Relationship Between Parental and Peer Attachment and Career Indecision

Amid the many challenges that today’s youth face, deciding on a career is one of great importance due to its wide-ranging and long lasting effects on life after graduation. Decisions about money, marriage, leisure, and friendship are just some of the factors that are affected by primary career decisions (Erikson, 1968; Feldman, 2003; Savickas, 1985, Super, Savickas, & Super, 1996). Due to this vital role of an early career choice, career indecision, which is an inability to make a decision about the vocation that one wishes to pursue, has been found to be a major source of anxiety and other negative emotions for young adults (Fuqua, Newman, & Seaworth, 1988; Fuqua, & Seaworth, 1987). Accordingly, research has drawn attention on finding the causes and solutions of these unwanted emotions that may occur with career indecision (Campagna & Curtis, 2007; Gati, Krausz, & Osipow, 1996) to better aid students in making satisfactory career choices.

Among methods of anxiety management, appropriate coping styles serve to keep anxiety within controllable limits and reduce the level of distress to aid career decision making (O’Hare & Beutell, 1987; O’Hare & Tamburri, 1987). In a study by O’Hare and Tamburri (1986), those students who experienced high trait anxiety, but did not use the appropriate coping skills such as establishing priorities and goals were likely to experience high state anxiety and were not likely to make a career decision. Not only has coping style been a focus as a moderator between anxiety and career decision making, variance in coping style has also surfaced as criterion of considering the heterogeneity of career undecided college students (Larson, & Majors, 1998). This is important because previous literature has proven that the “undecided” group is not a homogeneous one, but rather a heterogeneous one which needs classification or discrimination into subtypes (Larson, Heppner, Ham & Dugan, 1988). This need has risen because it has become vital that counselors be able to provide more informational and custom-made guidance according to each individual’s specific needs (Gati, Krausz, & Osipow, 1996). In other words, the demand for more tailored intervention strategies is high, as each individual’s profile differs even within the domains of career indecision (Fuqua, & Hartman, 1983; Kelly, & Pulver, 2003; Lucas, & Epperson, 1990; Sepich, 1987). Therefore, one of the purposes of the present study was to examine variables that are related to career indecision to better aid students in their career decision making.

Out of the many attempts in finding the causes and solutions for career indecision is a recent relational perspective that directs attention to the key role that interpersonal relationships play in human development (Bowlby, 1982; Josselson, 1992; Rosenbaum & Cohen, 1999). Relational theories assert that the primary human need is to develop and maintain close interpersonal relationships (Josselson, 1992), and unless this need is met, the individual would experience difficulty in carrying on other developmental tasks (Baumeister & Leary, 1995). Previous studies along these lines have found that interpersonal relationships were positively related to better health, reduced stress, and higher well-being among individuals (Broadhead et al., 1983; Cutrona, 1996). Prior studies have also found significant positive relations between interpersonal aspects and career development (Blustein, 2001, Blustein, Phillips, Jobin-Davis, Finkelberg, & Roarke, 1997; Way & Rossmann, 1996).

Among the theoretical framework that guides research on the benefits of interpersonal aspects on career domain is Bowlby’s (1982) attachment theory. Implications of attachment theory in the career domain suggest that individuals who have the accessibility and responsiveness of a trusted other would better cope with the difficulties expected during the career development process (Hazan & Shaver, 1990; Blustein et al., 1995). With regard to application of the career domain to attachment theory, empirical research has revealed the relations between attachment and career planning (Kenny, 1990), commitment to career choices (Blustein, Walbridge, Friedlander, & Palladino, 1991), and career self-efficacy (O’Brien, 1996).

A primary target of attachment in an individual’s life is parents. The central role of parents on young adults’ career development has been a focus of study in former career literature (Osipow, 1983; Roe, 1957; Super, 1957). Whiston (1996) found that both men’s and women’s career decision-making was positively related to how much familial support and encouragement they received. Still in another study using attachment theory, Lopez and Andrews (1987) suggest that young adults who are unable to psychologically separate from parents may suffer from continuous career indecision. Findings conducted with Korean young adults also suggest the helpful influence of parental attachment (Shim & Kang, 2007; Lee, 1997) in the career domain.

