Married women in missions: the effects of cross-cultural and self gender-role expectations on well-being, stress, and self-esteem.

The present study explored the effects of self-expectations and societal expectations of the host culture on the well-being of 37 married missionary women. The results did not support a relationship between the expectations of the host culture, and well-being. Homemakers appeared to be more relaxed and to experience life as more satisfying and interesting than women involved more actively in the missions task. The congruence of roles with self-expectations, role satisfaction, and freedom in choosing a role emerged as highly related to several indices of well-being. These findings highlight the centrality of freedom in choosing a role, and suggest that important subcultural differences in self-expectations exist in the Christian subculture which should be taken into account in research on women's issues.

**********

As today's world has grown smaller through the internationalization of politics and economy, cross-cultural living has become more common. Many individuals and families now face the challenges of living as sojourners in a different country, whether it be as part of the business world, through the State Department, or as missionaries.

The challenges of cross-cultural living have been studied in an attempt to facilitate the work of these cross-cultural sojourners. Some of this literature has focused on missionaries. Recent research in the area of missions and mental health has shown that missionaries as a group are subject to a great number of stressors, many of which are unique to living cross-culturally, or are exacerbated by these circumstances. Chester (1983) suggested that the resulting burnout may be more of a problem with missionary wives than with their husbands because of the role into which they feel they must fit in addition to the isolation and confinement (both cultural and social) that they often experience.

Additional data confirm Chester's hypothesis that women missionaries may face unique cross-cultural stressors. Gish (1983) had a sample of 549 missionaries rate a number of common stressors. Her results confirm Chester's hypothesis that burnout may be more of a problem among women on the field than among men. Of eleven significant gender differences on certain sources of stress, ten variables were significantly more stressful for women than for men. Among the variables identified in this way were coworker's attitudes toward their job, loneliness and isolation, self-acceptance, and communicating across the cultural and language barriers--all of which are related in some way to role issues. Carter (1999) replicated Gish's study with more stringent statistical cutoffs for significance, and similarly found that extended family concerns and relationship with spouse of partner were reported as more significant sources of stress for women than for men. These findings again highlight the role tensions between family demands and work that women experience.

Though no empirical studies have been done relating role issues to stress in married women missionaries, the existing theoretical literature suggests that there is a link between roles and stress. Married women missionaries may face more stress than their husbands because of the multiplicity of roles that they assume, as well as frequent role changes. O'Donnell (1987) presented a family life cycle approach to aid in understanding the missionary family. In almost every stage, the woman must face major role renegotiations. Initially, this involves balancing the responsibilities of childrearing with mission-related work and domestic duties. Later, as midlife career issues come to the forefront, she may want to develop herself further as a person beyond the roles of childrearing and homemaking. When the children leave the home, her marital role will need to adjust, and she will be free to become more active outside of the home. This last transition can be particularly difficult for women who have to that point found their main role identity as a mother.

Bowers (1984) further explored role-related issues of concern for women missionaries. She defined a role as "a cluster of behavior patterns that carries with it expectations on the part of the person filling the role, and also expectations on the part of others who are related to a person in the performance of the role" (p. 6). In an attempt to investigate role-related issues of concern among women missionaries, the present study focuses on role expectations, those of the missionary herself as well as those of the host culture.

The issue of role-related sources of conflict among women is not unique to this population, although it takes on unique characteristics in a cross-cultural context. In the following section theories of role conflict based on societal and self expectations will be explored. These general theories will set the stage for the literature on sources of role conflict for the missionary woman, which will be reviewed in the last part of the following section.

Common Sources of Role Conflict for Women

A great deal of psychological research and theorizing on the subject of women's roles in society has been generated in the past few decades, beginning in the sixties with the Women's Liberation movement, and continuing with the challenges the modern woman faces in attempting to reconcile new societal expectations with traditional role demands. This struggle has been most apparent in the area of gender-role expectations related to work outside of the home.

Astin (1984) proposed a comprehensive theory of work, which can serve as a useful framework in examining the area of gender-role expectations. Work is an important part of examining gender-role issues, because much of the controversy and tension involved in the topic of roles revolves around the degree of role involvement women have in work outside of the home. Astin's model is a need-based sociopsychological one, incorporating motivation, expectations, gender-role socialization, and the structure of opportunity. Briefly, her theory is defined as follows:

Work behavior is motivated activity intended to satisfy three basic needs: survival, pleasure, and contribution. Career choices are based on expectations concerning the accessibility of alternative forms of work and their relative capacity to satisfy the three basic needs. Expectations are shaped in part by early socialization through family, childhood play, school experiences, and early work experiences, and in part by the perceived structure of opportunity. Expectations developed through socialization and through early perceptions of the structure of opportunity can be modified by changes in the structure of opportunity ... (p. 119)

Expectations in regard to roles are central to this model of work behavior. This model also emphasizes how intertwined the two sources of expectations are, given that societal expectations influence to a large degree the self-expectations women have for themselves. Although societal expectations and self-expectations are impossible to completely differentiate, societal expectations will be used in reference to the broad expectations of women and their role that are held by a certain society. Self-expectations will more narrowly refer to the expectations that women have of themselves, based not only on the incorporation of societal norms, but also on the idiosyncratic way in which parental influences and life experiences have molded these expectations. In the following paragraphs, the literature on expectations will be roughly grouped into societal expectations and self expectations.

