Commentary: Focusing on Program Design Issues in Future Research on Court-Connected Mediation

Court-connected mediation often works very well. That is a fair con-clusion based on evidence summarized in Roselle Wissler’s meticulous review of court-connected mediation. Analyzing studies of small claims, general civil, and appellate mediation programs, her review suggests that mediation is usually evaluated very favorably and is rated as highly as or better than the alternatives on virtually all outcome indicators.1 In other words, almost all of these studies find that the results are either better in mediation or that there are no significant differences. This suggests that mediation has the potential to be quite effective in producing various desired results and that whether a mediation program actually generates such results depends on how well it is designed and fits with the local practice culture. Mediation is a highly variable process that program designers and users can readily adapt. As McEwen (1988) sug-gests, instead of “asking whether mediation works or not, we need to exam-ine how and why parties and lawyers ‘work ’ mediation in varying ways” (p. 3).2 McEwen’s suggestion can be extended to analyze how mediators and program designers “work ” mediation processes. Thus, this article ana-lyzes prior research to determine why some mediation programs did not outperform traditional litigation while others did.3 It illustrates how researchers and program designers might analyze past research findings, design programs to produce desired results, and then empirically test the effectiveness of such design efforts.4 This article also suggests that researchers increase the use of outcome measures in addition to traditional measures of efficiency, satisfaction, and perceived fairness. These include substantive justice, empowerment and recognition, and interest-based problem solving. Given intense con-cerns about mediator evaluation and party self-determination, researchers should also examine these issues further