Along with parents, another essential target of attachment in an individual’s life is peers (Berndt, 1996; Josselson, 1992; Sullivan, 1953). Previous studies have acknowledged the strong beneficial influence peer and friends have on an individual’s development and growth (Harter, 1999; Rubin, Bukowski &Parker, 1998). Given the fact that young adults spend most of their time with their friends, it can be assumed that they would talk more about their career anxiety and options with close friends than their parents. Therefore, close friendships and peer attachment may be seen as a fundamental source of support during the career development process (Berndt, 1996; Felsman and Blustein, 1999). Altogether, preliminary studies have implied the meaningful relationships between parent, peer attachment aid young adults’ career development (Hazan & Shaver, 1990; Kettersone & Blustein, 1997), however, the influence of parent and peer attachment specifically on career indecision remains unexamined. Thus, the present study provides further understanding of the potential role of parent and peer attachment in the process of career indecision.

In addition to the role of attachment in an individual’s career development, recent vocational literature has also documented the relation of social support and career self-efficacy among young adults (Lent et al., 2001). Social support has been defined as information from others that assures one is being loved, cared for, and valued (Cobb, 1976; Cohen & Wills, 1985). It may come from parents, spouses, friends, colleagues, and other social networks. These different sources of social support seem to furnish different domains of optimal functioning such as educational self-efficacy and career decision-making (Kush & Cochran, 1993; O’Brien et al., 2000). Thus far, relational perspectives have offered explanation of the beneficial relation of social support and career development; however, there are no studies that specifically investigate the role of career-related social support in the career indecision. Therefore, the present study seeks to add the career-related social support variable to one of the interpersonal factors that is expected to affect coping with career indecision.

As mentioned above, one of the purposes of the present study was to examine variables that are related to career indecision to better aid students in their career decision making. In order to assess individuals’ career indecision, the Coping with Career Indecision scale (CCI), developed by Larson and his colleagues (1994), was used in the present study. The CCI is unique in that it complements the shortcomings of the most widely used career indecision measures such as the Career Decision Scale (CDS; Osipow, 1987; Osipow, Carney, Winer, Yanico, & Koschier, 1976), My Vocational Situation (MVS; Holland, Daiger, & Power, 1980), and the Career Factory Inventory (CFU; Chartrand et al., 1990). It was designed to include assessment of the appraisal process concerned in career indecision, a factor commonly overlooked in the current career indecision measures, however, key to forecasting the results of the career decision-making process (Larson et al., 1994). The appraisal process is deemed important because it alters an individual’s approach or avoidance of decision-making (Heppner, 1988), and problem-solving appraisal in career development (Larson & Heppner, 1985).

The CCI consists of four sub factors that assess an individual’s coping with career indecision. The first factor of CCI, Subjective Career Distress and Obstacles (SCDO), touches on the affective factor that illustrates some of the distress involve in coping with career indecision. It includes avoidance behavior that keeps one from becoming career decided such as helplessness, procrastination, self blame, negativity, and indecisiveness (Larson et al., 1994). The second factor, Active Problem Solving (APS) factor can be understood as opposite of SCDO in that it reports a positive approach-oriented behavior that helps one to becoming career decided. In other words, the APS seems to reflect one’s goal-driven appraisal to resolve their career indecision. Academic Self-Efficacy (ASE), the third factor in CCI, incorporates one’s belief in successfully completing the academic requirements in reaching a career goal. This may include actions such as choosing a major and continuing in the educational process. The final factor, Career Myths (CM) has been acknowledged as an obstacle to career decision-making (Heppner & Krause, 1979). Larson et al. (1994) note that among the four factors, the roles that ASE and CM play in deciding career indecision are subject to examination. Additional investigation is required to determine the usefulness of these factors in forecasting or depicting career indecision. In this light, it seems critical to separately investigate the four subtypes of coping with career indecision in order to find each type’s unique role and relation to the interpersonal variables of interest.