Societal expectations. Societal expectations become part of women's experiences very early in life, when as children they learn what life roles, personality characteristics, and behaviors are reinforcing and suitable (Betz & Fitzgerald, 1987). These expectations continue throughout life in the form of societal expectations as to what is appropriate and what is deviant. Sometimes the societal expectations take the concrete form of limitations in the structure of opportunity in the field of work. However, these limitations have decreased as a result of increasing legislation prohibiting overt gender discrimination. It would appear that early socialization practices, and consequently current social gender-role expectations, have not changed as much as the actual occupational behavior of women. This disparity contributes to the conflict between societal expectations and self-expectations or actual gender-role behavior.

The difficulties implicit in this task of reconciling societal expectations with personal goals or aspirations have been the object of theoretical speculation and some empirical work. The gender role congruence hypothesis claims that greater congruence between role and societal gender-role expectations leads to better adjustment, and that incongruence in the form of "inappropriate" gender-role behaviors leads to distress (Tomlinson-Keasey, 1974). In our contemporary society, the working woman is repeatedly presented as an ideal, and research shows that the homemaker ideal of the 50s has given way to this new model in large segments of our society (Kerpelman & Schvaneveldt, 1999). In support of this hypothesis, a review of relevant studies showed that employed married women report better physical and mental health than unemployed married women (Betz & Fitzgerald, 1987; Froberg, Gjerdingen, & Preston, 1986; Keith & Schafer, 1998; Lennon, 1998).

In a more direct test of the role congruence hypothesis, Polasky and Holahan (1998) examined the discrepancy between what women reported as their actual behaviors and what the women thought that others expected from them in their maternal role. Their sample consisted of 103 primarily Caucasian, married, and professional women who had children aged 10 or younger. They found that a discrepancy between women's actual behaviors and others' expectations was associated with anxiety.

Self expectations. Self-expectations regarding a woman's role appear to be formed early on in life. Early socialization contributes to her understanding of what roles are reinforcing and suitable, leading, in psychoanalytic terminology, to the development of the ego ideal. In addition to the influences of society, each woman's self-expectation will be uniquely influenced by her particular upbringing and life experiences, resulting in highly personalized and unique goals for achievement, expression, contribution, etc. These self-expectations may or may not coincide with societal expectations, resulting in differing degrees of congruence.

Sederer and Seidenberg (1976) explored the internal dynamics of the conflict that can take place at the self-expectation level within a psychoanalytic perspective. These authors point out that the internal conflict is not only with society, but also with self. Basing their theory on Freud's idea of the ego ideal, they indicated that most women do not possess a uniform, ego-syntonic ideal or model for emulation and aspiration. The ego ideal is usually based on internal and cultural expectations to be like mother, and consequently, women must fight not only societal expectations, but strong internal expectations as well. The result of the struggle between ego and ego ideal is shame and a sense of inferiority on the part of the individual for having failed to achieve a goal set by the ego ideal. Katz and Shapiro (1976) clarify that the discrepancy may exist between a woman's goals and behavior or between her behavior and her experience of self.

The lack of congruence that can exist between a woman's role self-expectation and her actual gender role behaviors is elaborated on by Orlofsky and O'Heron (1987) in an article addressing implications of gender role for personal adjustment. These authors identified three levels of gender role: personality traits, behaviors, and attitudes, with the last corresponding to role self-expectations. The three levels might intuitively appear to be subject to socialization in similar ways; however, this study and others indicate that they are relatively independent of each other. These authors found that gender role attitudes did not appear to be related to self-esteem or adjustment, while gender role traits and behaviors were related to self-esteem and adjustment.

Greeley (1989), in an interesting analysis of data from a social survey collected through 1987, concluded that the clash between self-expectations and behavior does indeed result in distress. The author explored the declining morale noted among married women, as measured by the percent saying that they are "very happy." Through a series of statistical analyses, he concluded that working mothers who were born in the 1950s accounted for the greatest part of the noted decline in happiness. From other items on the marriage fulfillment survey, family satisfaction and job satisfaction, he concluded that the decline in happiness could not be attributed to these areas. In fact, the data showed that working wives had the highest job satisfaction in the population. Greeley concluded that the decline in "happiness" was probably the result of the cross pressures of trying to fulfill both job and home expectations well, and suggested that the pressure to do so had its source in the women's own high expectations.

Polasky and Holahan's (1998) study on expectations, mentioned previously, found that a discrepancy between women's own actual role and ideal expectations was associated with depression. Of interest is that this self-discrepancy resulted in depression, while discrepancy from societal roles was associated with anxiety, as noted above. Self-discrepancy theory, derived from contemporary self-concept theories, is based on the principle that individuals evaluate their actual selves in relation to their ideal/ought selves. To the extent that there is a difference between these two selves, individuals will experience affective responses such as depression, shame, guilt, and anxiety.

