Parental emotional support, science self-efficacy, and choice of science major in undergraduate women.

Former participants in a high school enrichment program for girls interested in science careers were surveyed 1 to 2 years after high school graduation. After reporting their college major, they completed measures of science self-efficacy and quality of emotional bonds with parents. Of the 41 women, 5 were not enrolled in college. Those actually majoring in science (n = 23) reported significantly higher science self-efficacy than those who were undecided or had chosen nonscience majors (n = 13). Science self-efficacy was significantly negatively associated with recollections of fathers as having been highly controlling and likely to use a "love withdrawal" parenting style.

**********

In U.S. colleges and universities, women constitute only 37% of physical science majors and 18% of engineering majors and earn only about 12% of all doctoral degrees in engineering (U.S. Department of Education, 2000, 2001). These data underscore a serious national problem. The potential contribution to engineering and physical sciences of women, who make up more than half the U.S. population, remains largely untapped (National Science Foundation [NSF], 1990). Research suggests that girls and women are discouraged from choosing a career in science or engineering (as well as from choosing other nontraditional careers) by a range of psychosocial factors that include occupational sex role stereotypes, doubts about the feasibility of combining family and a science career, lack of positive female role models, low social support, stage of feminist identity, and lack of information about careers in science and technology (Betz, 2001; Fitzgerald & Harmon, 2001; Flores & O'Brien, 2001; Luzzo & McWhirter, 2001). The level of women's career aspirations also tends to decrease as they grow older (O'Brien, Friedman, Tipton, & Linn, 2000), perhaps because of the increasing salience of the barriers described above.

The central thesis of this study was that the influence of parents constitutes a powerful factor in the balance of these forces that encourage or discourage girls from pursuing careers in engineering or science. Two theories lead to somewhat contradictory predictions about the nature of this parental influence: attachment theory and Roe's (Roe, 1957; Roe & Siegelman, 1964) theory of career development. Roe maintained that persons are drawn to "people-oriented" careers because of early experiences of being raised in a warm, loving, nurturing family (Roe, 1957; Roe & Siegelman, 1964). Conversely, individuals who are drawn to careers in the hard sciences tend to come from an emotionally colder home environment. In terms of research support for Roe's model, Green and Parker (1965) found that boys who perceived either parent as supportive and warm did tend to choose person-oriented occupations, whereas girls who reported cold parental relationships tended to choose careers with low person-orientation. However, many other studies have failed to confirm Roe's predictions. For example, although Roe and Siegelman found a significant relationship between nurturance/love in childhood and person-orientation in a sample of college students, the same study compared a sample of engineers with social workers and found that the social workers reported significantly more stress and less affection in their families than did the engineers. During the last 20 years, the number of studies explicitly testing Roe's model has sharply declined.

In contrast, over the past 15 years, there has been an upsurge in research using attachment theory to gain a better understanding of healthy adult development (for reviews, see Lopez & Brennan, 2000; Mallinckrodt, 2000). According to Bowlby's (1988) theory, caregivers who respond with relative consistency and warmth to a young child's needs foster a sense of a "secure base" that permits the child to regulate anxiety while exploring novel surroundings. Recent applications of attachment theory to the realm of career development predict that the emotional warmth and parental responsiveness that promote a sense of a secure base for exploration in young children may also be important for encouraging adolescent girls to explore the full range of career options (Blustein, Prezioso, & Schultheiss, 1995). A secure base of attachment may be especially important for girls and women to explore initially unfamiliar career options in nontraditional, male-dominated fields such as science and engineering (Ketterson & Blustein, 1997).