The earlier attention of the interpersonal perspective in the career domain has already produced numerous studies concerning the influence of parental attachment (Blustein, Prezioso, & Schultheiss, 1995; O’Brien, 1996; Vignoli, Croity-Belz, Chapeland, r. de Fillipis, & Garcia, 2005), peer attachment (Guay, Senecal, Gauthier, & Fernet, 2003; Felsman & Blustein, 1999) and social support (Kush & Cochran, 1993; O’Brien et al., 2000) on career development. However, there is a dearth of research that examines the roles these variables play in career indecision. Furthermore, given the fact that these variables have multiple causes and multiple effects, there is a need for research which uses a multivariate technique that allows simultaneous comparisons among these variables rather than conducting many statistical tests. In other words, it is critical that parental attachment, peer attachment, and social support be examined simultaneously as they are not variables that exist separately in an individual, but rather coexist in various combinations. Therefore, the purpose of the present study was to extend understanding of the interpersonal perspective of career development by investigating the relational matrix among mother, father, and peer attachment, career-related social support, and the four subtypes of coping with career indecision through canonical correlation analysis. In sum, it was hypothesized that mother, father, and peer attachment, along with career-related social support would be significantly related to coping with career indecision.

In addition, potential gender differences were also examined at the mean level of all variables and in the canonical correlational analysis. Sex differences among the mother, father, peer attachment, career-related social support, and the four subtypes of coping with career indecision were tested. Research on gender differences has shown that for females, high levels of both mother and father attachment were related to career development, while there has been inconsistent results for males (Blustein, Walbridge, Friedlander, & Palladino, 1991; Lee, 1999). However, the research does not report gender differences on the significance of peer attachment (Felsman & Blustein, 1999) in the career arena. In the present study, it was hypothesized that there would be significant gender differences among the relation of mother, father attachment to coping with career indecision, but none with peer attachment.

Method

Participants

One-hundred and ninety-five undergraduates at a four-year university in Seoul, Korea participated in the study as part of a fulfillment of course requirements. Ten cases were rejected due to missing data, resulting in a 185 sample for conducting the analysis. The final analysis sample consisted of 92 men and 93 women. The mean age of the sample was 22.39 years (SD = 3.2 years). The levels of final education for mother were as follows: 1.5% elementary school, 3.1% middle school, 40.5% high school, 43.6% university, 11.3% graduate school. The levels of final education for father were as follows: 0.5% elementary school, 1.5% middle school, 27.2% high school, 46.7% university, 24.1% graduate school.

Instruments

Coping with Career Indecision (CCI; Larson, Toulouse, Ngumba, Fitzpatrick, & Heppner, 1994).   The coping appraisal of career indecision was assessed by the Coping with Career Indecision (CCI; Larson et al., 1994). The Korean version, translated and revised in the study by Lee (2005) was used in the present study. Participants rated the extent to which they agreed or disagreed on each of the 35 items with scores ranging from 1 (never true) to 5 (always true). Though the original CCI is rated on a 6-point scale, a 5-point Likert scale was used, as it was deemed a more easily understandable measure of rating. This allowed a score range of 35 to 210 for the full scale. For convenience of interpretation, all items were reversely coded, indicating that higher scores depicted more positive appraisal to the career decision-making process.

The CCI consists of four subscales, that is, Subjective Career Distress and Obstacles (21 items), Active Problem-Solving (6 items), Academic Self-Efficacy (4 items), and Career Myths (4 items). Sample items from each subscale include, “I often feel down or depressed about selecting a major or career (Subjective Career Distress and Obstacles),” “I know the types of careers in which I could perform well (Active Problem-Solving),” “I am confident in my ability to succeed academically in the courses necessary to enter my chosen career (Academic Self-Efficacy),” “A career decision at this point is so important because it determines what I will be doing for the rest of my life (Career Myths).” Larson et al. (1994) reported an internal reliability coefficient of .89 for the full scale, and coefficients ranging from .69 to .90 for the subscales. In the present study, the internal consistency coefficient for the full scale was .85 and each factor was as follows: .85 for the Subjective Career Distress and Obstacles; .72 for the Active Problem-Solving; .67 for the Academic Self-Efficacy; and .40 for the Career Myth.