Given the importance of congruence, it would appear that freedom in choosing a role is an important component in achieving well-being for women. This is supported by an interesting study by Moen, Dempster-McClain, and Williams (1989). In a longitudinal study of a large sample of urban women, with data collected in two waves thirty years apart, they found that a greater number of roles was correlated with longevity. When the effects of specific roles on longevity were examined, the only role that showed significance was membership in voluntary organizations of activities. The authors offer various possible explanations for this finding, among them, that participation in an organization was an exercise of choice and control at a time when the emphasis on marriage and motherhood limited women's role choices. The authors suggest that choice of role involvement may be an important factor in studying the influence of roles on well-being. This conclusion is consistent with self-discrepancy theory.

Sources of Role Conflict for Women Missionaries

The findings from research done in a monocultural context appear relevant to women living in cross-cultural settings. However, there appear to be some unique characteristics to the cross-cultural setting, in particular the fact that the surrounding societal expectations may vary from those in the women's culture of origin. As with the general research in the area of women's roles, the sources of conflict can be grouped into societal and self-related sources of expectations.

Societal expectations. Role-expectations may be complicated for women overseas. Cultural factors may limit the extent of roles, or in some instances expand them. For example, in Islamic countries female missionaries may be limited to a ministry among females. Foyle (1987) suggested that single women working in cultures that hold very traditional views of women may find it easier to adapt to the situation than married women. Their singleness is an anomaly, and consequently they are considered social enigmas. In contrast, she suggests that married women have very strong role expectations imposed on them.

Self expectations. A second source of role conflicts is the tension between women's own expectations of what they could of should be doing, and the role they are in. Many of the women who come to the mission field are highly trained, very intelligent, and committed to the task they are to fulfill. They may feel a sense of responsibility regarding their stewardship of gifts and abilities. Yet often they are placed in roles on the basis of their gender rather than their abilities. Severe discrepancies between ideals and reality can easily lead to burnout (Chester, 1983) and discouragement.

Conflicts between role expectations and the actual role may be intensified by women's status as missionaries. Bowers (1984) stated that "the missionary role is a representative role; the missionary represents the Christian faith, the sending church, and his or her home country to the people of the country in which he or she serves. Generally speaking, the more representative a role is, the more pressure there is to fulfill role expectations" (p. 6).

Hypotheses

Three hypotheses were examined in the present study. The first two are related to congruence with societal role expectations. The third is relevant to congruence with self expectations.

(1.) Gender-role congruence theory, made relevant by the large gender-role differences often experienced by women overseas, was tested by the hypothesis that stress will increase, and well-being and self-esteem decrease, as congruence decreases between role and host culture expectations.

(2.) Gender-role congruence theory also suggests the hypothesis that incongruence from the role ideal endorsed by the culture of origin of missionary women may result in decreased well-being. Consequently, it was hypothesized that women who are exclusively homemakers will experience more stress than those who are more actively involved in the cross-cultural missions task, while women who have positions of authority outside the home will experience less stress, and greater well-being and self-esteem. This hypothesis is supported by Bowers' (1985) observation that women who are overseas primarily because of their husband's call have a harder time enduring the vicissitudes of missionary life, as well as the pertinent research on women's roles in society.

(3.) The role of self-expectations was examined by two hypotheses: (a) stress will increase, and well-being and self-esteem decrease, as congruence decreases between role and self-expectations; and (b) lower levels of role satisfaction and freedom in choosing a role will lead to greater stress and lower levels of well-being and self-esteem.

METHOD

Participants

Participants consisted of married missionary women who were currently serving overseas or had returned to their home culture within the past two months, and had served overseas for a minimum of one year. These requirements ensured that the missions experience had been long enough to permit awareness of the expectations, and recent enough to recall them clearly. These women represented a variety of evangelical mission organizations serving primarily in Latin America and in Muslim countries. In some cases, addresses were obtained from the organization, and the questionnaires were sent directly to each missionary. In other cases, the questionnaires were sent overseas in bulk, and were hand-distributed by a missions representative.

Of the 100 questionnaires that were distributed, 41 were returned. Four were excluded from the study, because they did not meet the criteria described above. All participants were married females. The age range for the 37 participants was 29 to 59 with a mean age of 38.65. Thirty of the women described themselves as Caucasian, two as Asian, and one as Hispanic. These women had been married for an average of 14.62 years, and had an average of 2.35 children. The average length of time that they had been at their current assignment was 5.63 years, and they had been working cross-culturally for an average of 10.53 years. In terms of Bowers' classification of roles (Appendix A), 19 of the women classified themselves as Homemakers, six as Background Supporters, eight as Teamworkers, and four as Parallel Workers.

Measures

A questionnaire was developed covering personal background information, missions experience, role information, and cultural information.

The actual roles in which women were functioning were assessed utilizing Bowers' (1984, 1985; see Appendix A) description of four categories in which women in missions can be classified: homemaker, background supporter, teamworker, or parallel worker. Each of the four roles was briefly described in the general questionnaire, and the respondent was asked to identify which of the roles described the primary role that she fulfilled.