A separate line of research suggests that, in addition to providing a secure attachment base for exploration, parental emotional responsiveness may be essential for promoting self-efficacy, especially for social outcomes (Mallinckrodt, 2000). Yet another line of inquiry has used self-efficacy theory (Bandura, 1977) to understanding influences on women's career choices and career development (for reviews, see Lent, Brown, & Hackett, 1994; Swanson & Gore, 2000). Self-efficacy is defined as the expectation that one can successfully execute the behaviors required to produce a desired outcome (Bandura, 1977). Self-efficacy influences not only an individual's initial choice of behavior but also the extent to which a person persists in the face of obstacles (Betz, 2001). Many studies have demonstrated that self-efficacy expectations affect the degree to which an individual persists in a given career (Betz, 2001; Diegelman & Subich, 2002). Perceived self-efficacy may influence (a) the likelihood of developing initial interest in a career area, (b) approach versus avoidance of various career choices, (c) career performance, and (d) career persistence (Betz, 2001; Betz & Hackett, 1983). Studies have repeatedly shown that self-efficacy has a significant effect on women's career aspirations and expectations (Betz, 2001). Women with high self-efficacy for nontraditional occupations may be more likely to commit to these work roles (Matzeder & Krieshok, 1998). Low self-efficacy is especially problematic with regard to women's choice of nontraditional careers, because it results in lower likelihood of their initially choosing such a career (Betz & Hackett, 1983) or, if chosen, of persevering to complete the requisite science and mathematics education (Lent, Brown, & Larkin, 1984).

Thus, attachment theory and social cognitive models of career development offer complementary perspectives on the role that parents can play in opening the full range of occupational choices to their daughters. Together, these two lines of research predict that parents who are emotionally warm and expressive, encouraging, and engaged, but who also permit gradually increasing autonomy for exploration, foster daughters who are the most likely to develop both a sense of self-efficacy and the attachment security that equip them to flourish in nontraditional careers for women. In addition, emotional warmth and encouragement from parents at an early age can directly influence achievement in math and science courses and, hence, overall career aspiration (Ferry, Fouad, & Smith, 2000).

Some research suggests that secure attachment of adolescent women to their mother and separation from both parents may have an important positive influence on career development (O'Brien, 1996). However, in an extension of their 1996 study, O'Brien et al.'s (2000) longitudinal work suggested that attachment to father may become more influential once a young woman enters college. One explanation for this is that fathers, compared with mothers, may be seen by daughters with high career aspirations as more of a career role model (Auster & Auster, 1981). Women with a strong attachment to fathers may tend to choose career paths that more fully use their abilities (O'Brien et al., 2000). Perhaps a balance between connection and autonomy with both parents throughout young adulthood may be the optimum environment to facilitate a young woman's career development and exploration of options (Fitzgerald & Harmon, 2001; O'Brien et al., 2000). Other research suggests that memories of parents' intrusive control; of their failure to permit age-appropriate autonomy; and of a "love withdrawal" pattern of discipline, in which parents threatened to withhold affection as a means of controlling behavior, are all associated with lower social self-efficacy and higher levels of interpersonal problems in college-age students (Mallinckrodt & Wei, 2000). A love-withdrawal parenting style may be thought of as the antithesis of unconditional positive regard (Swanson & Mallinckrodt, 2001). Thus, it appears that self-efficacy and optimum development are fostered by an attachment relationship in which parents frequently express encouragement and affection and refrain from intrusive control or threats to withdraw affection as a method of discipline (Mallinckrodt, 2000).

Although separate lines of theory and research point to the importance of attachment and self-efficacy in women's career choice, relatively few studies have examined both sets of predictors together. Therefore, the purpose of this study was to address the serious social problem of women's underrepresentation in science and engineering careers by surveying a sample of women who, when in high school, had expressed an interest in a career in science or engineering and had been selected for participation in a special program funded by the NSF to promote the science interests of young women. Among these former high school program participants, now all over the age of 18, we hypothesized that (a) self-efficacy for science academic tasks would be associated with their choice to follow up earlier interests and actually pursue an undergraduate major in science or engineering and (b) the level of self-efficacy for science-related educational requirements would be significantly positively correlated with emotional responsive parental bonds and negatively correlated with recalled bonds of intrusive control and the strongly conditional kind of parenting known as "love withdrawal." Finally, we explored whether bonds with mothers or fathers were relatively more strongly associated with science-related self-efficacy.

Method

Participants and Procedure

We began by obtaining the names of all 79 girls who had participated from 1997 to 1999 in three cohorts of an NSF-funded summer enrichment program hosted by a large midwestern public university. Participants had been nominated by their high school teachers on the basis of expressed interest and demonstrated aptitude in the physical sciences. They were selected based on grades in mathematics and science courses and a competitive essay about why they wanted to attend the program. The main goals of the program were to give hands-on science experience, provide positive female role models in the science professions, and provide a peer group of other young women interested in the sciences.