Career-Related Parent Support Scale (CRPSS; Turner, Alliman-Brissett, Lapan, Udipi, & Ergun, 2003).   Career-related social support was measured by the Career-Related Parent Support Scale (CRPSS; Turner et al., 2003). In order to expand the scope of the social support provider beyond parents, the “My parents” section of the items were changed to “There is someone who.” For example, “My parents tell me about the kind of work they do” was changed to “There is someone who tells me about the kind of work they do.” The Korean version, translated and revised by Kim (2004) was used in the present study. Participants rated the extent to which they agreed or disagreed on each of the 27 items with scores ranging from 1 (not at all) to 5 (very much so). This allowed a total score range of 27 to 135, with higher scores indicating more perceived career-related social support in the career decision-making process.

The CRPSS consists of four subscales, that is, Instrumental Assistance, Career-Related Modeling, Verbal Encouragement, and Emotional Support. Sample items from each subscale include, “There is someone who teaches me things that I will someday be able to use at my job (Instrumental Assistance),” “There is someone who tells me the kind of things that happen to them at work (Career-Related Modeling),” “There is someone who talks to me about what fun my future job could be (Verbal Encouragement),” “There is someone who knows I am sometimes scared about my future career (Emotional Support).” Turner et al. (1994) reported an internal reliability coefficient of .83 for the full scale, and coefficients ranging from .78 to .85 for the subscales. In the present study, the internal consistency coefficients for the full scale were .92, and each factor was as follows: .83 for the Instrumental Assistance, .87 for the Career-Related Modeling, .79 for the Verbal Encouragement, and .81 for the Emotional Support.

Inventory of Parent Attachment (IPPA; Armsden & Greenberg, 1987).  The Inventory of Parent Attachment is a subscale of the Inventory of Parent and Parent Attachment (IPPA; Armsden & Greenberg, 1987), and was used to measure the degree of attachment Korean undergraduates perceived regarding their parents. Mother and father attachment were measured separately with 25items each on a 1 (never true) to 5 (always true) scale. This allowed a total score range of 25 to 125 for each scale, with higher scores indicating more perceived attachment to mother and/or father.

The Inventory of Parent Attachment is constructed by three subscales, that is, Trust, Communication and Alienation. Sample items from each subscale include, “My mother/father accepts me as I am (Trust),” “I like to get my mother’s/father’s point of view on things I’m concerned about (Communication),” “My mother/father expects too much from me (Alienation).” The Korean version of the Inventory of Parent Attachment, translated and revised by Ok (1998) was used in the present study. Armsden and Greenberg (1987) reported an internal reliability coefficient of .86 to .91 for the subscales for the parent scale. In the present study, the internal consistency coefficient for the full mother scale was .92, and each factor was as follows: .84 for the Trust, .82 for the Communication, and .64 for the Alienation. The internal consistency coefficient for the full father scale was .93, and each factor was as follows: .89 for the Trust, .82 for the Communication, and .73 for the Alienation.

Inventory of Peer Attachment (IPPA; Armsden & Greenberg, 1987).  The Inventory of Peer Attachment is a subscale of the Inventory of Parent and Parent Attachment (IPPA; Armsden & Greenberg, 1987), and was used to measure the degree of attachment Korean undergraduates perceived regarding their peers. Peer attachment was measured with 25 items, on a 1 (never true) to 5 (always true) scale. This allowed a total score range of 25 to 125, with a higher score indicating more perceived attachment to peers.

The Inventory of Peer Attachment is constructed by three subscales, that is, Trust, Communication and Alienation. Sample items from each subscale include, “My friends are concerned about my well-being (Trust),” “When we discuss things, my friends consider my point of view (Communication),” “Talking over my problems with my friends makes me feel ashamed or foolish (Alienation).” The Korean version of the Inventory of Peer Attachment, translated and revised by Ok (1998) was used in the present study. Armsden and Greenberg (1987) reported an internal reliability coefficient of .72 to .91 for the subscales for the peer scales. In the present study, the internal consistency coefficient for the full peer scale was .90, and each factor was as follows: .85 for the Trust, .74 for the Communication, and .54 for the Alienation.

Procedures

With the approval of the Yonsei Psychology Departmental Review Committee for the purpose and procedure of the study, participants were recruited. Participants completed the aforementioned measures in addition to a form containing questions of demographic characteristics such as gender, age, parental education, and frequency of contact with close peers. The students were provided a consent form and assured of complete confidentiality. It was also explained that they could discontinue participation at anytime. Participants were told that the purpose of the study was to understand relational factors in career development. It took approximately 30 minutes to complete the whole procedure. The e-mail address of the researcher was given in case participants wanted results of the data analysis for the study.