Stress was measured with the General Well-Being Schedule (GWB; Fazio, 1977). This schedule contains 33 items. The first 14 items have Likert-type scales, the next four have rating bars on a scale of 0 to 10, and the last seven items are criterion-type behavioral and self-evaluation items. The six subscales of the GWB are Concerns About Health, Energy Level, Satisfying Interesting Life, Depressed-Cheerful Mood, Emotional-Behavioral Control, and Relaxed versus Tense-Anxious states. Scores can be derived for each of these subscales, as well as a total score. All scores are based on the first 18 questions, while the additional seven items provide important information regarding prior emotional distress and treatment. In this measure, a high score reflects a self-representation of well-being, while a low score indicates distress. Fazio (1977) found a test-retest correlation of .851 for the total scale, with a time interval of three months. The internal consistency of the test is also satisfactory, as shown by item to total scale score correlations of .912 for males and .945 for females. This high level of internal consistency indicates that the GWB is a homogeneous scale, with the subscales measuring different properties of the dimension tapped by this measure.

Self-esteem was measured with The Self-Esteem Scale (SES; Rosenberg, 1965). This scale was originally designed to measure adolescents' global feelings of self-worth or self-acceptance, but has since been applied to many other populations, becoming the standard against which new measures are evaluated (Robinson, Shaver, & Wrightsman, 1991). This measure consists of ten items that require that the respondent report feelings about the self on a four-point Likert scale ranging from Strongly Agree to Strongly Disagree. A total score is obtained by summing the responses to these items, resulting in a scale range of 10-40 with higher scores representing lower self-esteem, and lower scores indicating higher self-esteem. The internal consistency of the scale is acceptable, with Cronbach alphas ranging from .77 to .88 (Robinson, Shaver, & Wrightsman, 1991). Test-retest correlations of .85 have been obtained with a two-week interval between administrations.

Well-being was assessed with the Index of Well-Being (IWB; Campbell, Converse, & Rodgers, 1976). It is a two-part measure of self-reported well-being with life as currently experienced, and was developed with a national sample. The first part of the measure, known as the Index of General Affect, is a set of eight items on rating bars with a scale of one to seven, in which the respondent rates his of her life according to a set of contrasting adjectives. The second part of the measure is a single-item assessment of life satisfaction. A total score is obtained by weighting this last item 1.1 and adding it to the mean of the previous items. Scores can range from 0.0 to 14.7, with a low score indicating low well-being, and a high score indicating high well-being. The internal consistency of the scale is satisfactory, with a Cronbach alpha of .89 for the Index of General Affect. Test-retest correlations with a time interval of eight months were .56 for the Index of General Affect, and .43 for the one-item assessment of life satisfaction.

RESULTS

The first hypothesis was that stress would increase, and well-being and self-esteem decrease, as congruence decreased between role and host culture expectations. This hypothesis was not supported by the results. Congruence between the actual role and the expectations of the host culture were evaluated as follows: The role expectations of the host culture were determined through four items, in which the respondent indicated on a five-point Likert scale the degree of acceptability of each of Bowers' four roles for the host culture. Lower scores indicated greater acceptance of the role. Congruence was then evaluated by using the host culture's acceptance of the woman's actual role as a measure of congruence. Correlation coefficients were calculated between the role congruence score for the host culture, and the total scores for the GWB, SES, and IWB, as well as the subscales of the GWB. None of the resulting nine correlations were significant (see Table 1).

The second hypothesis was that women who were exclusively homemakers would experience more stress than those who were more actively involved in the cross-cultural missions task, while women who had positions of authority outside the home would experience less stress and distress, and greater self-esteem. This hypothesis was not supported by the results. This hypothesis was tested through a series of ANOVAs with tukey follow-ups, comparing the mean scores of women in the different role categories on each of the dependent variables. Two of the ANOVAs proved significant. The mean score of the Relaxed vs. Tense, Anxious subscale of the GWB was significantly higher for the Homemaker group than for the Background Supporter group, indicating that the Homemaker group was, in general, more relaxed than the Background Supporter group (F = 2.90, p < .05). The mean score of the IWB total score was also significantly higher for the Homemaker group than for the three other groups (F = 12.98, p < .00001), indicating that the Homemaker group reported higher well-being than the other groups. The ANOVAs are summarized in Table 2.

The third hypothesis, examining the role of self-expectations, consisted of two parts. The first was that stress would increase, and well-being and self-esteem decrease, as congruence decreased between role and self-expectations. This hypothesis was sup ported by the data. Congruence between role and self-expectations was assessed by comparing each respondent's description of the primary role she fulfills with her answer as to the role she would like to fulfill. In this way, two groups were formed: one in which the role and the role expectation were identical, and one in which there was a discrepancy between the actual role and the desired role. T-tests were run in order to compare the means between these two groups on the total scores for the GWB, SES, and IWB, as well as the subscales of the GWB. Of the nine t-tests, three proved to be significant, showing discrepancies between the means of the two groups in the GWB total score, as well as on the Satisfying, Interesting Life and Relaxed vs. Tense, Anxious subscales. In the three cases, the means for the congruent group were higher than those of the non-congruent group, indicating that women who were functioning in their desired role demonstrated higher satisfaction in terms of overall well-being, being satisfied with life, and being relaxed vs. tense of anxious. The remaining scales also showed a trend toward significance in the expected direction. The results of the t-tests are summarized in Table 3.

