Testing a model of men's nontraditional occupational choices. (Articles).

Researchers have compiled an extensive body of literature on the factors that influence the "traditionality," or the gender typical nature, of women 's career choices. However, little has been written about the variables that might influence men to enter gender atypical occupations. This study tested L. H. Chusmir's (1990) model of men's nontraditional occupational choice, using a longitudinal sample of college-age men in both gender traditional and nontraditional occupations. Liberal social attitudes, degree aspirations at Time 2, and socioeconomic status were directly predictive of nontraditional career choice, whereas the importance of a prestigious career at Time 1 and academic ability had an indirect influence on whether a traditional career choice was made.

**********

Although the labor market in the United States has traditionally been characterized by gender segregation, some men and women cross those gender lines and make gender atypical occupational choices. Literature in recent years has documented the increasing number of women who are entering male-dominated careers, stimulating an extensive body of research on the factors that may influence women's traditional career choices (Fassinger, 1990; O'Brien & Fassinger, 1993; Rainey & Borders, 1997). However, relatively little has been written about the factors that might influence men to enter female-dominated occupations (Chusmir, 1990; Hayes, 1989; Jome & Tokar, 1998; Lemkau, 1984; Tokar & Jome, 1998).

Whereas women who choose to enter male-dominated occupations are generally viewed as making a positive career move that offers increased opportunities for pay, advancement, and status (Hayes, 1986), the same perceptions do not hold true for men who enter female-dominated or nontraditional occupations. Men who enter traditionally female occupations face lower status and lower financial rewards (Chusmir, 1990; Hayes, 1989); they may also find their abilities, masculinity, and even sexual orientation questioned (Chusmir, 1990; Hayes, 1986). Yet, men do enter female-dominated jobs and are likely to continue to do so in increasing numbers as the labor market becomes more gender balanced (Fitzgerald & Cherpas, 1985).

Given the different perceptions of men and women who choose gender atypical occupations, it is likely that the variables that influence men's choice of gender atypical occupations are different than those that influence women's nontraditional occupational choices. An extensive review by Hayes (1986) summarized the reasons men gave for entering female-dominated occupations; these reasons included opportunities for less stressful occupations, increased options for other life choices, personal fulfillment, the greater economic stability of those jobs, and the increased opportunities for advancement to authority positions that might not be available in male-dominated occupations. These reasons imply a set of values and motivators that might be specific to men who choose nontraditional jobs. Nozik (1987) reported that nontraditional men put lower emphasis on achieving status than did men in jobs that are traditionally held by men; thus, the lesser importance of obtaining high status in prestigious jobs may also in fluence nontraditional occupational choices.

Chusmir (1990) proposed a model of men's nontraditional career choice that included personal influences, specifically background characteristics, attitudes, values, and needs; family influences of beliefs about family roles and relationship with children; and societal influences of gender role norms, modeling, and supportive others. A few studies have attempted to identify background or personal factors that differentiate career traditional men from career nontraditional men (e.g., Hayes, 1989; Lemkau, 1984; Mlott, Rust, Assey, & Doscher, 1986). These variables have included loss or absence of the male parental figure, socioeconomic status (SES), race, and educational level. Findings are equivocal regarding whether there are differences on those variables (Lemkau, 1984; Mlott et al., 1986).

Chusmir's (1990) model hypothesized that attitudes toward and participation in family roles would play an important role in nontraditional career choice. Kaufman (2000) reported that egalitarian women were less likely to intend to have children than women with traditional attitudes, whereas men with egalitarian attitudes were significantly more likely to intend to have children. Baruch and Barnett (1981) reported that nontraditional men participated more in child care than did traditional men. Although their classification of men as nontraditional was not based on career choice, it is plausible that men who place a high value on raising a family and perceive themselves as sharing in household and child-rearing tasks might choose female-dominated occupations, which are seen as allowing increased flexibility for child-rearing roles.

In addition, Chusmir's (1990) model emphasized the importance of societal factors on nontraditional career choice. Students experience many socializing influences when they are in college. College activities that are designed to broaden students' experiences and expose them to new cultures, ideas, and perspectives might affect their interest in nontraditional occupations. Parents are another important socializing influence. Hayes (1989) noted the importance of encouragement for continuing in school and the higher educational goals that the parents of nontraditional men had for these men. This support and modeling (especially by the mother) has been found to be a significant predictor of girls' career aspirations (Rainey & Borders, 1997); Selkow (1984) reported that elementary-aged girls and boys whose mothers were employed in nontraditional fields also tended to aspire to careers that were less gender typed.

