Data Analysis - A Multivariate Approach

• NCJ Number: 84208
• Author(s): J R Lincoln; J Miller
• Date Published: 1980
• Length: 33 pages
• Grant Number(s): 78-NI-AX-0135
• Sponsoring Agency: US Dept of Justice; US Dept of Justice NIJ Pub
• Sale Source: National Institute of Justice/
• Publication Type: Statistics
• Language: English
• Country: United States of America
• Note: Paper prepared for presentation at the Special National Workshop on Research Methodology and Criminal Justice Program Evaluation sponsored by the Criminal Justice Research Evaluation Program, LEAA, Baltimore, MD, March, 1980.

Annotation

Some issues and problems which arise when linear statistical models are applied to the analysis of interorganizational ties are discussed.

Abstract

Data are analyzed from an evaluation of three community-based interagency networks established under an LEAA pilot program that diverted youthful status offenders from secure detention to local treatment-oriented agencies. Surveys of agency staff and directors produced measures of agency attributes and interagency relations. Relations between organizations were measured by aggregating agency workers' reports of ties to personnel in other agencies. Two operationalizations of such agency-level ties -- total ties directed from one agency to another and a proximity measure computed on the binarized matrix of interagency ties -- yield somewhat different results. Effects of agency attributes (e.g., size and client composition) on interagency ties are evaluated. Main effects of initiating and receiving agencies' scores on the same attribute are estimated, as is the interaction of these scores. Previous pair analyses which assessed the impact of similarity and dissimilarity in the attributes of units forming pairs were sometimes subject to the same interpretation problems which plagued early studies of status inconsistency effects, i.e., no controls for the additive effects of attributes combined to form the dissimilarity index. Also explored were several procedures for evaluating the effect of reciprocity on the estimates of coefficients in models predicting pair relations from agency attributes. Tabular and graphic data and 22 references are provided. (Author summary modified)