The second part of the third hypothesis was that lower levels of role satisfaction and freedom in choosing a role would lead to lower self-image and greater emotional and physical distress. Role satisfaction was assessed by a single item which asked the respondent to what extent she experienced personal fulfillment in taking on the role that she did, on a five-point Likert scale ranging from "not at all" to" completely." This item was correlated with each of the dependent variables to determine their relationship. Of the nine resulting correlations, two proved significant. The correlation coefficient between role fulfillment and the Emotional-Behavioral Control subscale of the GWB was r = .40 (p < .05), indicating that the greater the feelings of personal fulfillment in taking on the role, the greater the report of emotional and behavioral control. The correlation coefficient between role fulfillment and the SES total score was also significant (r = -.34, p < .05), indicating that the greater the feelings of personal fulfillment, the higher the woman's self-esteem. The correlation coefficients for this hypothesis are summarized in Table 1.

Freedom in choosing the woman's role was also assessed by a single item, which was answered on a five-point Likert scale ranging from "none" to "a lot." The statistical analyses revealed significant correlations for this item and the amount of fulfillment experienced (r = .33, p < .05), as well as for the GWB total (r = .50, p < .01), the Cheerful vs. Depressed Mood (r = .46, p < .01), Relaxed vs. Tense, Anxious (r = .50, p < .01), and Emotional-Behavioral Control (r = .47, p < .01) subscales of the GWB, the IWB total (r = .42, p < .05), and the SES total (r =-.47, p < .01). These results indicated strong support for this part of the hypothesis, with seven of nine correlations showing significance, and indicating that freedom in choosing a role correlated with fulfillment, general well-being, the absence of depression and anxiety, feelings of being in control of emotions and behavior, lack of stress, and higher self-esteem.

In addition to the measures addressing the hypotheses of the study, an additional item was included in the questionnaire to assess the extent to which the woman and her husband agreed on her role. This item, which is on a five-point Likert scale, from "not at all" to "completely," was significantly correlated with both the GWB total (r = .34, p < .05) and the Relaxed vs. Tense, Anxious subscale of the GWB (r = .41, p < .05). Spousal support also appears to be a factor in overall well-being and the absence of anxiety.

DISCUSSION

The present study is an extension of past research in women's roles, and is based on the assumption that expectations are central to an understanding of role-related issues. More specifically, the assumptions are that self-expectations and the expectations of others interact with the roles that women assume, and can be sources of stress or satisfaction. The present study explored the effects of multiple expectations on the well-being of married missionary women. The focus was on the expectations regarding a role, including both self-expectations and the expectations of the host culture. The hypotheses examined in this study stemmed both from the largely anecdotal material related to women in missions, and from the research in women's roles discussed above.

The first hypothesis was that stress would increase, and well-being and self-esteem decrease, as congruence decreased between role and host culture expectations. There are several possible reasons for the lack of support for this hypothesis. First, women missionaries may be prepared to expect a discrepancy between their role and the expectations of the surrounding culture. Second, women missionaries may tend, in general, to assume roles that are acceptable within the framework of their host culture, thus avoiding conflict in that area. The discrepancy scores may not have been broad enough in range, and the sample size may have been too limited, to provide sufficient statistical power and thus empirically explore this hypothesis. An alternative explanation is that, while the gender-role expectations of the host culture are occasional irritants and make good missionary stories, they really have little impact on women's overall well-being.

The second hypothesis was that women who were exclusively homemakers would experience more stress than those who were more actively involved in the cross-cultural missions task, while women who had positions of authority outside the home would experience less stress and distress, and greater self-esteem. This hypothesis was not supported by the data. On the contrary, the data indicated that the Homemaker group appeared to be more relaxed and to experience life as more satisfying and interesting than other groups who had more out-of-the-home involvement. It should be noted that many of the women in the Homemaker group indicated, through choice and through notes to the experimenter, that they considered involvement in this role temporary, and looked forward to moving into other roles as soon as their children became older. The additional finding that freedom in choosing roles appeared to be an important determinant in well-being and self-esteem is relevant here; the Homemakers in this study apparently chose this role, and did not feel constrained by it because they felt that other options were open to them at a later time. Past studies in other samples may have found contrasting results because of the confound of freedom of choice. In other words, freedom in choosing a role, and its influence on the attitude toward working exclusively in the home, may be more important in terms of its impact on well-being and self-esteem than the actual role behavior itself.

