This article reports on two studies. The analysis of the first study, a survey of 250 mediators, finds four distinct groups of mediator “clusters,” based on self‐reported strategies. These four clusters are described in detail and mediators' self‐defined labels are then correlated with the four clusters. There is little consistency between the labels mediators give their approach and the cluster into which they actually fall in this survey. The analysis of the second study, which involved observation and coding of actual mediations, finds that those mediators who were observed to use any directive strategies tended to use mostly directive strategies and those mediators who were observed to use any elicitive strategies tended to use mostly elicitive strategies throughout the observed mediation case. This challenges the notion that mediators may use both directive and elicitive strategies together in the same mediation.

This empirical study explores the effects of matching mediators and mediation participants by gender and by racial or ethnic identity group. It considers both the effect on a participant of being present in a mediation session where there is no mediator of the same gender or racial/ethnic group and the effect of being present when there is also a mediator who matches the gender or race/ethnicity of the other participant. The results show that failing to match disputants and mediators by gender has negative effects on mediation satisfaction measures and that those effects increase when the mediator's gender also matches that of the other participant. In contrast, failure to match by racial or ethnic group has little effect, but when an unmatched participant faces both an opposing participant and a mediator who share a racial or ethnic identification, mediation satisfaction decreases in several respects.

This study uses behavioral observation and pre‐and post‐mediation questionnaires to measure the impact of mediator behaviors on participant attitudes and case outcomes in 130 court‐connected custody mediations involving 270 participants and 30 mediators. As a quasi‐experimental design, regression analysis controlled for a broad range of participant attitudinal and case characteristics. Mediator reflecting and eliciting strategies were associated with positive outcomes, while directing strategies had significant negative effects. Proportionally greater time spent in caucus was associated with increased participant trust in the mediator but more negative attitudes among participants. The article considers implications for mediators and court mediation programs.

No abstract

This article reviews the importance of research and understanding of mediation styles on quality assurance, ethical practice, and accuracy of research. Three studies are reviewed. One finds that while there are patterns of stylistic practice in mediation, there is no agreement on the definitions for different styles. The second finds that mediators tend to practice in either a directive or a reflective style within a given mediation, rather than using a mix of strategies. The final study highlights how different mediator strategies affect participant satisfaction with the process. Implications for practice and future research are discussed.Questions about mediator style have significant implications for mediator ethics, quality assurance, and the ability of researchers to identify the outcomes of mediation. Yet little empirical research has been performed to understand mediator style and its implications for practice. This article outlines the dangers of ignoring mediator style, discusses research that points to confusion about the labels that are used, and provides direction for future research.The research described in this article intentionally did not begin with an attempt to study certain styles. Instead, the goal was to understand if there were, in fact, patterns of behaviors or tactics that mediators tended to use together and then to see if these patterns could accurately be tied to any commonly used term to define them. In fact, throughout this article, I resist using the names of specific styles to describe the groups of mediator behaviors that were observed because to do so would imply that there are accepted definitions of styles. The primary thesis of this article is that there are not standard accepted definitions. The research described here supports this assertion and points to problems associated with the lack of standard definitions.The research described in this article comes from three separate studies. The first was a survey of mediators practicing in a variety of settings throughout Maryland. The second involved observations of live mediations and coding the behaviors used by the Negotiation and Conflict Management Research