In December 2001, surveys were mailed to the last known address of these 79 former participants, who were by that time all over the age of 18. Surveys were returned unopened from 10 addresses because participants' families had moved and left no forwarding addresses. Three waves of surveys were mailed to the remaining 69 addresses. Each survey packet contained a $5 McDonald's gift certificate as an incentive. It is unclear how many of the 69 remaining participants received surveys, but 41 (59%) provided usable data. The mean age of these women was 19.22 years (SD = 1.04, range = 18-21). With regard to ethnic identification, 32 (78%) indicated White, 4 (10%) Asian American, 2 (4%) biracial, 1 (2%) Native American, 1(2%) Latina, and 1 did not respond to this item.

Instruments

Parental Bonding Instrument (PBI). The PBI (Parker, Tupling, & Brown, 1979) contains 50 statements describing parental behaviors that tap adults' memories of their parents' emotional responsiveness (measured on the Care subscale), and intrusive control (measured on the Overprotection subscale). Respondents are directed to make ratings based on "your earliest memories until you were 16 years old" regarding how similar the particular parent was to the descriptive statement. Respondents use a 4-point scale (1 = very unlike, 2 = moderately unlike, 3 = moderately like, 4 = very like). The PBI is arranged in two parallel forms of 25 items each to provide separate ratings of one's mother and father. Respondents who experienced more than one person in the role of mother or father were directed to rate the person "you believe had the greater impact on you." For each of the parallel forms (mother and father), 12 items make up the Care scale and 13 items make up the Overprotection scale. Retest reliabilities (3-week interval) of .76 and .63 were reported, respectively, for the Care and Overprotection scales. Split-half reliabilities of .88 and .79 were also obtained, respectively, for the Care and Overprotection scales. Evidence of concurrent validity was established through significant correlations of PBI scores with ratings made by independent judges who conducted personal interviews (Parker et al., 1979) and between self-ratings and interviews with mothers and siblings (Parker & Lipscombe, 1981). Internal reliability (coefficient alpha) for the present study ranged from .87 (Father Overprotection) to .95 (Mother Care).

Love-Withdrawal Scale (LWS). The LWS (Mallinckrodt & Wei, 2000; Swanson & Mallinckrodt, 2001) is a 20-item scale designed to supplement the PBI by capturing memories of parents' highly contingent style of expressing affection. Two parallel forms of 10 items each assess mothers and fathers separately. The LWS uses the same instructions and 4-point response scale as the PBI. Whereas the PBI Care subscale assesses a parent's general level of emotional warmth, the LWS assesses whether the affection that was expressed was contingent on "correct" behavior and whether parents withheld affection as a means of discipline (e.g., "My mother/father made me work very hard to earn [her/his] love and affection"). Retest reliabilities (2-week interval, n = 30) of .83 and .94. were reported for the Mother LWS and the Father LWS, respectively, with internal reliability (coefficient alpha) of .90 for both subscales (Mallinckrodt & Wei, 2000). Internal reliability (coefficient alpha) for the present study was .90 and .91, for mother and father subscales, respectively.

Self-Efficacy for Technical/Scientific Fields (SETSF). The SETSF (Lent et al., 1984) was developed to assess an individual's science self-efficacy. Items tap the belief that one can complete the education and training requirements for 15 science and engineering fields as well as perform the duties for these same 15 occupations. In the present study, only the two academic requirements subscales were used, because we reasoned that many respondents might not be familiar with the specific job duties of these occupations. The self-efficacy level for the Educational Requirements-Level (ER-L) subscale is calculated by summing the number of science occupations for which the participants believed they could complete the educational requirements. In contrast, the strength of self-efficacy beliefs about educational requirement, measured on the Educational Requirements-Strength (ER-S) subscale, is calculated by determining the average score on 15 individual, 10-point scales (1 = completely unsure, 10 = completely sure) that assess one's confidence in completing the educational requirements for each of the 15 specific fields. Thus, the ER-L describes the participant's belief in his or her ability to complete a task at all, whereas the ER-S goes beyond this to describe the strength of that self-efficacy belief for each individual item (Lent et al., 1994). The following test-retest correlations over an 8-week, post-follow-up period were reported by the Lent et al. (1984): for ER-L, .76, and for ER-S, .89. The coefficient alphas used to determine internal consistency reliability for ER-L and ER-S were .79 and .89, respectively (Lent et al., 1984). The ER-S scale also correlated significantly with the ER-L scale, r = .81 (Lent et al., 1984). In the current study, internal reliabilities (coefficient alpha) were .91 and .93 for the ER-L and ER-S scales, respectively.