These findings, which contradict a stable pattern of research findings in the secular literature (e.g., Betz & Fitzgerald, 1987; Froberg, Gjerdingen, & Preston, 1986; Keith & Schafer, 1998; Lennon, 1998), also suggest the importance of subcultural factors in the area of gender-role expectations. Cross-cultural research has found important differences in the area of maternal employment (Gottfried, Gottfried, & Bathurst, 1995). Only meager data are available looking at the influence of religion on gender-role stereotyping. Two studies suggest that women of Jewish background are more strongly career oriented than are those of other religious backgrounds (Betz & Fitzgerald, 1987). Berz and Fitzgerald (1987) suggest that religion is related to gender-role expectations because religions influence the traditionality of gender role and life style attitudes. It is possible that the results of the present study reflect gender-role values of a conservative Christian population that encourage mothers of young children to remain at home.

The third hypothesis, examining the role of self-expectations, was supported by the data. The key findings were that higher congruence between self-expectations and actual role leads to greater feelings of well-being and lower reported distress. In particular, higher congruence of role with self-expectations was correlated with reports of finding life satisfying and interesting, and with reported feelings of relaxation and absence of anxiety. Role satisfaction and freedom in choosing a role likewise contributed to greater well-being and self-esteem, and decreased stress.

These findings, in combination with those mentioned above, would seem to suggest that congruence with self-expectations may be more important than congruence with the expectations of the host culture. While missionary women may find gender-role limitations a source of occasional frustration, it would appear that these societal sources of tension have a minimal impact on long-term well-being and self-esteem. Discrepancy from self-expectations, in contrast, appears to be detrimental to well-being. From a psychodynamic perspective, self-expectations are related to the ego ideal which is formed relatively early, and is consequently more central to one's experience of self than the expectations of others who are less central in one's life. Although the expectations of significant others, such as husbands or parents, were not evaluated extensively in this study, it would be interesting to see whether these would prove to be more influential than the expectations of the host culture in terms of the implications for well-being.

The results of this study appear to have some implications for mission organizations. On a very basic level, mission agencies can provide women with role guidance and the opportunity to examine their self-expectations that is ordinarily given to men entering the mission field, in order to enhance their options and thus their freedom in choosing a role. This might be particularly important at developmental transition points, such as when children start school or leave the home. Gish (1983) recommends on-site, inservice training in relevant areas as one possible contribution. For example, a woman whose children are leaving the home may need assistance in locating suitable work or training for new responsibilities.

A second implication is that, to maximize well-being and consequently effectiveness, mission agencies should allow women to make their own decisions regarding their work involvement outside of the home. It appears that self-expectations are an important determinant of well-being among women, and self-expectations will vary significantly from individual to individual.

Two areas stand out as potential topics for further study. In the first place, given the apparent lack of impact of the expectations of the host culture, it would seem important to study the impact of role expectations of those more intimately connected with married women, namely, their husbands. It would seem probable that the expectations of spouses would have a greater effect on variables such as well-being, satisfaction, and self-esteem. This is partially supported by the finding, mentioned above, that spousal support was correlated with well-being. Secondly, anecdotal reports of tensions related to gender-role expectation discrepancies with the host culture were not reflected in the outcome measures of well-being, stress, and self-esteem. Women appear to be coping with these differences adequately, but the way in which this occurs is not explored in the present study. Missionaries differ from immigrants in a number of ways, one of which is the relationship with the country of origin, and the differences might lead to differences in adjustment. For example, they may have the expectation of returning to their country of origin and consequently escaping from the host culture expectations. A study comparing sojourners, such as missionaries, with immigrants might reveal differences in the ability to adapt to attitudinal differences, and would contribute to an understanding of how this adjustment takes place.

Some limitations to the present study should be noted. The present study is correlational in nature, and consequently, the direction of cause and effect can only be speculated upon, but not demonstrated empirically. The small sample size and the convenience sampling which occurred may limit the generalizibility of the results, although attempts were made to utilize participants who would experience very different societal expectations than those in their culture of origin. Given these factors, as well as the virtual absence in the literature of studies on gender-role expectations for cross-cultural sojourners, this study should be considered preliminary, and attempts should be made to refine the hypotheses and replicate the results in future studies.

However, and in spite of these caveats, the present study appears to have contributed in several important ways to an understanding of role issues for women missionaries. Freedom in choosing a role emerged as a key factor for well-being, fulfillment, satisfaction, and self-esteem. Fulfillment in exercising a role was also important in terms of well-being. Congruence between the actual role and the women's self-expectations also appears to be important, while congruence with the expectations of the host culture did not emerge as crucial factors. The actual role involvement of women does not seem to be as important as factors such as of freedom of choice, fulfillment, and congruence with self-expectations.

The unexpected finding of better outcomes for homemakers is also important in that it contradicts a large body of research and suggests the importance of subcultural factors such as religious values as mediating variables. Betz and Fitzgerald noted in 1987 that research documenting the effects of religion on women's career development was sorely needed. Although research since that time has documented the impact of religious variables on college students' intention to engage in careers (Jones & McNamara, 1991; Morgan, 1987; Morgan & Scanzoni, 1987), this is, to the authors' knowledge, the first to look at the impact of religious variables in women actually engaged in careers. Future studies should explore this relationship more directly by comparing the relationships between employment status and outcome variables in both religious and nonreligious samples that have been matched on other relevant variables.

