Transformative Women, Problem‐Solving Men? Not Quite: Gender and Mediators' Perceptions of Mediation

In various countries, mediation is increasingly gaining significance as a less expensive and often more satisfactory alternative to litigation (Boulle and Nesic 2001), as well as a tool for resolving conflicts in business, communities, and schools (Kovach 2000). At face value, any given mediation process, particularly one with legal consequences, is intended, first and foremost, to help parties produce a stable agreement. Indeed, most definitions of mediation in literature emphasize this instrumental (also known as problem‐solving) aspect of mediation and present it as a powerful tool for resolving disputes and satisfying parties' conflicting interests (Bush and Folger 1994; Moore 1996).

Other definitions also include that mediation can be a means to enhance communication and maintain the relationship (Fuller 1971; Kovach 2000). Some ambitious advocates, however, aim even higher. They emphasize the potential of mediation to transform disputants (Riskin 1982; Menkel‐Meadow 1991; Bush and Folger 1994) and to affect social justice (Wahrhaftig 1982; Herrman 1993; Moore 1994).

Robert A. Baruch Bush and Joseph P. Folger (1994) are strong advocates for “the transformation story” of mediation, arguing that mediation can bring about two changes in the disputants: empowerment and recognition. Empowerment is achieved in mediation when disputants are given maximum autonomy in their decisions and when their ability to define their own interests and options is enhanced. Essentially, their self‐worth and ability to face conflicts and other difficulties are strengthened. Mediators help parties achieve recognition by enhancing disputants' interpersonal communication and attention, thereby strengthening their ability to relate with concern to others and to see their points of view. By these two processes, disputants may grow “compassionately strong” and achieve moral growth. A settlement may well be reached in transformative mediation, but it is only desirable if it is the true autonomic choice of the parties — it is more of a by‐product than a set goal (Bush and Folger 1994).

While the mediation definitions presented above, both instrumental and transformative, are mainly concerned with individual results, two other polarized views of mediation focus on its social potential. Some scholars emphasize its ability to unite and empower representatives of weak groups and, thus, increase social justice (Wahrhaftig 1982; Herrman 1993; Moore 1994). Others, among them feminist critics, warn against the privatization of conflict resolution in closed‐door, interest‐based mediation, which may foster social inequality and injustice (Harrington 1985; Fineman 1988). While this theoretical argument is of sociological and political significance, it is the instrumental versus transformative theories of mediation that give rise to specific practices in the field.

Early theories of mediation practice tended to emphasize the impartiality of the mediator. His or her role was to encourage disputants to resolve their conflict by themselves and to create their own agreement. The mediator was portrayed as a facilitator, who assists the negotiation and encourages communication, without any personal interests in the matter of conflict or any preference regarding the terms of agreement (Fisher and Ury 1981; Stulberg 1981; Silbey and Merry 1986; Smith 1994).

In this “mythic world of mediation,” as Deborah M. Kolb (1994) called it, the resolution of the dispute is completely voluntary and free from the mediator's influence. This “mythological” practice was expected to achieve the instrumental and most commonly agreed goal of mediation: an agreement. As Susan S. Silbey and Sally E. Merry (1986: 7) put it: “The mediator  . . . faces a dilemma: to settle a case without imposing a decision . . . the tension between the need to settle and the lack of power to do so.” Indeed, many mediation scholars have contested these assumptions about mediators' noninfluential position and have shown that mediators can often be directive and even dominant. They rephrase disputants' arguments, raise options and convince participants to accept them (Vanderkool and Pearson 1983), emphasize the disadvantages of the alternative court procedure, control the communication process in terms of both topics and turn taking, and design the meetings' setting (Silbey and Merry 1986). At times, they even pressure the parties to accept a recommended agreement, over disputants' own preferred solution (Dingwall 1988). These practices probably more often occur during a problem‐solving mediation, which is the most widely documented, though transformative mediation has attracted both theoretical and practical attention since its development (Seul 1999).

From the range of views presented above regarding mediation goals and ensuing practices, mediators may choose freely, working as they do in a closed‐door room. We are interested in their choices and in the personal attributes that affect these choices. The creators of the transformative approach believe that mediators have “personal ideologies” (Della Noce, Bush, and Folger 2002), which are implicit or explicit perceptions of social frameworks and conflicts, and, accordingly, of the role of mediation. Bush and Folger (1994) have described three alternative worldviews that they believe are the basis of mediation ideologies. The individualistic worldview sees people as “unique, separate and autonomous” (Bush and Folger 1994: 238), defining their own needs and interests, and aiming to satisfy them. Conflict is an inevitable obstruction and, for the sake of both parties, had better be resolved. This gives rise to a more pragmatic, instrumental attitude toward mediation, the goal of which is reaching an agreement that satisfies conflicting personal interests and resolves the conflict.