Chusmir's (1990) model also included societal influences of gender role expectations. Several studies have examined gender role attitudes and reported differences on masculinity scores (Lemkau, 1984), androgyny scores (Lemkau, 1984), and sex role attitudes (Hayes, 1989). Jome and Tokar (1998) reported that career nontraditional men scored lower than career traditional men did on a measure of male role norms. They noted that findings reported in the literature showed consistently that men who had chosen female-dominated occupations had lower adherence to traditionally masculine gender roles.

Although gender role expectations can be seen as a socializing influence, views about the appropriate social roles of men and women can also be seen as attitudinal sets that influence nontraditional career choice. Career atypical men have been found to be less gender typed (Hayes, 1989; Lemkau, 1984). These men may be characterized as having more liberal gender attitudes that challenge traditional stereotyped perspectives of appropriate gender roles. These beliefs can be viewed from the attitudes and values component of Chusmir's (1990) model and can be seen as an important explanatory factor in men's nontraditional occupational choice. It is possible that liberal attitudes on a range of social issues (in addition to gender roles) might be a broader factor, of which nontraditional gender attitudes are only a component. In order to test this possibility, social attitudes as a predictor of traditional career choice were included in the model tested in the current study.

Unfortunately, studies that examined men's nontraditional occupational choice have been limited by samples in which only a limited number of female-dominated careers were represented (e.g., Galbraith, 1992), by comparisons of nontraditionally employed men with women in the same fields, and by cross-sectional designs that only allowed for a snapshot picture of men's occupational choices. The present study tested Chusmir's (1990) proposed model of men's nontraditional occupational choice, using regression-based path analysis; a longitudinal sample of men were assessed in their 1st year of college and 4 years later. It was hypothesized that background characteristics (e.g., perceived academic ability, SES, educational aspirations, importance of a prestigious job), societal influence variables (e.g., "traditionality," or the gender typical nature of the mother's career; participation in college activities that focused on diversity; ideological perspectives on social issues), and importance of raising a family wou ld predict the gender traditionality of male students' chosen careers. Men with more exposure to diversity experiences in college, to more liberal ideologies, and to mothers in less traditional careers who placed more emphasis on the importance of raising a family and less emphasis on the importance of a prestigious career would be more likely to choose a gender nontraditional career. Some of these effects (e.g., educational aspiration, importance of family, importance of prestigious career, social ideology) were hypothesized to be direct, whereas others (e.g., perceived academic ability, SES, mother's career, experiences in college) were hypothesized to be indirect and to influence traditionality of career choice through their impact on other variables. The hypothesized model is shown in Figure 1.

Method

Participants

Data for the present study were obtained from respondents to the 1986 and 1990 Cooperative Institutional Research Program (CIRP; Higher Education Research Institute, 1986) surveys. The overall sample consisted of 4,408 students who completed the initial CIRP survey when they began college in the fall of 1986 (Time 1); approximately 4 years later, they completed a follow-up instrument (Time 2). The sample for the current study consisted of 354 students who met the criteria for choosing traditionally male and female careers. One hundred and fifty-four men who represented the 10 most female-dominated careers (defined as having 52% or more women in those careers according to 1995 Bureau of Labor Statistics data) and 200 men who represented the 10 most male dominated occupations (defined as comprising 72% or more men) in the sample were selected for the analyses. The majority of the participants were White (89%) and of traditional college age; that is, 98% were 18 or 19 at the time of the first data collection.

Measures

The CIRP survey elicited a broad array of information on students' family background, high school experiences, initial goals and occupational aspirations, and personal characteristics. The follow-up survey collected information on involvement in college-related activities, extracurricular activities, attitudes, and family and occupational goals. Twelve distinct variables were identified and operationalized, using data available in the CIRP-86 and CIRP-90 database.

Career Choice

The dependent variable of participants' career choice at the time of the second data collection was categorized as either traditional or non- traditional, based on the percentages of men in those careers (Bureau of Labor Statistics, 1995), as previously specified.

Descriptive and Background Variables

Information about SES and mother's career was collected from the CIRP86. The value for SES reflects a composite of parental income and parents' education levels. Items were converted to z scores and summed to create the scale. The alpha coefficient for the composite SES scale was .60. The traditionality of the mother's career was determined by the percentage of women in that occupational field according to 1995 Bureau of Labor Statistics data. Jobs that are more traditional were those that had higher percentages of women employed in them.

