Predicting Outcomes in Mediation - The Influence of People and Process (From Final Report of the Divorce Mediation Research Project, 1984 - See NCJ-98054)

• NCJ Number: 98064
• Author(s): N A Thoennes; J Pearson
• Date Published: 1984
• Length: 22 pages
• Grant Number(s): 90-CW-634
• Sponsoring Agency: US Dept of Health and Human Services
• Corporate Author: Assoc of Family and Conciliation Courts
• Publication Format: Document
• Language: English
• Country: United States of America

Annotation

This paper explores the relative importance of dispute characteristics versus perceptions of mediator skill and expertise in explaining successful resolutions in mediation.

Abstract

The 387 disputants who participated in the longitudinal survey were mediating custody or visitation problems at the Los Angeles Conciliation Court, Minnesota's Hennepin County Superior Court, or the Family Division of the Connecticut Superior Court. An initial questionnaire was administered immediately prior to the start of mediation; a second questionnaire was administered 3 months later. The questionnaires were either self-administered or conducted by phone during the fall of 1981 and early 1982. Factor analysis was used to create composite indexes from a variety of pre-existing characteristics suggested in the literature as instrumental in the outcome of mediation. Factor analysis was also employed to create indexes of mediator effectiveness as perceived by clients. Discriminant analysis using these indexes indicates that the ability to predict outcomes of mediation is limited. However, the three indexes of pre-existing characteristics that contribute most to predicting outcomes are the duration of the dispute, its intensity, and the quality of the relationship with the ex-spouse. The two indexes of mediator effectiveness that most improve predictions are the mediator's ability to promote communication and his ability to provide disputants with insights into the attitudes and behaviors of the parties involved. Finally, the indexes of mediator behaviors are as effective as the indexes of pre-existing characteristics in predicting outcomes. Two figures, 3 tables, and 34 references are included. (Modified author abstract)