An alternative view is expressed in the organic worldview that sees people as part of a bigger social framework, expected to serve the collective welfare. Here, conflict is seen as a threat to the social fabric, and the relationship must be maintained even at the expense of personal interests. The organic worldview does not enhance any known attitude toward mediation, because the neglect of parties' personal interests would go sorely against its core.

The relational worldview combines the other two worldviews, seeing people as both separate and connected, capable of both self‐interestedness and concern for others. According to Bush and Folger's conception of this worldview, conflict presents a unique opportunity to express and develop both abilities, as it involves the conflicting motivations of competition and collaboration (Thompson and Nadler 2002). This relational view is the basis for setting the goal of transformation in mediation, which at the same time empowers the individual and teaches recognition of others (Bush and Folger 1994).

Because individuality and connectedness seem to affect attitudes toward conflict and because women and men are thought to fall into different places on the spectrum between individuality and connectedness — both in traditional stereotype and in some empirical research — gender could be a personal attribute that may differentiate between mediators in regard to these attitudes.

Gender scholars discern between stereotypically feminine “expressive” traits and stereotypically masculine “instrumental” traits. For example, nurturance and kindness are deemed to be expressive traits, while independence and competitiveness have been classified as instrumental (Bem 1974; Spence and Helmreich 1980; Eagly 1995; Halpern and Parks 1996). Another widespread contention is that gender affects self‐construal — men view themselves as more independent, while women see themselves as more interdependent. Even at childhood games, girls avoid conflict and emphasize collaboration more than boys do (Babcock and Laschever 2003). These characteristics are sometimes expressed in the way that women and men use language. Deborah Tannen (1990) showed that men's language tends to express a more competitive perception, and women's language tends to express a more interdependent perception. Robin Lakoff (1975) wrote that women's speech tends to show less confidence and to be phrased to avoid conflict, although it should be noted that much of this research is now at least one or two generations old.

Instrumental versus expressive traits are relevant to conflict resolution and, therefore, to negotiation and mediation. While few scholars of mediation have focused on gender, a large body of negotiation research has looked at the role of gender in negotiation behavior. Studies have shown that women are more likely to emphasize the quality of the interpersonal contact created in the negotiation, while men are more likely to emphasize their own gains and successes (Halpern and Parks 1996; Florea et al. 2003; Kray and Thompson 2004). Men tend to achieve higher personal profits and also feel more comfortable negotiating than women (Walters, Stuhlmacher, and Meyer 1998; Stuhlmacher and Walters 1999; Babcock and Laschever 2003).

Gender as a variable of composition was also studied, and its effects on both negotiation process and outcomes were reported. Male dyads tend to be more competitive and task oriented than female, while female dyads tend to be more collaborative in negotiation and personal in conversation. In mixed‐sex dyads, women were found to emphasize the relationship more, and achieve less profit, than men (Kray and Thompson 2004). In fact, in one study, female adolescents rated their negotiation sessions as successful even when they did not reach agreement, so long as harmony was kept and the relationship was fostered (Florea et al. 2003).

Linda Babcock and Sara Laschever (2003) have argued that women's interpersonal abilities allow them to collaborate and achieve creative solutions but that in mixed‐sex dyads, where the other party is competitive, these abilities may be exploited as weaknesses. Men were shown as more successful than women in using “tit for tat” strategy in matrix games (e.g., prisoner's dilemma). They collaborated only with a collaborative partner and competed with a competitive one, a profitable strategy in matrix games (Watson 1994; Walters, Stuhlmacher, and Meyer 1998). Men's willingness to collaborate was therefore attributed to an assertive, self‐serving attitude rather than to a social, interdependent one (Walters, Stuhlmacher, and Meyer 1998).

Nevertheless, these gender differences in negotiation are notably inconsistent (Walters, Stuhlmacher, and Meyer 1998), particularly in the laboratory (Babcock and Laschever 2003). Their inconsistency is typical of psychological research (Eagly 1995), and in response, some scholars have explored under what conditions, and through which mechanisms, gender matters (Bowles, Babcock, and McGinn 2005). For example, status and situational power (such as better alternatives) sometimes interact with gender in affecting negotiation performance and often have results beyond gender, rendering the latter insignificant (St. John 1996; Kray et al. 2004; Bowles, Babcock, and McGinn 2005).

However, other reported moderators of gender‐typed negotiation behavior support theories of female interdependence and male independence. In a carefully constructed experiment (Bowles, Babcock, and McGinn 2005), the researchers showed that when negotiating on behalf of a third party, women negotiated higher gains than men and than did women negotiating for themselves. When negotiating on behalf of another, this study found women to be more, not less, competitive. Several explanations for this successful competitiveness were tested, and some research indicates that women were more highly motivated when caring for another's needs, although additional studies suggest that women recognize that it is more culturally acceptable for them to negotiate competitively on behalf of others and that they fear a social and financial backlash when they behave more assertively for themselves (Bowles, Babcock, and Lai 2007; Tinsley et al. 2009).