Data analysis

First, a series of t-tests were used in order to examine whether significant sex differences among the mother, father, peer attachment, career-related social support, and the four subfactors of the CCI existed. Next, canonical correlational analysis was conducted to determine the relationships between the predictor (mother, father, and peer attachment) and criterion variables (subjective career distress and obstacles, active problem solving, academic self-efficacy, and career myths). The squared canonical correlation (R_c²₎ depicts the strength of the correlation between pairs of synthetic (or latent) scores based on the original variables. These synthetic scores are derived by the formation of linear variables (Stevens, 1996; Thomson, 1991). The number of canonical correlations in an analysis depends on the lesser number of variables in the two sets. Thus, the possible number of canonical correlations is three in the current analysis. The procedure of interpreting the canonical data in this study followed the step-by-step guidance of Sherry and Henson (2005). After determination of statistical significance with the R_c², significant values were then further inspected by both the standardized canonical coefficients and the canonical variate-variable correlations. These correlations provide the foundation for later interpretation of the synthetic variables, when the coefficients determine which variables are redundant. Those variables that are largest in magnitude for both the coefficients and correlations form the basis for interpretation of the canonical variate. Further, variables with a high correlation may be found to be redundant with other variables in the set of their standardized coefficient is small. Consequently, these correlations do not contribute to naming the canonical variate. Standardized coefficients and canonical variate-variable correlations above .30 were considered important for the interpretive process.

Results

Gender differences

Results of the series of t-tests are presented in Table 1. Findings show that a significant sex difference only existed in the peer attachment variable (t (191) = -1.447, p < .05). No other significant sex differences were found among the remaining variables.

Canonical correlation analysis with total sample

The means, standard deviations, and Pearson correlation for the total sample are reported in Table 2. Overall, the interpersonal variables and coping with career indecision styles were significantly correlated.

A canonical correlation analysis was conducted using the three attachment variables as predictors of the 4 coping with career indecision variables to evaluate the multivariate shared

relationship between the two variable sets (i.e., interpersonal attachment and coping with career indecision). The analysis yielded four functions with squared canonical correlations (R_c²) of .368, .083, .003, and .001 for each successive function. Collectively, the full model across all functions was statistically significant using the Wilks’s λ = .577 criterion, F(16, 541.38) = 6.664, p<.001. Because the Wilks’s λ represents the variance unexplained by the model, 1- λ yields the full model effect size in an r² metric. Thus, for the set of four canonical functions, the r² size was .423, which indicates that the full model explained a portion, about 42% of the variance of shared between the variable sets.

The dimensions reduction analysis allowed the researcher to test the hierarchal arrangement of functions for statistical significance. As noted, the full model (Functions 1 to 3) was statistically significant. Functions 2

to 4, 3 to 4, and 4 to 4 did not explain a statistically significant amount of

shared variance between the variables sets, F(9, 433.36) = 1.831, p=.061, F(4, 358.00) = .179, p=.949, and F(1, 180.00) = .164, p=.686.

Given the R_c² effects for each function, only the first function was considered noteworthy in the context of this study (36.8% of the shared variance). The last three functions only explained 8.3%, 0.3% and 0.1%, respectively, of the remaining variance in the variable sets after the extractions of the prior functions.

Table 3 presents the standardized canonical function coefficients and structure coefficients for Function 1. The squared structure coefficients are also given. Looking at the Functions 1 coefficients, one sees that relevant criterion variables were primarily active problem solving and subjective career distress and obstacles. This conclusion was supported by the squared structure coefficients. These coping with career indecision styles also tended to have larger canonical function coefficients. Furthermore, all of these variables’ structure coefficients had the same sign, indicating that they were all positively related.

Regarding the predictor variable set in Function 1, the CRSS variable was the primary contributor to the predictor synthetic variable, with a secondary contribution by father attachment. Because the structure coefficient for CRSS was negative, it was positively related to all of the coping with career indecision styles except career myths. Father

attachment was positively related to the coping with career styles, again except career myths.