Perceived Academic Ability and Educational Aspirations

A composite measure of perceived academic ability was created from self-ratings on academic, mathematical, and writing abilities; intellectual self-confidence; and average high school grades. Responses to ability ratings ranged from 1 (lowest 10%) to 5 (highest 10%); high school grade point averages ranged from 1 (D) to 8 (A or A+). Scores on all items were converted to z scores and summed. The internal reliability coefficient for the perceived academic ability (Ability) scale was .71. Educational aspirations at both data collection times were determined by asking participants at both the 1986 and 1990 data collection times to mark the highest level of education they planned to attain. Educational aspiration was conceptualized as interval level data: a low score of 1 represented no degree, and 6 represented doctoral or professional degrees.

Social Ideology and College Experiences

Measures of ideology toward social issues (Social) and college experiences (College) were created by summing relevant items on the survey. The Social scale was created by taking the mean score of 17 items that elicited students' opinions on a variety of social issues (e.g., equal opportunities for women, abortion, cohabitation, death penalty, and the prohibition of homosexual relations). Items ranged from 1 (disagree strongly) to 4 (agree strongly), with high scores indicating a politically liberal ideology. The College scale was formed by calculating the mean score of seven items (e.g., enrolled in women's study or ethnic studies course, participated in social or ethnic student organizations, participated in campus demonstrations, or attended racial awareness workshops). Items were given a code of 1 if students did not mark that they had engaged in the activity and were given a code of 2 if students indicated that they had engaged in the activity. Alpha coefficients were calculated for the newly created scal es; they were .59 and .76 for the College and Social scales, respectively.

Importance of Career and Family Goals

Calculating the mean score of five career-related goals and values at both data collection times created the variables of importance of a prestigious career (Prestige) at Time 1 and Time 2. On a 4-point scale (1 = not important to 4 = essential), participants rated the importance of items assessing financial success, recognition from colleagues, and having administrative responsibilities. The alpha coefficients for the Prestige scale were .69 and .70 for Time 1 and Time 2, respectively. At both administrations, participants responded to a single item (1 = not important to 4 = essential) asking about the importance of raising a family. Higher scores indicate a greater importance of family.

Data Analysis

Data were analyzed using regression-based path analysis. Path analysis is especially well suited for the analysis of nonexperimental data and, because of the composite nature of variables used in the present study (Keith, 1988), was more appropriate than approaches such as covariance structures analysis (e.g., LISREL; Joreskog & Sorbom, 1984). Regression-based path analysis differs from such approaches as LISREL in that it is more appropriate for exploratory testing of causal relationships, whereas "LISREL should only be used at a relatively late stage of a research program when 'crucial' tests of complex hypotheses are needed" (Kerlinger, 1986, P. 614). Chusmir (1990) noted that his framework was an untested model; as such, "until more is understood about the strength of each variable ... answers must remain tentative" (p. 12).

The background factors ability, SES, Time 1 assessment of degree aspiration, and the traditionality of mother's career choice were included as exogenous variables in the model. These variables represent both the personal influences component of the model and early societal influences of significant others. Time 1 measures of the importance of having a successful, prestigious career and the importance of raising a family were included to assess early personal emphases on career and family-oriented goals. Engaging in college activities that address a range of life experiences (e.g., participating in diversity-themed activities) assessed later societal influences that affect individuals as they progress through college. The last block included Time 2 measures of degree aspirations, the importance of family, and the importance of having a prestigious career; a measure of social attitudes was also included. Longitudinal data allow for the assessment of change that has occurred over time because of intervening expe riences. Time 2 measures of family importance, prestigious career, degree aspirations, and social attitudes take into account the effect of socializing experiences on individuals throughout their college years.

Results

It was hypothesized that social attitudes, the importance of a prestigious career and raising a family at Time 2, and educational aspirations at Time 2 would directly predict the traditionality of participants' career choice. The socializing influences of the mother's career and college experiences, background characteristics of perceived academic ability and SES, and early career and family goals were hypothesized to have indirect effects because of their influence on social attitudes and later family and career importance. Zero-order correlations were calculated to examine relationships among the model variables. The means, standard deviations, and correlation matrix are displayed in Table 1.