Results from another study (Kray et al. 2004) indicate that men and women were aware of their instrumental versus expressive stereotypes and that these affected their negotiation performance. In this experiment, when it was implied that male stereotypic behavior was profitable in negotiation, women did not perform as well in negotiation as men. Conversely, when it was implied to participants that interpersonal traits were effective in negotiation, the men gained less than the women.

Therefore, not only have researchers found gender‐typed behavior in numerous negotiation studies, supporting the independent versus interdependent self‐construal theory, but also recent studies into the moderators and mechanisms of gender differences in negotiation further support this notion. To use Bush and Folger's (1994) terminology, men seem rather to consistently express in their behavior an individualistic worldview. Certainly so when they are being competitive, but even when they collaborate, it is likely to be an assertive, self‐serving choice. Women seem to shift between an organic worldview (when they emphasize relationship over negotiation performance) and a relational worldview (when they collaborate and achieve both). Intervening variables, however, such as status, show women to be capable of an individualistic competitive approach as well.

Because gender has been shown to have some effects on negotiation behavior, it could also make a difference in mediation, as a form of “guided” negotiation. It can be predicted to affect the disputants' behavior as negotiators. But we are interested in its effect on mediators, who facilitate or direct a negotiation between parties, rather than directly participate in one. It takes interpersonal skills and orientation to mediate at all and, even more so, to fulfill the potential of mediation to transform the disputants and enhance their relationship. As women have been shown in the literature to have a stronger interpersonal orientation than men and, more specifically, to emphasize the relationship aspects of negotiation more than men do, we may expect to find gender differences in mediators' views of conflict and its potential, and in their perceptions about the goals and practices of mediation on the continuum between problem solving and transforming.

Initial studies in the field of mediation and gender show that female and male mediators do conceptualize their roles somewhat differently. A survey of mediators in the United States (Herrman et al. 2003) indicated that female mediators ascribed more importance to parties' emotions and nonverbal behavior, compared with their male peers, who emphasized a more instrumental, settlement‐oriented attitude. In other studies, female mediators reported being more facilitative in their style, while male mediators testified taking more control over the process toward achieving a settlement (Weingarten and Douvan 1985; Dewhurst and Wall 1994). Cheryl Picard (2002: 256) found that both male and female mediators described their role as “facilitator” but in fact meant different things by it. Women emphasized communication and spoke either of “facilitating communication” or “communication and process,” while men tended to mention only “facilitating process.” One qualitative study of gender and mediation had been previously reported in Israel, but it reached ambiguous conclusions (Desivilya, Ady‐Nagar, and Ben‐Bashat 2004).

Further research into the effect of gender on mediators' perceptions is clearly called for. We are interested in gender differences along the instrumental to transformative continuum. If, as much of the research has shown, we can expect that female mediators will display more of an emphasis on interpersonal skills than will their male counterparts and will more strongly emphasize the role of relationships in negotiation, does that lead them to aspire differently than men in their goals, beyond agreement and toward transformation? Furthermore, gender‐typed mediation goals may lead to gender‐typed mediation practice styles. If women mediators' goals are more transformative and less instrumental than are men's, we would also expect them to endorse facilitative, rather than directive, practices more than men do.

In designing a study to answer these questions, we also attempted to overcome one of the limitations in some of the previous studies comparing female and male mediators, namely, their usage of small samples and fully qualitative research designs (e.g., Weingarten and Douvan 1985; Desivilya, Ady‐Nagar, and Ben‐Bashat 2004). While they were rich in content, further research is needed to replicate and generalize findings about the significance of mediators' gender. In light of this, our study included a large sample of mediators, representing diverse professional backgrounds and areas of mediation. Furthermore, it was designed to utilize both qualitative and quantitative methods, allowing for authenticity and empirical validity at the same time.

Our hypotheses, based on the literature presented above, are below.

Although our research was theoretically grounded on, and meant to test, some specific hypotheses, it was designed as an explorative phenomenological field study, based on mediators' own descriptions of their perceptions. We conducted two studies; the first was qualitative and interpretational, based on the content analysis of interviews with twenty mediators. From that content, we formulated a detailed questionnaire for the second study, which was then answered by a large and representative sample of mediators. Thus, we were able to not only reflect mediators' authentic views and concepts, rather than impose our own assumptions, but also to statistically validate our results and to allow for generalization.