Canonical correlation analysis with male sample

The means, standard deviations, and intercorrelations for the for the male sample are reported in Table 4. A canonical correlation analysis was conducted using the three attachment variables as predictors of the 4 coping with career indecision variables to evaluate the multivariate shared relationship between the two variable sets (i.e., interpersonal attachment and coping with career indecision). The analysis yielded four functions with squared canonical correlations (R_c²) of .355, .192, .020, and .009 for each successive function. Collectively, the full model across all functions was statistically significant using the Wilks’s λ = .507 criterion, F(16, 257.26) = 4.006, p<.001. Because the Wilks’s λ represents the variance unexplained by the model, 1- λ yields the full model effect size in an r² metric. Thus, for the set of four canonical functions, the r² size was .493, which indicates that the full model explained a portion, about 49% of the variance of shared between the variable sets.

The dimensions reduction analysis allowed the researcher to test the hierarchal arrangement of functions for statistical significance. As noted, the full model (Functions 1 to 3) was statistically significant. Functions 2 to 4 was also statistically significant, F(9, 207.02) = 2.404, p=.013. Functions 3 to 4, and 4 to 4 did not explain a statistically significant amount of shared variance between the variables sets, F(4, 172.00) = .630, p=.641, and F(1, 87.00) = .781, p=.379.

Given the R_c² effects for each function, only the first two functions were considered noteworthy in the context of this study (35.5% and 19.2% of shared variance, respectively). The last two functions only explained 2.0%, and .09% respectively, of the remaining variance in the variable sets after the extractions of the prior functions.

Table 5 presents the standardized canonical function coefficients and structure coefficients for Function 1 and Function 2. The squared structure coefficients are also given as well as the communalities (h²) across the two functions for each variable. Looking at the Functions 1 coefficients, one sees that relevant criterion variables were primarily active problem solving and subjective career distress and obstacles. This conclusion was supported by the squared structure coefficients.  These coping with career indecision styles also tended to have larger canonical function coefficients. Furthermore, all of these variables’ structure coefficients had the same sign, indicating that they were all positively related.

Regarding the predictor variable set in Function 1, the CRSS variable was the primary contributor to the predictor synthetic variable, respectively with a secondary contribution by father attachment. Because the structure coefficient for CRSS was negative, it was positively related to all of the coping with career indecision styles except career myths. Father attachment was positively related to the coping with career indecision style, again except career myths.

Moving on to Function 2, the coefficients in Table 5 suggest that the criterion variables of relevance were academic self-efficacy and subjective career distress and obstacles. These variables’ structure coefficients had the same sign, indicating that they were all positively related. As for interpersonal variables, father attachment was now the dominant variable, along with peer attachment. These variables were also positively related. Looking at the structure coefficients for the entire function, father attachment and peer attachment were positively related to academic self-efficacy and subjective career distress and obstacles.

Canonical correlation analysis with female sample

The means, standard deviations, and intercorrelations for the females samples are reported in Table 6. A canonical correlation analysis was conducted using the three attachment variables as predictors of the 4 coping with career indecision variables to evaluate the multivariate shared relationship between the two variable sets (i.e., interpersonal attachment and coping with career indecision). The analysis yielded four functions with squared canonical correlations (R_c²) of .416, .084, .033, and .000 for each successive function. Collectively, the full model across all functions was statistically significant using the Wilks’s λ = .517 criterion, F(16, 260.32) = 3.919, p<.001. Because the Wilks’s λ represents the variance unexplained by the model, 1- λ yields the full model effect size in an r² metric. Thus, for the set of four canonical functions, the r² size was .483, which indicates that the full model explained a portion, about 48% of the variance of shared between the variable sets.

The dimensions reduction analysis allowed the researcher to test the hierarchal arrangement of functions for statistical significance. As noted, the full model (Functions 1 to 3) was statistically significant. Functions 2 to 4, 3 to 4, and 4 to 4 did not explain a statistically significant amount of

shared variance between the variables sets, F(9, 209.45) = 1.185, p=.306,

F(4, 174.00) = .738, p=.567, and F(1, 88.00) = .000, p=.993.