Path analysis was used to test the model because of its application to issues involving linear relationships and the need to determine the direct and indirect effects of predictor variables on criterion variables. Path coefficients were calculated for all possible relationships and are presented as standardized partial regression coefficients (beta weights). In the first step of the analysis, the criterion variable was traditionality of students' career choice; the predictor variables were all other variables in the model. This first step calculates the variables that are directly predictive of traditionality of career choice. Subsequent regression analyses were carried out in the same manner until all endogenous variables had served as the criterion variable for variables to the left of it in the model. These analyses allow for the later calculation of the indirect effects. Indirect effects were evaluated for significance, using the GEMINI (Wolfie & Ethington, 1985) statistical program.

In the first regression equation, the variables in the model explained 17% of the variance in career choice. Social attitudes, degree aspirations at Time 2, and SES were predictive of career choice ([beta] = .15, -.29, and -.17, respectively). The hypothesis regarding the direct prediction of career choice was partially supported. The direction of the beta weights indicated that more politically liberal attitudes were related to gender nontraditional occupational choice, whereas higher SES and higher degree aspirations were related to occupations that were gender traditional. No other variables directly predicted the traditionality of men's career choice. The subsequent regression equations indicated that college activities, a socializing influence, directly predicted social attitudes ([beta] = .3 1), degree aspirations at Time 2 ([beta] = .16), and prestigious career at Time 2 ([beta] = -.10). Mother's career, another socializing influence, directly predicted social attitudes. SES directly predicted the impo rtance of family at Time 1, prestigious career at Time 1, and degree aspirations at Time 2 ([beta] = .17, .12, and .13, respectively) in addition to its influence on career choice. Perceived academic ability predicted both later degree aspirations and family importance ([beta] = .18 and .12, respectively).

Standardized regression coefficients can be considered measures of effect size, and Pedhazur (1982) suggested that betas of larger magnitude could be viewed as meaningful even if they were not statistically significant. Although there is no standard of evaluating standardized regression coefficients, betas of .10 or greater are often used to indicate a substantial effect size. The variable importance of having a prestigious occupation had a beta of -.10 and, therefore, might have had an impact on career choice. The direction of the beta weight suggests that the greater importance of a prestigious career was related to choosing a gender traditional occupation. Standardized and unstandardized path coefficients for all possible relationships in the model are presented in Table 2. The tested model indicating significant paths and the additional path from prestigious career to career choice is displayed in Figure 2.

Although a variable may not directly influence an outcome, it may still have an indirect effect on the outcome variable through its impact on determinants of the outcome. These indirect effects must also be tested. Indirect effects (IB, expressed as betas) in this model were estimated using the GEMINI (Wolfie & Ethington, 1985) program. The GEMINI program analyzes recursive structural equation models and provides standard errors of all direct and indirect effects. The program estimates all of the equations in the model, calculates all of the total indirect effects implied by the model, and then calculates their large-sample asymptotic standard errors (Wolfle & Ethington, 1985). I hypothesized that the socializing influences of a mother's career and college experiences and background characteristics of perceived academic ability and SES and early career and family goals would have indirect effects on traditionality of career choice. This hypothesis was partially supported. Perceived academic ability (IE = -.05 , p < .05) and the importance of having a prestigious career at Time 1 (IE = -.07, p < .05) had significant indirect effects on traditionality of occupational choice. The influence of academic ability was mediated by degree aspirations at Time 2. The influence of prestigious career was probably carried by the later-measured importance of prestigious career, although this path was less clear because the direct effect of prestigious career at Time 2 did not attain statistical significance. There were no other indirect influences on traditionality of career choice.

Discussion

Chusmir's (1990) model of nontraditional career choice--which focuses on personal, family, and societal influences--seemed to be partially supported in the current sample because men with more ideologically liberal social attitudes were more likely to choose occupations that had higher percentages of women working in them. These social attitudes, including attitudes about gender roles, might indicate that these men were reference group independent (Wade, 1998) and allowed themselves to have greater role flexibility to choose occupations that seemed satisfying to them, even if those occupations have not typically been considered as appropriate for men.