Our respondents were twenty practicing Israeli mediators, in various domains, including family, work, and business: eleven men and nine women, ages thirty‐eight to sixty. Half the respondents had a legal professional background, and the rest had backgrounds in social sciences, management, and economics.

Each mediator was interviewed in a semistructured interview that included nearly thirty open standard questions about goals, mediator's perception of his or her role, preferred practices, criteria for success and failure, case descriptions, etc. The interviews lasted two hours on average and were conducted by a team comprising a legal expert and a psychologist, to enrich interviewers' perspectives. Two independent judges, expert in mediation, then analyzed the mediators' answers and partitioned them to the smallest units of meaning, usually short sentences or statements.

The statements that both judges agreed on (i.e., both chose the exact same statement as a separate unit of meaning) were then classified into categories based on semantic similarity, leaving out statements that did not fit any category. Three judges, two psychologists, and a lawyer‐mediator then formulated the comprehensive questionnaire that would be used in Study Two, representing each category by two to three of the statements originally classified into it that all members of the team completely agreed on. The team organized the statements into a series of subquestionnaires on various mediation topics that represented not only the issues we addressed in our hypotheses but also other topics raised by interviewees (see details in Study Two). We used the interviewees' own phrasing to formulate the statements in the questionnaires.

In each questionnaire, respondents were asked to score their agreement with each statement on a six‐degree scale (1 = “strongly disagree” to 6 = “strongly agree”). Additionally, for some of the questionnaires, respondents were further required to rate the three most important statements (e.g., if statements described different mediation goals, respondents rated their own top three goals among them). This was designed to control for the confounding effect of social desirability on answers, which might have led respondents to refrain from expressing their true preferences and to endorse just as strongly all statements (e.g., “reaching a stable agreement” is just as important as “enhancing parties' communication”). In a pretest of the questionnaire, experienced mediators answered it and then offered their comments, which led to some corrections in the final version. A demographic questionnaire was added, including questions about gender, age, education, profession, mediation training, seniority in mediation, and area of mediation.

To represent the population of mediators in Israel, we sampled respondents in clusters according to professional background. We sent questionnaires to a total of 489 mediators, and 43 percent of them (189 mediators) returned the questionnaires, a high return rate for mailed questionnaires. Participants were practicing mediators from the mediator pool of the Israeli Ministry of Justice and other authorized mediation institutes. One hundred twenty‐four of them were men and sixty‐three were women (two failed to report their gender). Their ages ranged from thirty to seventy, with 50 percent of the sample above the age of fifty. Seniority in mediation was variable and ranged from fewer than three years to more than eleven years, but most (58.2 percent) had five or fewer years of seniority. Professional backgrounds included social sciences, law, management, and economics. Mediation areas included family, work, business, and community. A description of the questionnaires follows (for full versions, please see  Appendix One).

Respondents scored their agreement with items representing perceptions of the essence of mediation that were identified in the first study. A factor analysis yielded four factors that explained 59 percent of the variance. A transformative orientation was indicated by the mediator's interest in strengthening the disputants and their relationship (reliability: Cronbach's alpha = 0.85; examples: “mediation is a tool to create communication,”“mediation allows for participants' moral growth”). This factor included half the items in the questionnaire because transformative views were expressed frequently and richly in prior interviews.

An instrumental orientation was indicated by mediator's interest in the goal of settling a specific dispute (reliability: alpha = 0.59; example: “mediation is a flexible procedure to resolve a conflict”). A social positive orientation was expressed by the mediator's support for statements that endorsed the idea that mediation can strengthen weak groups and lessen power inequality (reliability: alpha = 0.70; example: “mediation reduces power inequality in society”), and the social negative orientation involved statements expressing the fear that mediation in fact increases power inequality (reliability: alpha = 0.59; example: “mediation increases power inequality in society”).¹

Respondents were asked to agree with statements articulating some of the different mediation goals that had been noted in the preliminary interviews. A factor analysis revealed three factors, which explained 69 percent of the variance. An agreement emphasis was reflected by statements involving settlement of the dispute (reliability: alpha = 0.57; example: “to bring about the resolution of the dispute”). A relationship goal was reflected in statements concerning parties' communication and mutual recognition (reliability: alpha = 0.80; examples: “to build trust between the parties,”“to create behavior change between the parties”), and a process emphasis was indicated by support for statements concerning the process itself rather than its conclusion; mediators with this goal are less concerned with interpersonal and more with intrapersonal processes (reliability: alpha = 0.68; example: “to help parties reveal their interests”).

Respondents scored their endorsement of items representing success criteria that were mentioned in the preliminary study. Three of the same factors that emerged in this analysis also emerged in the Mediation Goal Questionnaire (MGQ), explaining 67 percent of the variance: an agreement (reliability: alpha = 0.46; example: “achieving an agreement that will show stability over time”); the relationship (reliability: alpha = 0.84; examples: “when communication between the parties is enhanced,”“when hostility between the parties is reduced”); and the process factor, which can include both disputant's intrapersonal processes and mediator's ability to lead a respectful, helpful procedure (reliability: alpha = 0.51; examples: “when the procedure was pertinent,”“when parties better understand alternative solutions”).