Given the R_c² effects for each function, only the first function was considered noteworthy in the context of this study (41.6% of the shared variance). The last three functions only explained 8.4%, 3.3% and 0%, respectively, of the remaining variance in the variable sets

after the extractions of the prior functions.

Table 7 presents the standardized canonical function coefficients and structure coefficients for Function 1. The squared structure coefficients are also. Looking at the Functions 1 coefficients, one sees that relevant criterion variables were primarily active problem solving, academic self-efficacy, and subjective career distress and obstacles. This conclusion was supported by the squared structure coefficients. These coping with career indecision styles also tended to have larger canonical function coefficients. Furthermore, all of these variables’ structure coefficients had the same sign, indicating that they were all positively related.

Regarding the predictor variable set in Function 1, the CRSS variable was the primary contributor to the predictor synthetic variable, respectively with a secondary contribution by mother attachment.

Because the structure coefficient for CRSS was negative, it was positively related to all of the coping with career indecision styles except

career myths. Mother attachment was positively related to the coping with career indecision styles, again except career myths.

Discussion

The purpose of the present study was to extend understanding of the interpersonal perspective of career development by investigating the relational matrix among mother, father, and peer attachment, career-related social support, and the four subtypes of coping with career indecision through a multivariate analytic procedure. To reach this purpose, a canonical correlation analysis reflecting the relationships between interpersonal variables and subtypes of coping with career indecision were conducted. The results of this study reported previous relations between aspects of interpersonal relationships and career indecision, and introduced new findings pertaining to the relations between networks of interpersonal variables and coping with career indecision variables. The canonical correlations were statistically significant, allowing interpretation of the notable amount of variance in the findings.

The results of the study imply that the supportive relational matrix for Korean college students includes both career-related social support and parental attachment. The findings reported in this study, which lend support to previous literature on the effective role relational aspects play in career development (Blustein, 2001; Blustein, Phillips, Jobin-Davis, Finkelberg, & Roarke, 1997), suggest that students who receive career-related encouragement from social networks and have well established attachment with their parents are better able to deal with the difficulties of career indecision. In particular, when the interpersonal variables were considered along with those of coping with career indecision, the strongest contributors to the predictor and criterion synthetic variables were career-related social support and active problem solving, respectively. This finding lends support to previous research that has shown social support is strongly related to engagement in career decision-making (Kush & Cochran, 1993) and preparation in educational stages for specified career interests (Lapan, Hinkelman, Adams, & Turner, 1999). Those who scored high on career-related social support also showed higher scores on active problem solving. Considering that active problem solving is defined as a positive approach toward problem-solving behavior (Larson, Toulouse, Ngumba, Fitzpatrick, & Heppner, 1994) that will eventually lead to becoming career decided (e.g., “I know the types of careers in which I could perform well,” “I have a clear idea how to go about choosing a major (field of study) and selecting a career”), it can be inferred that students surrounded by people who help them choose and engage in educational activities, share their own work and vocational activities, encourage them to learn, and are thoughtful of positive/negative emotions concerning career decision (Turner et al., 2003), are going to be better informed and emotionally boosted to take on a proactive approach to pull oneself out of the state of career indecision. They may seek interpersonal relationships for support and buffer of anxiety that is inevitable in making important developmental decisions that are difficult to predict.

It is interesting to note that although there was no significant sex difference in the level of mother attachment (t(192)=-2.080, p=.283) or father attachment (t(192)=.248, p=.952), there was a difference in secondary contribution to the predictor synthetic variable between males and females in the canonical correlational analysis. Father attachment was the secondary contributor for the male’s predictor synthetic variable, while it was mother attachment for females. There was a greater gap between the contribution of mother and father attachment among the males than females. This indicates that while mother and father attachment probably share similar weight in influence on females’ coping with career indecision, father attachment distinctly has more influence than mother attachment on males’ coping with career indecision. The results for the females are consistent with previous studies that have revealed that high levels of both mother and father attachment was related to career development (Blustein, Walbridge, Friedlander, & Palladino, 1991; Lee, 1999).  Results for the males are consistent with the findings of Blustein et al., (1991) in that for males, father attachment was more predictive of progress in career development. It can be inferred that the father-son relationship seems especially important in the career development of young adult males. The present findings may also suggest that, for males, adopting one’s father’s beliefs may be a guide to better coping with career indecision.