Results

Of the 41 women who returned surveys, 5 reported being not currently enrolled in a postsecondary institution. Of the remainder, 31 were enrolled in a 4-year college or university, and 5 were enrolled in a community college or professional or technical school. Of these 36 women, 23 reported majoring in engineering or one of the physical or biological sciences, whereas 13 reported either majoring in one of the social sciences, humanities, or business (n = 11) or were undecided about their major (n = 2). Our first hypothesis held that science self-efficacy would be associated with choice of a science major. Two t tests were conducted to compare women who had chosen a science or engineering major (n = 23) with those who were undecided or had chosen other majors (n = 13). Table 1 shows that science majors were significantly higher in both science self-efficacy level and strength of self-efficacy beliefs than nonscience majors. The effect sizes (eta squared) for these group differences were 0.117 and 0.139, respectively. A post hoc power analysis indicated that this sample size (N = 36), with two-tailed tests and alpha = .05, yields statistical power of .55 and .62, respectively, for effect sizes of this magnitude.

Our second hypothesis held that, regardless of a respondent's choice of major, science-related self-efficacy would be positively correlated with the parental bond Care (Mother Care and Father Care) subscales and negatively correlated with the PBI Overprotection (Mother Overprotection and Father Overprotection) subscales as well as with love withdrawal. Results of Pearson product-moment correlations used to test this set of hypotheses for the entire sample of 41 women are shown in Table 2. Contrary to expectations, there were no significant correlations between mothers' parental bonds and self-efficacy. However, as expected, strength of science self-efficacy was significantly negatively associated with memories of father's overprotection (i.e., intrusive control) and love withdrawal. Thus, the second set of hypotheses received moderate support, in that there were significant correlations between one of the two measures of self-efficacy and negative emotional bonds (i.e., control and love withdrawal) for fathers. (It should be noted that all correlations in Table 2 were in the expected direction.)

Discussion

This study surveyed a group of young women who had, while in high school 2 to 4 years previously, expressed sufficient interest in and aptitude for a physical science or engineering career to be recommended by their teachers for a prestigious summer enrichment program sponsored jointly by the leading public university in their state and the NSF. Our intention was not to investigate the impact of this program (in part, because we could not sample an appropriate control group), but rather we were interested in examining possible factors related to the decisions of these young women to follow through on their initial interest by majoring in engineering or one of the physical or biological sciences.

As expected by our first hypothesis, self-efficacy for completing educational requirements for science and engineering careers was significantly higher among women who had chosen these majors than for women who had not. This result is perhaps not surprising and is congruent with a growing body of research on the role of self-efficacy in facilitating women's pursuit of nontraditional careers (Fitzgerald & Harmon, 2001; Lent et al., 1984). However, although it is to be expected that women who chose to major in science would have greater self-efficacy in those academic pursuits, perhaps of greater interest are the findings of associations between levels of self-efficacy and parental bonds. Although bonds with mothers were not significantly associated with self-efficacy, memories of fathers who were high in intrusive control and who used threats to withdraw affection as a frequent means of discipline were significantly negatively associated with self-efficacy for science career educational requirements. This love-withdrawal style of parenting is of critical importance when examining young women's efficacy for nontraditional fields of study. Love withdrawal is, in essence, the opposite of Rogers's concept of unconditional positive regard (Swanson & Mallinckrodt, 2001). Parents with a love-withdrawal style of parenting provide only conditional support. For young women to feel confident and comfortable pursuing nontraditional career paths, it is crucial that they receive unconditional positive regard to actualize their potential. The conditional expression of approval and affection that is part of a love-withdrawal parenting style may result in a young woman's decision to avoid these potentially more difficult career options, regardless of her skill or abilities. No doubt these negative parenting styles can result in lower self-efficacy for a broad range of other academic pursuits in addition to the science/technical fields. However, because the barriers to female participation in science/technical fields are especially high, love-withdrawal or other negative parenting styles that limit self-efficacy may be particularly problematic for these specific career fields. Contrary to our second set of hypotheses, memories of fathers' warmth and emotional expressiveness were not significantly positively associated with self-efficacy. Finally, Roe's (1957) model might be interpreted to predict a significant negative association between emotional expressiveness and self-efficacy. No support for this prediction was obtained.