Respondents scored their agreement with items representing criteria for mediation failure, formulated from the interviews. A factor analysis yielded two factors that explained 66 percent of the variance. The first was the parties; these statements attributed failure to the parties' behavior (reliability: alpha = 0.70; examples: “when no change occurred in parties' positions,”“when parties do not cooperate with the mediation procedure”). Those respondents who attributed the failure to the mediator agreed with statements describing the mediator's failure to manage the process appropriately (reliability: alpha = 0.70; example: “when the procedure itself was managed unprofessionally”).

This questionnaire included a list of statements related to elements of mediation practice: “explaining to the parties about the mediation,”“presentation of the conflict by each party,”“reflection by the mediator,”“private meetings with each party,”“identifying interests,”“raising solution options,”“evaluating options,” and “processing an agreement.” Here, respondents did not score their agreement with each but were merely asked to rate the three most important elements.

Respondents were presented with a list of things that disputants often expect from a mediator (and were mentioned in Study One) and asked to score how strongly they agreed that each expectation existed. A factor analysis yielded two factors that explained 61 percent of the variance. The judicial factor involves the perception that the mediator is a judge who is there to end the dispute (reliability: alpha = 0.65; examples: “[expect to] have the mediator acknowledge their justification,”“[expect] that the mediator will help them to resolve the conflict”). The therapeutic factor depicts the mediator as a facilitator who is there to enhance the process and the relationship (reliability: alpha = 0.61; examples: “[expect] that the mediator will contribute to reduce conflict's intensity,”“[expect] that the mediator will assist in creating communication between the parties”). Another item, “that the mediator will be attentive to them,” was also initially included as one of the therapeutic factor statements, but its reliability was low, and therefore, we analyzed it separately as a third factor, attentiveness.

Respondents scored their agreement with statements concerning the reasons that people choose to work as mediators. Examples include: “Mediation fascinates and challenges me,”“I want to work with people,” and “I identify with the notion of mediation.” A factor analysis yielded one factor that explained 45.30 percent of the variance. As all questionnaire statements were loaded on that factor (reliability: alpha = 0.80), we calculated a mean motivation score for each respondent based on the entire questionnaire.

Respondents scored a list of statements representing different aspects of job satisfaction. From among the list, two groups of items showed high intragroup correlation, and therefore, we analyzed each as a separate scale. The satisfaction scale directly describes level of satisfaction (reliability: alpha = 0.83; example: “in general, I'm highly satisfied with my work as a mediator”). The sense of competence scale describes feeling able to do the job (reliability: alpha = 0.65; example: “I feel skilled enough to mediate”).

First, we must note gender differences in some of our background data, both for the sake of describing our population and as it may affect our results. Male and female mediators differed significantly in professional background (Pearson's chi‐square = 16.75, degrees of freedom = 3, p < 0.001), with more women than men in social sciences, more men than women in economics, and similar rates in law and management. They also differed in mediation area (Pearson's chi‐square = 37.26, degrees of freedom = 4, p < 0.001), with more women than men practicing family, community, or a combination of areas, and more men than women practicing work and business mediation. Of these two variables, professional background showed effect on some of our dependent measures in preliminary analyses. Therefore, to control for its potential confounding effect, professional background was assessed against gender in all our measures. That is, we performed a series of two‐way multivariate analyses of variance (MANOVAs) with gender and professional background as independent variables and each questionnaire's factor scores as dependent variables. The MANOVAs yielded no interactions between gender and professional background; because this article is not concerned with the latter, we proceed to report only the gender main effects.

Table One shows all factors' mean scores for male and female mediators, as well as the statistical gender difference (or lack thereof) between the means of each factor.

As can be seen in Table One, most questionnaires' factors showed gender differences. We also calculated the overall gender effect for each questionnaire, and the results of those calculations are subsequently reported in the text, along with references to the specific results for each factor appearing in Table One.

The two‐way MANOVA yielded a main effect for gender (F = 6.63, p < 0.001, correlation coefficient = 0.14). Compared with their male peers, female mediators endorsed more strongly the transformative orientation and somewhat more strongly the social positive orientation. Men and women, however, did not differ in their scores on the instrumental orientation, which both scored highest, nor in their disagreement with the social negative orientation, which scored lowest (see Table One). Therefore, Hypothesis One was supported and Hypothesis Two was not.