This issue is of more than simply academic interest. The oft-repeated finding that homemakers are more depressed than employed women has led some to strongly advocate that this role be discouraged by career counselors. For example, Betz and Fitzgerald (1987) wrote,

The persistence of the notion that nonemployment (i.e., being a housewife) is a legitimate option for adult women is a direct result of the traditional notion that a woman's place is in the home. We submit that this position is a sexist one as it is applied exclusively to women ... and that it is detrimental to their well-being, as it deters women from obtaining economic security over which they have firm control. (p. 90)

The finding of the present study that homemakers in a conservative Christian group are, in fact, better off in some ways than employed women suggests that congruence with strongly held values and the freedom to make choices consistent with those values are more important than employment status.

While many questions were left unanswered by the present study, it provides direction for future research that may ultimately contribute to a greater understanding of the dynamics of women missionaries and their families in their lives as sojourners.

APPENDIX A

Bowers' (1984) Classification of Married Women Missionaries' Roles

Homemaker: This woman is primarily a full-time wife and mother. Her main focus is on the home and the support and nurture of her family. She is the enabler of her husband in his work. She may have very young children and/or may teach her own school-age children.

Background Supporter: She actively supports her husband and his work. She is moderately involved in outside activities, many of which relate to her husband's assignment. Her main focus may be on work that can be carried out within the home, such as entertaining, listening/counseling, home classes, or language classes.

Teamworker: Her main focus is on a team work with her husband, and on helping to reach the goals of their combined ministry, whether working side-by-side of separately. She may be involved in a variety of activities, some of which relate directly to her husband's work. She may have part-time paid employment, but it does not detract from her sense of teamwork with her husband.

Parallel Worker: She sees her role as distinct from her husband's role. She may work within the same organizational structure as her husband's assignment. Both husband and wife are involved in the childrearing.

Table 1
Correlation Coefficients for Role Congruence of Host Culture and
Role Satisfaction, with Stress, Well-Being, and Self-Esteem

                     Congruence with       Role
                      Host Culture     Satisfaction

GWB
Total Score                .14             .24

Freedom from              -.21            -.09
health concern

Energy level               .10             .31

Satisfying,                .29             .17
interesting life

Cheerful vs.               .32             .18
depressed mood

Relaxed vs. tense,         .04             .21
anxious

Emotional-                 .27             .40 *
behavioral control

IWB                       -.22             .05

SES                       -.09            -.34 *

* p < .05

Table 2
Analyses of Variance Comparing Role Groups

                      Home-     Background      Team
Measure               Maker     Supporter      Worker

GWB
Total Score           76.37       63.67         70.62

Freedom from          11.84        9.67          9.37
health concern

Energy level          12.37       11.00         11.62

Satisfying,            6.05        4.67          6.75
interesting life

Cheerful vs.          17.95       15.83         17.62
depressed mood

Relaxed vs.           16.05       10.83         13.50
tense, anxious

Emotional-            12.11       11.67         11.75
behavioral control

IWB                   11.60        7.82          7.41

SES                   18.29       17.17         17.00

                     Parallel                Probability
Measure               Worker     F Value     of F Value

GWB
Total Score           76.50        1.39          .26

Freedom from          11.50        1.00          .40
health concern

Energy level          13.50         .34          .79

Satisfying,            7.00        1.94          .14
interesting life

Cheerful vs.          18.00         .58          .63
depressed mood

Relaxed vs.           13.75        2.90          .05
tense, anxious

Emotional-            12.75         .32          .81
behavioral control

IWB                    5.87       12.98          .00

SES                   16.50         .24          .87

Table 3
T-tests Comparing Congruent and Non-Congruent Groups on
Strees, Satisfaction, and Well-Being

                     Congruent-   Noncongruent    t-
Measure                Group         Group       Value

GWB
Total Score            78.20         67.06        2.52

Freedom from           11.70         10.00        1.33
health concern

Energy level           12.90         11.17        1.29

Satisfying,             6.80          5.24        2.75
interesting life

Cheerful vs.           18.35         16.59        1.57
depressed mood

Relaxed vs.            16.05         12.47        2.77
tense, anxious

Emotional-             12.40         11.59        1.31
behavioral control

IWB                     9.64          9.24         .38

SES                    16.62         18.82       -1.40

                      Degrees     Probability
Measure              of Freedom    of t Value

GWB
Total Score              35           .00

Freedom from             35           .10
health concern

Energy level             35           .10

Satisfying,              3S           .00
interesting life

Cheerful vs.             35           .06
depressed mood

Relaxed vs.              35           .00
tense, anxious

Emotional-               35           .10
behavioral control

IWB                      33           .35

SES                      35           .08

REFERENCES

Astin, H. S. (1984). The meaning of work in women's lives: A sociopsychological model of career choice and work behavior. The Counseling Psychologist, 12(4), 117-126.

Betz, N. E., & Fitzgerald, L. E (1987). The career psychology of women. Orlando, FL: Academic Press.