Unlike Lemkau's (1984) finding that men in atypical occupations had more education than men in sex-typical occupations, in the current study, higher educational aspirations were related to choice of traditional careers. Lemkau's findings were based on a small sample of nontraditional and traditional career men, and few of the occupations in Lemkau's typical sample required graduate education. In the current sample, the gender traditional occupations were more likely to require advanced degrees than were the gender nontraditional careers; therefore, the relationship between degree aspiration and traditional career choice was not unexpected. Perceived academic ability was a significant indirect predictor of the traditionality of occupational choice, with higher academic ability predicting the choice of gender traditional occupations. Although Chusmir's (1990) model does not explicitly mention the role of academic ability, many other theories of career choice and development (e.g., Gottfredson, 1981; Holland, 19 85; Super, 1953) consider ability to be a key component of the career choice process. In the current study, perceived academic ability had no direct effect on traditionality of career choice, but rather it had an indirect effect through its impact on later educational aspirations and social attitudes. Academic ability is typically seen as a necessary precursor to the choice of many professional careers that require advanced degrees; thus, the relationship between ability and degree aspiration is a logical one. The negative relationship between ability and social ideology is a less obvious one. Perhaps there is a third variable that accounts for this relationship.

Likewise, the unexpected direct effect of SES on traditionality of career choice may be a function of the advanced educational requirements of many of the traditional careers. It is possible that men from a higher SES would be more able to afford the additional educational training for those careers than would men from a lower SES. It is also possible that men from higher SES backgrounds experienced greater societal expectations from family and friends that they would enter professional, well-paying careers. In the current study, these types of careers tended to be gender typical.

Results of my study showed that the importance of a prestigious career had a nonsignificant trend toward being predictive of traditional career choice when the effect size was considered. This finding is consistent with Nozik's (1987) finding that nontraditional men placed less emphasis on achieving status than did men in traditionally male jobs and suggests that nontraditional career men have lower needs to achieve high career status and may value other components of their occupations (i.e., collegiality, social outlets, altruism) more than status attainment. However, this finding must be considered as tentative and requires further research.

Chusmir (1990) hypothesized that men in nontraditional occupations would place a higher priority on family than would career traditional men. The current study found no direct or indirect relationship between the importance of raising a family and occupational choice. However, there were significant correlations among the importance of raising a family, degree aspirations, and prestigious career. It is possible that these young men sought a prestigious (and presumably well-paying) job to be able to support their families and did not, as has been discussed in women's career development (Betz, 1993; Betz & Fitzgerald, 1987; Farmer, Rotella, Anderson, & Wardrop, 1998; Stickel & Bonett, 1991), consider the goals of family and career success to be in conflict with each other. It is also possible that the measure of the importance of raising a family did not adequately assess the level of expected involvement in that role. Items assessing the type and extent of participation in the role, the anticipated time commit ment, and the value of child rearing might have more accurately measured the importance of the parental role and the effect it would have on occupational choice.

Contrary to expectations, the influence of societal role models assessed by the traditionality of mothers' careers did not have the expected indirect influence on the sons' choices of nontraditional careers. However, sons who had more liberal attitudes on a number of social positions were more likely to have mothers who worked in nontraditional careers; thus, perhaps mothers' career choices had an effect on the traditionality of sons' career choices that was not tested in the current study. Further research could explore this relationship more explicitly and expand the focus of mothers' behavior as role models, from the narrow specification of the traditionality of her own employment to a more inclusive assessment of her attitudes toward traditional roles and societal values.

Implications for Counseling

Men who are considering careers in gender nontraditional occupations are likely to be influenced by a combination of personal characteristics, family factors, and societal influences. My results indicate the importance of social attitudes, educational aspirations, SES, perceived academic ability, and prestigious career in predicting nontraditional career choice. Counselors are encouraged to assist their clients in examining these multiple factors when making career decisions.

As the job market changes and careers become available in sectors that have previously been seen as less open to men, career counselors must be prepared to assist men in considering a wider range of occupational options. Helping men in career counseling to understand how their social attitudes, values about prestigious career, and related issues may influence their career choices is a part of gender aware therapy (Good) Gilbert, & Scher, 1990). Evaluating the combination of work and family roles and considering how the roles are prioritized so that both roles can be more fully accommodated might help young male clients clarify their values. Raising a family and having a prestigious career were related in the current study, yet Good and Mintz (1990) noted that men who had greater conflict between work and family reported increased depression. Young men who have not yet started their own families might benefit from discussions about the time commitments related to parenting roles so that they can make more info rmed choices about combining their work and family roles.

Because of common stereotypes of men in nontraditional careers, counselors themselves must be aware of their own values regarding the choice of gender nontraditional work. If men who consider and choose these careers fear prejudice and ridicule (Chusmir, 1990; Haemmerlie, AbdulWakeel, & Pomeroy, 1985; Hayes, 1986), then counselors must be careful not to fall into the similar trap of assuming that the choice of gender nontraditional work implies a negative evaluation. Indeed, counselors may have to assist young men as they struggle with internalized messages about appropriate and inappropriate jobs.