The two‐way MANOVA yielded a main effect for gender (F = 3.99, p < 0.01, correlation coefficient = 0.07). Female mediators scored higher than their male peers on the process goal, emphasizing the importance of both mediation room process and the relationship of the disputants. We found no gender difference in the scores for the agreement goal, which both genders scored highest (see Table One). Again, Hypothesis One was supported and Hypothesis Two was not.

The two‐way MANOVA yielded a main effect for gender (F = 3.94, p < 0.05, correlation coefficient = 0.07). The difference resulted because the female mediators scored the relationship factor, describing success in terms of relationship enhancement, higher than did their male peers. We found no gender difference in the agreement and process factors (see Table One). Hypothesis One was partially supported (relationship enhancement) and Hypothesis Two was not.

The two‐way MANOVA yielded a main effect to gender (F = 4.29, p < 0.05, correlation coefficient = 0.05). When assessing to what extent the mediations described in the questionnaire were unsuccessful, women scored both the mediator failure factor and the parties' failure factor higher than did the men (see Table One). These questions did not directly address our hypotheses but represent the ways in which failure was discussed in interviews. Still, it yielded an interesting gender difference, which we will discuss later.

The two‐way MANOVA revealed a main effect for gender (F = 4.93, p < 0.01, correlation coefficient = 0.08). Compared with male mediators, female mediators more strongly attributed to their clients each type of expectation (see Table One): attentiveness, therapeutic‐like actions, and the expectation for judicial actions. The gender difference we found does not confirm our hypotheses about the instrumental–transformative or the directive–facilitative continuums, although it could be argued that it indicates that women mediators are more strongly aware of all disputants' expectations, thus generally confirming the basic premise that they are more interpersonally oriented.

The data for this questionnaire do not appear in Table One, as in this case we asked respondents to rate the three most important elements rather than to score each one. Men and women differed in their ratings of four out of eight items. More women than men chose the statements “presentation of the conflict by each party” (Pearson's chi‐square = 3.38, degrees of freedom = 1, p < 0.06) and “reflection by the mediator” (Pearson's chi‐square = 6.23, degrees of freedom = 1, p < 0.01). These activities — mediator reflection and encouraging disputants' self‐expression — are facilitative, process‐focused activities. Conversely, more men than women chose, as their most important mediation elements, “explaining to the parties about the mediation” (Pearson's chi‐square = 4.14, degrees of freedom = 1, p < 0.04), which is a directive element, and “private meetings with each party” (Pearson's chi‐square = 5.34, degrees of freedom = 1, p < 0.02), an element that allows the mediator more control of the process but does not enhance direct communication between the parties. Therefore, the female mediators showed a stronger preference than did the male mediators for facilitative and communication‐building techniques, while the male mediators showed a stronger preference than female mediators for directive practices. Hypotheses Three and Four were therefore supported.

The elements in which we found no gender differences were basic building blocks of problem‐solving mediation but did not clearly exclude more facilitative or transformative‐type activities (e.g., “identifying interests” or “processing an agreement”). Moreover, they did not specify a style, facilitative or directive. Therefore, the lack of difference in these statements is not clearly salient to our hypotheses.

Information from the Mediator Motivation Questionnaire (MMQ), mediator satisfaction questions (in Mediator Satisfaction Questionnaire [MSQ]), and mediator sense of competence questions (in MSQ) did not address our hypotheses, although we did find some interesting gender differences.

A two‐way ANOVA by gender and professional background yielded a gender difference in the mean motivation score (F = 16.69, p < 0.001, correlation coefficient = 0.09). When assessing their identification with different reasons for mediating, female mediators scored higher, overall, than their male peers (see Table One), indicating that they may feel more strongly the different inducements for mediating, which may indirectly suggest higher motivation.

A two‐way ANOVA by gender and professional background yielded a gender difference (F = 4.49, p < 0.05, correlation coefficient = 0.03). Femalemediators reported a somewhat higher level of job satisfaction than did male mediators (see Table One).

A two‐way ANOVA by gender and professional background did not yield a significant gender difference.

In this large and representative sample of Israeli mediators, female mediators showed a stronger tendency toward more facilitative and transformative approaches to mediation than did their male counterparts, although they were just as agreement oriented as their male peers and held agreement to be the most important goal of mediation. Women's orientation toward mediation was no less instrumental (problem solving and agreement oriented), but the women in our study displayed more interest in mediation approaches that can be categorized as more facilitative, relationship oriented, and transformative than did their male counterparts. Similarly, women and men both endorsed agreement as the most important goal and success criteria, but the women attached more importance than did the men to relationship enhancement and to process factors.

The same propensity was reflected in their preferences of practice elements. Women, more strongly than men, endorsed facilitative behaviors, such as letting the parties tell their stories, while men preferred directive techniques that allow more control, such as private meetings, more than women did. Female mediators also indicated a greater awareness of their disputant's expectations. Whether parties expect judge‐like behavior, simple attentiveness, or therapeutic‐like attendance, women scored higher than did men on their attentiveness to disputants' expectations.