These findings suggest a rather pessimistic conclusion, namely, that negative emotional bonds with fathers might do more to harm a daughter's developing sense of science self-efficacy than positive emotional bonds can do to promote this self-efficacy. These findings also suggest that the paternal bond, as opposed to the maternal bond, may be more influential in connection with science self-efficacy. Longitudinal research suggested a similar trend. O'Brien et al. (2000) reported that for adolescent women, mothers had a direct effect on career self-efficacy, but for women 5 years later (in college), fathers had the only significant direct effect on career self-efficacy. In college, the maternal effects appeared not to be significant. Our findings about the relative influence of fathers compared with mothers are also consistent with a number of other studies. The initial choice of college major for students of both sexes may be more heavily influenced by fathers than by mothers (Weishar, Green, & Craighead, 1981). O'Brien (1996) reported a correlation between emotional separation from father and important career constructs such as congruity of career choice. In Roe and Seligman's 1964 study, the female engineers identified with their fathers and reported good relationships. Relative freedom from conflict with the opposite-sex parent has been associated with positive vocational identity for both men and women (Lopez, 1989). However, there is some evidence that for women, attachment to and conflictual separation (i.e., "one's reported freedom from guilt, anxiety, mistrust, responsibility toward, or resentment of one's parents"; Blustein, Walbridge, Friedlander, & Palladino, 1991, p. 41) from both parents is significantly positively associated with career commitment (Blustein et al., 1991).

A number of methodological factors limit the strength of conclusions that can be drawn from this study. Although we surveyed all 79 women who participated in the NSF-funded program, the resulting sample of 41 women was small. Statistical power was correspondingly reduced, and the increased chance of Type II error should be considered carefully--especially before concluding that positive bonds with fathers, or bonds of any type with mothers, are not important in influencing daughters' science self-efficacy. Our effective return rate was no doubt somewhat higher than 59%, given that we used the addresses where students had lived while they were in high school 1 to 4 years previously. At least a few of the 28 surveys that were not returned may never have reached the intended recipient. Nevertheless, the selective return rate may have limited generalizability of these findings. Five women reported not being currently enrolled in college. Of course, at some point in the future, they may yet enter college. If our study was repeated in a few years, the results might be considerably different with respect to the variables that predict graduation and a satisfying career in science or engineering. The study relied entirely on self-report measures. Mood effects or other confounds based on common method error may have inflated correlations. In addition, according to Lent et al. (1984), the recommended method of deriving self-efficacy belief strength scores by taking the average of all 15 individual occupational scales penalizes respondents who indicate very high confidence scores in only a few areas.

Finally, both methodological strengths and weakness stem from our decision to sample only girls who attended a particular NSF-funded program rather than a broader sample of female students. First of all, this sample is unique in the fact that these young women were singled out by their teachers as having potential in the sciences. However, despite the recognized potential of these women, not all majored in the sciences. Thus, it is important to emphasize that women may choose to pursue (or not pursue) these nontraditional fields based not only on the influence of their parents but also based on the influence of their teachers. In the case of this particular enrichment program, the women who were selected had demonstrated scholastic aptitude in the sciences. For those who did not go on to major in science (which the program would not view as "successes"), there may have been other factors at work--namely, the attachment bond. Another issue related to this specialized sample is the fact that it was limited to a specially selected group of young women who are not at all representative of the general college population. However, it is exactly this type of young woman--that is, who has been singled out as having special potential for a science career by her high school teachers--that seems crucial to understand when studying the factors influencing choice of science majors. Our study suggests that even after participating in the carefully designed enrichment activities of this program, barely half (23 of 41) of the women in this specially selected cadre went on to major in a "hard" science or engineering. Despite the small sample size and other limitations, the findings of this study offer at least tentative suggestions for future researchers and career counselors. Parental attachments may have a powerful influence on self-efficacy and perhaps also on the career development of girls and young women. Much more research is needed to examine the specifics of both positive and negative influences of fathers and mothers on their daughters' career development. The present study tentatively suggests that relationships with fathers may be especially deserving of research scrutiny in connection with daughters' interest in science careers. The influence of parental bonds may have considerable potential to enhance or detract from the impact of even the best conceived interventions designed to address the serious problem of underrepresentation of women in nontraditional careers. Interventions like the program we studied may benefit from more direct interaction with parents, with special attention perhaps paid to involving fathers. For girls who lack an actively engaged father figure, other male relatives could be encouraged to become involved in their career development. Although the program provided young women with female role models in the sciences, it might have also been beneficial to connect participants with an encouraging male mentor with whom they could keep in touch after completion of the program.