Counselors who are interested in addressing potential barriers against men in nontraditional jobs might choose to take a more proactive stance and work with employers to address bias and discrimination in the workplace; these counselors might also form partnerships with men working in nontraditional occupations who could serve as role models, mentors, and information sources for clients. Interventions could be designed to increase students' (even those at the junior high school level) positive attitudes toward the suitability of nontraditional careers (Savenye, 1990).

Limitations

Although the current study improved on previous work by having a sample that had chosen a range of both female- and male-dominated occupations and by following them over a period of 4 years, it was also limited by the difficulty of finding a sample of men in nontraditional occupations. To achieve an adequate range of nontraditional occupations, a definition of 52% or more women employed in the occupation was used as the criterion for female-dominated careers. The percentage of women in the nontraditional occupations used in the current study ranged from 52 to 84, with a mean of 55. Thus, even though nontraditional participants were in female-concentrated occupational choices, they were not limited to occupations defined as gender atypical (e.g., nursing, elementary education, or child care occupations). In fact, none of the career nontraditional men in the sample had chosen nursing as an occupation. Many of the occupations selected by the current sample could be considered gender neutral rather than female do minated. This could weaken the evaluation of Chusmir's (1990) model. However, finding differences that predicted traditionality of career choice in the current sample suggests that even greater differences might be found when using a more conservative definition of gender atypical jobs. The difficulty in finding adequate numbers of men in gender atypical careers highlights the need for continued focus on this population. Future research should examine men who are employed in both gender neutral and female-dominated occupations. In addition, although this was a national sample, it was racially homogenous, and generalizations about men's gender nontraditional occupational choice should be confined to young White men because the variables that influence men of color to choose nontraditional occupations might be quite different.

Consistent with the procedure in previous studies that used national educational databases, scales for the current study were developed from items in the CIRP database and have not been subjected to rigorous test construction and validation. However, composite measures were internally consistent, face valid, and demonstrated some construct validity through interscale correlations in the expected directions. Nevertheless, future research could use measures, especially those concerning mother's influence and family importance, that are more detailed and have established psychometric data.

Summary

The personal and societal influence components of Chusmir's (1990) model received support in the current study because degree aspirations, social attitudes, SES, ability, and prestigious career were predictive of traditionality of career choice. However, the variables accounted for a limited amount of variance (17%), suggesting that other variables more proximal to career choice and perhaps improved measures of the variables included in the current study are needed to predict men's choices of nontraditional careers.

TABLE 1

Means, Standard Deviations, and Correlation Matrix for Model Variables
(N = 354)

Variables         M     SD    1    2        3     4        5        6

 1. Career      1.44   0.50      .14 **  -.04  -.10     -.31 **   .04
 2. Social      2.79   0.40              -.16  -.06      .06      .33 **
 3. Family2     2.98   0.94                     .24 **   .18 **  -.04
 4. Prestige2   2.43   0.63                              .09     -.09
 5. Degree2     5.55   1.66                                       .21 **
 6. College     8.06   1.27
 7. Family1     2.99   0.88
 8. Prestige1   2.60   0.56
 9. Degree1     5.33   1.49
10. MCareer    70.38  28.37
11. Ability     0.18   0.84
12. SES        -0.07   1.65

Variables         7        8        9       10       11       12

 1. Career      .05      .00     -.08     -.05     -.16 **  -.19 **
 2. Social     -.06     -.02      .10     -.16 **  -.07      .08
 3. Family2     .45 **   .17 *    .16 **   .12 *    .16 **   .13 *
 4. Prestige2   .07      .47 **   .05      .03     -.04      .07
 5. Degree2     .08      .11 *    .31 **   .02      .21 **   .20 **
 6. College    -.03      .02      .17 **  -.08      .06      .05
 7. Family1              .26 **   .10      .08      .08      .19 **
 8. Prestige1                     .09     -.03     -.04      .12 *
 9. Degree1                               -.03      .15 *    .17 **
10. MCareer                                         .03      .11 *
11. Ability                                                  .10
12. SES

Note. Carrier = traditionality of student's career; Social = social
attitudes; Family2 = importance of raising a family at Time 2; Prestige2
= importance of a prestigious career at Time 2; Degrees2 = degree
aspirations at Time 2; College = participation in deversity-related
college activities; Family1 = importance of raising a family at Time 1;
Prestige1 = importance of a prestigious career at Time 1; Degree1 =
degree aspirations at Time 1; MCareer = percentage of women in mother's
occupation; Ability = perceived academic ability; SES = socioeconomic
status.