These differences support our premise that because women have been shown in previous studies to display more cooperative, relationship‐oriented negotiation behavior than men, they would likely be more tuned into the relationships of the people in the mediation room — the relationship between the disputants themselves as well as the relationships between them and the mediator. Our results indicate that women mediators are more sensitive to disputants' expectations (as far as acknowledging them) and more ready to take the opportunity to facilitate a process that is meant to reach more than just a signed agreement.

The gender differences we found in respondents' professional backgrounds and current mediation areas could also be interpreted as supporting the premise that female mediators are more interested in relationships than male mediators. More women than men in our sample were educated in the social sciences, and more women than men chose to handle family mediations, while more men than women mediated business disputes. It could be claimed that these tendencies reflect a stronger interpersonal tendency on the part of the women mediators, which may go along with stronger therapeutic propensities, although social expectations and economic considerations could also play important roles. Clearly, further research is needed to test these interpretations.

Both male and female mediators shared the same primary mediation goal, which is an instrumental one: to help parties reach agreement and settle their dispute. Transformation and relationship enhancement came second. Thus, we found that, contrary to our expectations, female mediators were no less task oriented than their male counterparts. This may be because they all share the same profession and status, which research in several fields has shown to be no less — if not more — predictive of cognition and behavior than gender (St. John 1996; Aguinis and Adams 1998; Kray et al. 2004; Bowles, Babcock, and McGinn 2005). In other words, just as businesswomen approaching negotiation may adapt themselves to the norms and practices that prevail in a business environment, it is possible that our female respondents adapted themselves to their particular environment.

Further, we must remember that the participants in our study mediate court cases, and it is in the interests of courts to reduce litigation and the costs associated with it, and thus, failure to reach a settlement is considered a failure. This strong contextual feature, which emphasizes an agreement as the major success factor, may override personal gender‐typed tendencies. Additionally, our findings are in line with the literature that shows problem‐solving orientation to be the most popular among mediators (Seul 1999).

We found some additional intriguing gender differences. As stated before, we chose to base our quantitative study on a prior qualitative, explorative study, in which interviews with male and female mediators helped us to create content that we then formulated into questionnaires. This method enriched our study with findings for which we did not form hypotheses. One was that female mediators agreed more strongly with statements that mediation suits their abilities and interests and that they identify with its notion and cause. Further research is required to evaluate whether gender affects mediators' level of motivation, and if so, what explains this effect.

Female mediators' higher level of identification with reasons for mediating may explain their somewhat higher level of job satisfaction, compared with their male peers. They may be more satisfied because they earn their living doing a job that they more strongly feel suits their ideals and abilities. This explanation could be tested by further study.

Another intriguing difference was that women scored higher on the failure criteria questionnaire — they rated higher than men items of both factors of failure, those centered on the parties (“not being cooperative”) and those more directly concerning the mediator (“not being professional”). It seems that they are more inclined than the men to see these typical mediation situations as failure. It may indicate that they are more critical of their own performances and of the mediation process altogether than are their male peers. We wonder whether this also means that in practice, female mediators worry more than male mediators about failure or are quicker to describe their mediations as failing. This result, however, is a preliminary thread, and the questions it raises require further study.

In male and female mediators' sense of their own competence, we found no explicit difference. But the confirmation of our hypotheses that female mediators are more oriented to facilitative, relationship‐oriented processes than are male mediators suggests that when they report feeling competent on the job, male and female mediators may mean different things. For men, “competent” may mostly mean “able to reach a stable agreement” or “able to direct a negotiation that leads to an agreement.” For women, it may mean all this but also “able to strengthen the parties” or “able to enhance their relationship.” Cheryl Picard (2002) has shown that male and female mediators use the same word to mean different things. We suggest, for further study, the possibility that male and female mediators mean different things when they speak of competence.

Female mediators seem to have a broader range of perceptions about their job than do male mediators. Our mediators did not subscribe to a theoretical dichotomy between instrumental and transformative orientations (Bush and Folger 1994); rather, they all endorsed both approaches — and other approaches along the spectrum — though to differing degrees. But this is particularly true for female mediators. While they were no less instrumental in their inclinations, they expressed more interest in transformative approaches and goals than did the male mediators.

One implication of our findings concerns mediators' training. It is worthwhile for the trainers to be aware of the possibility of gender‐typed attitudes toward mediation, to discuss different approaches toward mediation, and to encourage trainees of both sexes to be aware of the spectrum of mediation goals and practices. Further, trainees could benefit from an encouragement to try approaches that are new or less comfortable for them, thereby enriching their “toolkit.”