It is more speculative to apply these findings to the general population of girls in middle and high school. Perhaps career guidance counselors can benefit girls' career development by providing parents (both mothers and fathers) with information on how they can best support their daughter's career aspirations and capacity to choose from a full range of fulfilling career paths. Our findings suggest that career counselors may need to operate at the level of the family system or the local community to promote mentoring, in addition to providing their traditional interventions. Career counselors interested in helping female adolescents may need to intervene to educate not only the young women themselves but also the people in their lives who can have the potentially greatest impact on career choice--their parents and extended family.

TABLE 1 Differences in Science Self-Efficacy Between Science,
Nonscience, and Undecided Majors

               Undeclared or
               Nonscience       Science Majors         Effect
Self-Efficacy  Majors (n = 13)    (n = 23)             Size
Measure        M     SD         M      SD       t(34)  [[eta].sup.2]

ER-L           8.85  5.56       11.91  3.26     2.09*  .117
ER-S           4.69  2.48        6.44  2.00     2.31*  .139

Note. N = 36. ER-L = Educational Requirements-Level subscale; ER-S =
Educational Requirements-Strength subscale.
*p < .05.

TABLE 2 Correlations of Parental Bonds With Science Self-Efficacy

Parental Bond     M     SD    ER-L    ER-S

Mother (n = 41)
 Care             2.58  0.61   .09     .15
 Overprotection   0.85  0.51  -.02    -.12
 Love-withdrawal  1.45  0.63  -.09    -.16
Father (n = 40)
 Care             2.40  0.67   .11     .23
 Overprotection   0.76  0.55  -.25    -.38*
 Love-withdrawal  1.44  0.60  -.21    -.32*

Note. ER-L = Educational Requirements-Level; ER-S = Educational
Requirements-Strength.
*p < .05.

References

Auster, C. J., & Auster, D. (1981). Factors influencing women's choice of nontraditional careers: The role of family, peers, and counselors. Vocational Guidance Quarterly, 29, 253-263.

Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215.

Betz, N. E. (2001). Career self-efficacy. In F. T. L. Leong & A. Barak (Eds.), Contemporary models in vocational psychology (pp. 55-77). Mahwah, NJ: Erlbaum.

Betz, N. E., & Hackett, G. (1983). The relationship of mathematics self-efficacy expectations to the selection of science-based college majors. Journal of Vocational Behavior, 23, 329-345.

Blustein, D. L., Prezioso, M. S., & Schultheiss, D. P. (1995). Attachment theory and career development: Current status and future directions. The Counseling Psychologist, 23, 416-432.

Blustein, D. L., Walbridge, M. M., Friedlander, M. L., & Palladino, D. E. (1991). Contributions of psychological separation and parental attachment to the career development process. Journal of Counseling Psychology, 38, 39-50.

Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. New York: Basic Books.

Diegelman, N. M., & Subich, L. M. (2002). Academic and vocational interests as a function of outcome expectations as posited in Social Cognitive Career Theory. Journal of Vocational Behavior, 59, 394-405.

Ferry, T. R., Fouad, N. A., & Smith, P. L. (2000). The role of family context in a social cognitive model for career-related choice behavior. Journal of Vocational Behavior, 57, 348-364.

Fitzgerald, L. F. & Harmon, L. W. (2001). Women's career development: A postmodern update. In F. T. L. Leong & A. Barak (Eds.), Contemporary models in vocational psychology (pp. 207-230). Mahwah, NJ: Erlbaum.

Flores, L. Y., & O'Brien, K. M. (2001). The career development of Mexican American adolescent women: A test of Social Cognitive Career Theory. Journal of Counseling Psychology, 49, 14-27.