Bowers, J. M. (1984). Roles of married women missionaries: A case study, International Bulletin of Missionary Research, 8(1), 4-Z

Bowers, J. M. (1985). Women's roles in mission: Where are we now? Evangelical Missions Quarterly, 21(4), 352-360.

Campbell, A., Converse, P. E., & Rodgers, W. L. (1976). The quality of American life: Perceptions, evaluation and satisfaction. New York: Russell Sage.

Carter, J. (1999). Missionary stressors and implications for care. Journal of Psychology and Theology, 27(2), 171-180.

Chester, R. M. (1983). Stress on missionary families living in "other culture" situations. Journal of Psychology and Christianity, 2(4), 31)-3Z

Fazio, A.E (1977). A concurrent validation study of the NCHI general well-being schedule (Dept. of HEW Publ. No. HRA-78-1347). Hyattsville, MD: National Center for Health Statistics.

Foyle, M. (1987). Stress factors in missionary marriages. Evangelical Missions Quarterly, 23, 20-31.

Froberg, D., Gjerdingen, D., & Preston, M. (1986). Multiple roles and women's mental and physical health: What have we learned? Women & Health, 11(2), 79-96.

Gish, D. (1983). Sources of missionary stress. Journal of Psychology and Theology, 11(3), 236-242.

Gottfried, A. E., Gottfried, A. W., & Bathurst, K. (1995). Maternal and dual-earner employment status and parenting. In M. H. Born-stein (Ed.), Handbook of Parenting: Biology and Ecology of Parenting (Vol. 2) (pp. 139-160). Mahwah, NJ: Lawrence Erlbaum.

Greeley, A. (1989). The declining morale of women. Sociology and Social Research, 73(2), 53-58.

Jones, B. H., & McNamara, K. (1991). Attitudes toward women and their work roles: Effects of intrinsic and extrinsic religious orientations. Sex Roles, 24(1/2), 21-29.

Katz, C., & Shapiro, R.B. (1976). Sex-role identification and success. Contemporary Psychoanalysis, 12(2), 251-257.

Keith, P. M., & Schafer, R. B. (1998). Marital types and quality of life: A reexamination of typology. Marriage and Family Review, 27(1/2), 19-35.

Kerpelman, J.L., & Schvaneveldt, P. L. (1999). Young adults' anticipated identity importance of career, marital, and parental roles: Comparisons of men and women with different role balance orientations. Sex Roles, 41(3/4), 189-217.

Lennon, M. C. (1998). Domestic arrangements and depressive symptoms: An examination of housework conditions. In B. P. Dohrenwend (Ed.), Adversity, Stress, and Psychopathology (pp. 409-421). New York: Oxford University Press.

Moen, P., Dempster-McClain, D., & Williams, R. M. (1989). Social integration and longevity: An event history analysis of women's roles and resilience. American Sociological Review, 54, 635-64Z

Morgan, M. Y. (1987). The impact of religion on gender-role attitudes. Psychology of Women Quarterly, 11,301-310.

Morgan, M. Y., & Scanzoni, J. (1987). Religious orientations and women's expected continuity in the labor force. Journal of Marriage and the Family, 49, 367-379'

O'Donnell, K. (1987). Developmental tasks in the life cycle of mission families. Journal of Psychology and Theology, 15(4), 281-290.

Orlofsky, J. L., & O'Heron, C. A. (1987). Stereotypic and non-stereotypic sex role trait and behavior orientations: Implications for personal adjustment. Journal of Personality and Social Psychology, 52(5), 1034-1042.

Polasky, LJ., & Holahan, C. K. (1998). Maternal self-discrepancies, interrole conflict, and negative affect among married professional women with children. Journal of Family Psychology, 12(3), 388-401.

Robinson, J. P, Shaver, E R., & Wrightsman, L. S. (Eds.). (1991). Measures of personality and social psychological attitudes: Vol. 1. Measures of Social Psychological Attitudes Series. San Diego: Academic Press.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

Sederer, L, & Seidenberg, R. (1976). Heiress to an empty throne. Contemporary Psychoanalysis, 12(2), 240-257.

Tomlinson-Keasey, C. (1974). Role variables: Their influence on female motivational constructs. Journal of Counseling Psychology, 21(3), 232-237.

HALL, M. ELIZABETH: Address: Rosemead School of Psychology, Biola University, 13800 Biola Ave., La Mirada, CA 90639. Title: Associate Professor of Psychology. Degrees: BA, MA, PhD, Biola University. Specializations: Psychodynamic therapy; missions and mental health; women and work.

DUVALL, NANCY S.: Address: Rosemead School of Psychology, Biola University, 13800 Biola Ave., La Mirada, CA 90639. Title: Professor of Psychology. Degrees: BA, MA, PhD, University of North Carolina, Chapel Hill. Specializations: Psychodynamic therapy; integration of psychology and theology; psychotherapy and spirituality; missions and mental health.

Correspondence concerning this article may be addressed to M. Elizabeth Lewis Hall, PhD, Rosemead School of Psychology, 13800 Biola Ave., La Mirada, CA 92639. Email: [e-mail]