Our findings also raise an intriguing question: if female mediators are more interested than male mediators in facilitative practices and in transformation in mediation, does that imply that they would be more appropriate than male mediators for mediations that are particularly educational in nature, such as mediations at school? While we believe that competent mediators of either gender should be flexible enough to adapt themselves to different environments, it would be interesting to research whether male and female teachers handle disputes differently and whether they create differing degrees of transformation in their students when they help them resolve disputes.

More research would overcome limitations in our study. One limitation is that we obtained mediators' perceptions at only one time in their careers. We cannot know how their perceptions developed and how specific mediation experience has moderated whatever impact gender may have had. We suggested, based on previous research, that our mediators' specific experience did indeed moderate hypothesized gender differences — but we do not have those particular data. A longitudinal study evaluating perceptions at the beginning of mediators' careers and then again later on, involving mediators from the court system and outside of it, would better highlight the interplay between gender and environment in mediators' perceptions.

Finally, further research is recommended to overcome the limitation of self‐reporting. Mediation rooms are still rarely open to direct observation. Such observation is invaluable in order to find out how mediators' perceptions shape their actual practices, whether women are actually more transformative and facilitative than men in practice, and how that relates to their success in reaching agreements and in other measures, such as the length of mediation and the parties' satisfaction. One study (Payne et al. 2000) found that female mediators achieved more final and stable agreements than did their male counterparts (although both reached initial agreements) and that a facilitative style rather than a directive one was related to such success. More research is needed before the relationship among mediator gender, perceptions, actual practices, and mediation success can be better understood.

(Note: After completing the questionnaire, respondents were asked to rate the three most significant or true statements [example is shown in the mediation elements questionnaire]. Respondents were asked to rate their answers on a Likert scale with 1 indicating strong disagreement and 6 indicating strong agreement.)

• Mediation is a technique for resolving a conflict.

• Mediation is a flexible procedure to resolve a conflict.

• Mediation increases power inequality in society.

• Mediation is a tool to create communication between people.

• The procedure's secrecy and the lack of outside control over mediation allow for coercion and manipulation over weak parties.

• Mediation encourages one to help oneself.

• Mediation reduces power inequality in society.

• Mediation strengthens private people against the system.

• Mediation empowers the parties by strengthening their awareness of their self‐worth.

• Mediation enhances people's ability to recognize and attend to the other's needs.

• Mediation is an economical procedure to resolve a conflict.

• Mediation is a way of life.

• Mediation allows to perpetuate injustice against the weak and to ignore the achievements of groups.

• Mediation is an educational experience.

• Mediation allows for participants' moral growth.

• Mediation changes the character of the parties, and of society as a whole.

• To get the parties to agree.

• To bring about the resolution of the dispute.

• To create communication between the parties.

• To build trust between the parties.

• To create behavioral change between the parties.

• To empower the parties.

• To create a better society.

• To help parties reveal their interests.

• To help parties understand the various alternatives for resolving the conflict.

• Achieving an agreement.

• Achieving an agreement that will show stability over time.

• When a party understands his/her own and the other party's interests.

• When communication between the parties is enhanced.

• When a negotiation was carried out between the parties.

• When a party feels that his/her narrative was recognized.

• When parties better understand alternative solutions for resolving the conflict.

• When the procedure was pertinent.

• When hostility between the parties is reduced.

• When parties do not cooperate with the mediation procedure.

• When the procedure itself was managed unprofessionally.

• When important issues at dispute were not resolved.

• When no change occurred in parties' positions.

• When communication was not created between the parties.

• When parties did not reach a settlement.

• Explaining to the parties about the mediation.

• Presentation of the conflict by each party.

• Reflection by the mediator.

• Private meetings with each party.

• Identifying interests.

• Raising solution options.

• Evaluating options.

• Processing an agreement.

(Note: Respondents were asked to choose the three most important from the above list.)

• That the mediator will help them to resolve the conflict.

• That the mediator will acknowledge their justification.

• That the mediator will be attentive to them.

• That the mediator will bring about the resolution of the conflict.

• That the mediator will assist in creating communication between the parties.

• That the mediator will contribute to reduce conflict's intensity.

• I have the right personality for it.

• I identify with the notion of mediation.

• Mediation gives me a sense of satisfaction.

• I want to work with people.

• Mediation is a way of life for me.

• Mediating presents the finest exercise of my abilities.

• Mediating earns me a living.

• Mediation fascinates and challenges me.

• In general, I'm highly satisfied with my work as a mediator.

• In general, I'm satisfied with my performance as a mediator.

• I often consider retiring from mediation.

• I feel skilled enough to mediate.

• I succeed in mediation.

• I have the right abilities for mediation.

• I need to extend my knowledge in mediation.