* p < .05. ** p < .01.

TABLE 2

Direct Effect Coefficients for Model Variables

Variable       [R.sup.2]  1     2        3      4         5

 1. Career        .17         .18       .02    -.08    -.09
                             (.15 **)  (.04)  (-.10)  (-.29 **)
 2. Social        .15

 3. Family2       .23

 4. Prestige2     .23

 5. Degree2       .17

 6. College       .04

 7. Family1       .05

 8. Prestige1     .03

 9. Degree1       --

10. MCareer       --

11. Ability       --

12. SES           --

Variable          6          7          8         9          10

 1. Career       .02        .05        .05       .00       -.00
                (.05 **)   (.09)      (.06)     (.01)     (-.01)
 2. Social       .10       -.02       -.03       .01       -.00
                (.31 **)  (-.05)     (-.04)     (.05)     (-.14 **)
 3. Family2     -.03        .43        .11       .06        .00
               (-.05)      (.40 **)   (.06)     (.09)     (0.8)
 4. Prestige2   -.05       -.05        .53       .01        .00
               (-.10 *)   (-.07)      (.48 **)  (.02)      (.04)
 5. Degree2      .20        .01        .21       .24        .00
                (.16 **)   (.00)      (.08)     (.23 **)   (.02)
 6. College                -.07        .02       .14       -.00
                          (-.05)      (.01)     (.16 **)  (-.07)
 7. Family1                                      .00        .00
                                                (.06)      (.07)
 8. Prestige1                                    .03       -.00
                                                (.08)     (-.04)
 9. Degree1                                       --         --

10. MCareer                                       --         --

11. Ability                                       --         --

12. SES                                           --         --

Variable          11         12

 1. Career      -.05       -.05
               (-.09)     (-.17 **)
 2. Social      -.05        .02
               (-.10)      (.10)
 3. Family2      .13        .00
                (.12 *)    (.01)
 4. Prestige2   -.02        .01
               (-.02)      (.02)
 5. Degree2      .29        .12
                (.16 **)   (.13 *)
 6. College      .05        .03
                (.03)      (.04)
 7. Family1      .06        .09
                (.05)      (.17 **)
 8. Prestige1   -.04        .04
               (-.06)      (.12 *)
 9. Degree1       --         --

10. MCareer       --         --

11. Ability       --         --

12. SES           --         --

Note. Career = traditionally of student's career; Social = social
attitudes; Family2 = importance of raising a family at Time 2; Prestige2
= importance of a prestigious career at Time 2; Degree2 = degree
aspirations at Time 2; College = participation in diversity-related
college activities; Family1 = importance of raising a family at Time 1;
Prestige 1 = importance of a prestigious career at Time 1; Degree 1 =
degree aspirations at Time 1; MCareer = percentage of women in mother's
occupation; Ability = perceived academic ability; SES = socioeconomic
status.

* p < .05. (**)p < .01.

References

Baruch, G., & Barnett, R. (1981). Father's participation in the care of their preschool children. Sex Roles, 7, 1043-1055.

Betz, N. E. (1993). Women's career development. In F. L. Denmark & M. A. Paludi (Eds.), Psychology of women: A handbook of issues and theories (pp. 627-684). Westport, CT: Greenwood.

Betz, N. E., & Fitzgerald, L. F. (1987). The career psychology of women. New York: Academic Press.

Bureau of Labor Statistics. (1995). Employed persons by detailed occupation, sex, race, and Hispanic origin. Employment and Earnings, 42(1), 175-180.

Chusmir, L. H. (1990). Men who make nontraditional career choices. Journal of Counseling & Development, 69, 11-16.

Farmer, H., Rotella, S., Anderson, C., & Wardrop, J. (1998). Gender differences in science, math, and technology careers: Prestige level and Holland interest type. Journal of Vocational Behavior, 53, 73-96.

Fassinger, R. E. (1990). Causal models of career choice in two samples of college women. Journal of Vocational Behavior, 36, 225-248.

Fitzgerald, L. F., & Cherpas, C. C. (1985). On the reciprocal relationship between gender and occupations: Rethinking the assumptions concerning masculine career development. Journal of Vocational Behavior, 27, 109-122.

Galbraith, M. (1992). Understanding career choices of men in elementary education. Journal of Educational Research, 85, 246-253.