Green, L. B., & Parker, H. J. (1965). Parental influence upon adolescents' occupational choice: A test of an aspect of Roe's theory. Journal of Counseling Psychology, 12, 379-383.

Ketterson, T. U., & Blustein, D. L. (1997). Attachment relationships and the career exploration process. The Career Development Quarterly, 46, 167-178.

Lent, R. W. Brown, S. D., & Hackett, G. (1994). Toward a unifying social cognitive theory of career and academic interest, choice, and performance. Journal of Vocational Behavior, 45, 79-122.

Lent, R. W., Brown, S. D., & Larkin, K. C. (1984). Relation of self-efficacy expectations to academic achievement and persistence. Journal of Counseling Psychology, 31, 356-362.

Lopez, F. G. (1989). Current family dynamics, trait anxiety, and academic adjustment: Test of a family-based model of vocational identity. Journal of Vocational Behavior, 35, 76-87.

Lopez, F. G., & Brennan, K. A. (2000). Dynamic processes underlying adult attachment organization: Toward an attachment theoretical perspective on the healthy and effective self. Journal of Counseling Psychology, 47, 283-300.

Luzzo, D. A., & McWhirter, E. H. (2001). Sex and ethnic differences in the perception of educational and career-related barriers and levels of coping efficacy. Journal of Counseling & Development, 79, 61-67.

Mallinckrodt, B. (2000). Attachment, social competencies, social support, and interpersonal process in psychotherapy. Psychotherapy Research, 10, 239-266.

Mallinckrodt, B., & Wei, M.-F. (2000, June). Attachment, social competencies, and characteristic patterns of interpersonal problems. In B. M. Straus (Chair), Relationship of attachment, interpersonal perceptions and problems. Symposium presented at the annual conference of the Society for Psychotherapy Research, Chicago.

Matzeder, M. E., & Krieshok, T. S. (1998). Career self-efficacy and the prediction of work and home role salience. Journal of Career Assessment, 3, 331-340.

National Science Foundation. (1990). Women and minorities in science and engineering. Washington, DC: Government Printing Office.

O'Brien, K. (1996). The influence of psychological separation and parental attachment on the career development of adolescent women. Journal of Vocational Behavior, 48, 257-274.

O'Brien, K. M., Friedman, S. M., Tipton, L. C., & Linn, S. G. (2000). Attachment, separation, and women's vocational development: A longitudinal analysis. Journal of Counseling Psychology, 47, 301-315.

Parker, G., & Lipscombe, P. (1981). Influence on maternal overprotection. British Journal of Psychiatry, 138, 303-311.

Parker, G., Tupling, H., & Brown, L. B. (1979). A parental bonding instrument. Journal of Medical Psychology, 52, 1-10.

Roe, A. (1957). Early determinants of vocational choice. Journal of Counseling Psychology, 4, 212-217.

Roe, A., & Siegelman, M. (1964). The origin of interests. Washington, DC: American Personnel and Guidance Association.

Swanson, J. L., & Gore, P. A., Jr. (2000). Advances in vocational counseling theory and research. In S. D. Brown & R. W. Lent (Eds.), Handbook of counseling psychology, (3rd ed., pp. 233-269). New York: Wiley.

Swanson, L. B., & Mallinckrodt, B. (2001) Family environment, love withdrawal, childhood sexual abuse, and adult attachment. Psychotherapy Research, 11, 455-472.

U.S. Department of Education, National Center for Education Statistics. (2000). Entry and persistence of women and minorities in college science and engineering education (NCES 2000-601). Washington, DC: Author

U.S. Department of Education, National Center of Education Statistics. (2001) Digest of education statistics (NCES 2001-034). Washington, DC: Author.

Weishar, M. E., Green, B. J., & Craighead, L. W. (1981). Primary influencers of initial vocational choices for college women. Journal of Vocational Behavior, 18, 67-78.

Anne B. Scott and Brent Mallinckrodt, Department of Educational, School, and Counseling Psychology, University of Missouri-Columbia. The authors thank Mary Heppner and Lisa Flores for their assistance and Meera Chandrasekhar for her cooperation and permission to contact the former participants in the program. Correspondence concerning this article should be addressed to Anne B. Scott, Department of Educational, School, and Counseling Psychology, 16 Hill Hall, University of Missouri, Columbia, MO 65211 (e-mail: [e-mail]).