The male sample was the only one that produced a significant second function. We can see that in Function 2, academic self-efficacy takes place of active problem solving in the criterion variable set and peer attachment replaces career-related social support. It can be inferred that when it comes to practical academic ability, the security provided by close friends is a more helpful aspect to coping with career indecision than a vague subject of social support. These results offer support to previous findings that suggest the sense of belongingness received from close peer relationships offer security and psychological support that facilitate commitment to academic commitment in the career domain (Ainsworth, 1989; Berndt, 1996). In sum, for the male students, the security provided close peer attachment may be an essential ingredient in the relational context that functions to increase academic self-efficacy.

In sum, it can be inferred that Korean college students are able to more actively solve career-related problems and feel less stress concerning the career decision-making process when emotional ties to fathers are healthy and when encouragement and practical career-related support from surrounding social networks is accessible. Regarding sex difference, similar suggestions may be made for the male students. Their levels of active problem solving increased while levels of career-related distress decreased when provided with strong emotional ties to father and career-related social support. In the case of female students, positive emotional ties to mothers and career support from surrounding networks increased levels of active problem solving and academic self-efficacy.

We can learn from the variables that peer attachment was only moderately useful in the model with the total sample. It only made a marginal contribution as a predictor, thereby suggesting that it may not be notably related to coping with career indecision in the present study. Past studies have revealed inconsistent results concerning peer relationships and career development (Felsman & Blustein, 1999; Guay, Senecal, Gauthier, & Fernet, 2003; Greenberg, Siegel, & Leitch, 1983). Findings from the study by Guay, Senecal, Gauthier, and Fernet (2003) report that support from peers plays a notable role in fostering the development of students’ career decision-making behaviors. In contrast, Felsman and Blustein (2003) discovered a rather modest amount of shared variance between peer relatedness and career development, suggesting caution while interpreting these relationships. The findings from this study are consistent with the latter study in that peer attachment had only moderate significance in the relations of interpersonal aspects and coping with career indecision. An interesting fact, however, was that females reported more peer attachment (t (191) = -1.447, p < .05) than males. This result lends support to previous research on gender differences that found that females reported higher levels of peer attachment than males, and that peer attachment partially mediated the relationship between parent attachment and life satisfaction only for females (Ma & Huebner, 2008). In sum, while parent attachment had significant roles in predicting career indecision, peer attachment had little play in the present study. Considering that Korean culture greatly emphasizes a child’s role of showing respect to elders and making efforts to make parents happy, it seems reasonable that attachment to parents has more significant influence on Korean college students’ career-decision making than attachment to peers.

Although the present study provides empirical results of a relationship between interpersonal variables and coping with career indecision, several limitations require caution in interpretation. First, the internal consistency coefficient for career myth within the CCI was a mere .40 in the present study. The internal consistency coefficient for this dimension scored a similar low in a study with other Korean college students as well (Lee, 2005), suggesting that the career myth factor is the most tentative. Additional research is required to examine the utility of this factor in predicting a range of career indecision among Korean college students. Second, due to the correlational design of the study, causality in the relationship cannot be inferred. Future longitudinal research is therefore required to test alternative hypotheses concerning the causal order among the variables of interest. Third, the generalizability of the findings is limited to Korean college students attending a high level university as the participants in the study. Further research should thus be implemented to see whether these findings apply to young adults with other levels of education. Another limitation is that the study was conducted with self-report measures, which warrants susceptibility of subjective interpretations of self by the participants. Futures research would benefit from qualitative methods to investigate a deeper nature of interpersonal relationships that seems to be important in coping with career indecision. Despite the limitations, the findings presented in the study provide a notable view of interpersonal aspects and coping with career indecision, resulting in theoretical assumptions and intervention approaches that incorporate these two domains of life.

Regarding counseling practice, the findings presented in this study suggest the benefits of interventions that connect interpersonal and career issues. Interventions and mentoring programs that provide sufficient career-related social support may be especially helpful for those students living far away from home and social networks. Moreover, counselors may want to encourage those students having adjustment difficulties to join groups or clubs that are available in order to enhance their interpersonal support systems.