Good, G.E., Gilbert, L.A., & Scher, M. (1990). Gender aware therapy: A synthesis of feminist therapy and knowledge about gender. Journal of Counseling & Development, 68, 376-380.

Good, G. E., & Mintz, L. B. (1990). Gender role conflict and depression in college men: Evidence for compounded risk. Journal of Counseling & Development, 69, 17-21.

Gottfredson, L. S. (1981). Circumscription and compromise: A developmental theory of occupational aspiration. Journal of Counseling Psychology, 28, 545-579.

Haemmerlie, F. M., Abdul-Wakeel, A., & Pomeroy, M. (1985). Male sex bias against men and women in various professions. The Journal of Social Psychology, 125, 797-798.

Hayes, R. (1986). Men's decisions to enter or avoid nontraditional occupations. The Career Development Quarterly, 35, 89-101.

Hayes, R. (1989). Men in female-concentrated occupations. Journal of Organizational Behavior, 10, 201-212.

Higher Education Research Institute. (1986). Cooperative institutional research program codebook for 1986 freshmen survey and 1990 follow-up survey. Los Angeles: Author.

Holland, J. L. (1985). Making vocational choices: A theory of vocational personalities and work environments (2nd ed.). Englewood Cliffs, NJ: Prentice Hall.

Jome, L. M., & Tokar, D. M. (1998). Dimensions of masculinity and major career choice. Journal of Vocational Behavior, 52, 120-134.

Joreskog, K. G., & Sorbom, D. (1984). LISREL VI: Analysis of linear structural relationships by the method of maximum likelihood: User's guide. Morresville, IN: Scientific Software.

Kaufman, G. (2000). Do gender role attitudes matter? Family formation and dissolution among traditional and egalitarian men and women. Journal of Family 21, 128-144.

Keith, T. Z. (1988). Using path analysis to test the importance of manipulable influences on school learning. School Psychology Review, 17, 637-643.

Kerlinger, F. N. (1986). Foundations of behavioral research (3rd ed.). New York: Holt, Rinehart & Winston.

Lemkau, J. P. (1984). Men in female-dominated professions: Distinguishing personality and background features. Journal of Vocational Behavior, 24, 110-122.

Mlott, S. R., Rust, P. F., Assey, J. L., & Doseher, M. S. (1986). Performance of male nursing students on the MMPI, Fantasy, and self-esteem inventories. Psychological Reports, 58, 371-374.

Nozik, S. R. (1987). The relations of sex role self-concept and work values to traditional occupational choice among employed men. Dissertations Abstracts International, 47(7-B), 3094.

O'Brien, K. M., & Fassinger, R. E. (1993). A causal model of career orientation and career choice of adult women. Journal of Counseling Psychology, 40, 456-467.

Pedhazur, E. J. (1982). Multiple regression in behavioral research: Prediction and explanation (2nd ed.). New York: Holt, Rinehart & Winston.

Rainey, L. M., & Borders, L. D. (1997). Influential Factors in career orientation and career aspiration of early adolescent girls. Journal of Counseling Psychology, 44, 160-172.

Savenye, W. C. (1990). Role models and student attitudes toward nontraditional careers. Educational Technology Research & Development, 38(3), 5-13.

Selkow, P. (1984). Effects of maternal employment on kindergarten and first-grade children's vocational aspiration. Sex Roles, 11, 677-690.

Stickel, S. A., & Bonett, R. M. (1991). Gender differences in career self-efficacy: Combining a career with home and Family. Journal of College Student Development, 32, 297-301.

Super, D. E. (1953). A theory of vocational development. American Psychologist, 8, 185-190.

Tokar, D. M., & Jome, L. M. (1998). Masculinity, vocational interests, and career choice traditionality: Evidence for a fully mediated model. Journal of Counseling Psychology, 45, 424-435.

Wade, J. C. (1998). Male reference group identity dependence: A theory of male identity. The Counseling Psychologist, 26, 349-383.

Wolfle, L. M., & Ethington, C. A. (1985). GEMINI: Program for analysis of structural equations with standard errors of indirect effects. Behavior Research Methods, Instruments, & Computers, 17, 581-584.

Suzanne H. Lease, Department of Counseling, Educational Psychology and Research, The University of Memphis. Correspondence concerning this article should be addressed to Suzanne H. Lease, Department of Counseling, Educational Psychology and Research, The University of Memphis, Memphis, TN 38152 (e-mail: [e-mail]).