Secure Attachment and Career Indecision: The Mediating Role of Emotional Intelligence

Attachment and Career Indecision

Attachment refers to the emotional bond developed in the early child–parent relationship (Ainsworth, 1989; Bowlby, 1982). Bowlby (1982), who first introduced attachment theory, proposed that optimal psychological development occurs when infants have access to responsive caregivers who provide a trustworthy secure environment. In the context of these attachment experiences, children establish cognitive representations, or internal working models, of their relationships with others that for the most part are relatively stable across the life span (Cozzarelli, Karafa, Collins, & Tagler, 2003; Waters, Merrick, Treboux, Crowell, & Albersheim 2000).

Securely attached children typically develop a balanced sense of dependence and autonomy and the capability to self-regulate distressing emotions that extend into adulthood. In contrast, children who lack consistent external support and comfort often develop an insecure attachment orientation characterized by lack of trust in others and inability to effectively self-regulate distressing emotions. Research findings have provided support for the positive association of a secure attachment orientation to the ability to regulate emotions in children and adults (Borelli et al., 2010; Kim, 2005).

During the career decision-making process, individuals may experience fear of commitment (Blustein, Prezioso, & Schultheiss, 1995), uncertainty about the future, or perfectionism about making a choice (Gati, Amir, & Landman, 2010). However, an internalized sense of security from supportive parental relationships is likely to increase college students’ ability to manage negative emotions related to career choice and promote decision-making (Wright, Perrone-McGovern, Boo, & White, 2014). Consistent with this perspective, research findings with college students (Mojgan, Abdul Kadir, Noah, & Hassan, 2013; Tokar, Withrow, Hall, & Moradi, 2003) and adolescents (Emmanuelle, 2009) have indicated that students who report secure attachment orientations are less likely than their insecurely attached peers to report career decision-making difficulties.

Attachment and Emotional Intelligence

Emotional intelligence refers to “the ability to monitor one’s own and other’s feelings and emotions, to discriminate among them and to use this information to guide one’s thinking and actions” (Salovey & Mayer, 1990, p. 189). Emotional abilities may be arranged in four hierarchical domains that start with basic abilities and move toward more complex abilities: (a) accurately perceiving emotions in oneself and others; (b) using emotions to facilitate thinking; (c) understanding emotion, emotional language, and emotional signals; and (d) managing emotions to reach specific goals (Mayer & Salovey, 1997; Mayer, Salovey, & Caruso, 2008).

Consistent with attachment theory, secure individuals tend to utilize emotional management strategies that lower distress and enhance positive emotions, while insecure individuals are more likely to use emotional management strategies that perpetuate negative emotions and increase distress (Hazan & Shaver, 1987; Mikulincer & Florian, 2001; Shaver & Clark, 1994). Given that managing emotions is a key component of emotional intelligence, we can infer that secure individuals will likely demonstrate greater emotional intelligence abilities than insecure individuals. Several studies have reported links between attachment and emotion regulation-related components of emotional intelligence (Kafetsios, 2004; Kim, 2005; Mikulincer & Orbach, 1995). Compared to their less secure counterparts, securely attached individuals are more accurate in the perception and decoding of facial expressions of negative emotions (Magai, Distel, & Liker, 1995) and the perception of nonverbal messages from their partners (Kafetsios, 2004). Secure individuals also have demonstrated the capacity to regulate negative feelings (Feeney, 1995), avoid negative emotion bias (Magai et al., 1995), and access negative emotional memories without being overwhelmed (Mikulincer & Orbach, 1995).

Taken together, research findings suggest that securely attached college students are likely to demonstrate higher levels of overall emotional intelligence capacities than their less securely attached peers. Therefore, it seems reasonable to expect that among college students, secure attachment will be positively related to emotional intelligence. However, attachment and emotional intelligence may differ in their malleability. Only about 20–30% of adults are likely to report substantial changes in attachment orientation across time and factors associated with such changes have not been clearly identified (Cozzarelli et al., 2003; Fraley, 2002). In contrast, children, adolescents (Di Fabio & Kenny, 2011; Durlak, Weissberg, Dymnicki, Taylor, & Schellinger, 2011), and college students (Pool & Qualter, 2012) are responsive to emotional skills training. For example, an 11-hr intervention that included short lectures, role-plays, group discussions, and readings resulted in significant improvements in the understanding and managing of emotions among college students in England (Pool & Qualter, 2012). Therefore, an increased understanding of the extent to which emotional intelligence helps explain the observed association of attachment security to career indecision may provide useful avenues for the development of interventions to increase emotional intelligence and facilitate career decision-making among college students with unsecure attachment orientations.

The Present Study

Theoretical approaches and empirical evidence (Bowlby, 1982; Mikulincer & Florian, 2001) suggest that individuals with secure attachment orientations are generally skillful at using emotional management strategies that lower distress and increase positive mood, which are the hallmarks of emotional intelligence. In turn, the ability to manage negative emotions may facilitate career decision-making, as the process commonly causes an increase in anxious feelings (Brown et al., 2003; Di Fabio & Kenny, 2015; Di Fabio & Palazzeschi, 2009). While the relation of attachment to global indexes of career indecision is well established, we know less about the paths that link attachment to dimensions of career indecision. Therefore, the purpose of this study was to examine to what extent emotional intelligence, which involves awareness of emotions and capacity to integrate emotions with thoughts and actions (Goleman, Salovey, & Mayer, 1990), mediates the observed relation of secure attachment to each of the three domains of career decision-making difficulties identified by Gati et al. (1996). Unidimensional attachment, rather than attachment styles (secure, anxious, and avoidant; Braunstein-Bercovitz et al., 2012; Tokar et al., 2003), was considered in this study because the focus was on level of security and not on patterns of attachment. Prior studies have indicated gender differences in attachment (Mulligan & Lavender, 2010), emotional intelligence (Brown et al., 2003; McIntyre, 2010), and career-related factors (Mau, Perkins, & Mau, 2016; Migunde, Othuon, & Mbagaya, 2015). Based on these findings and the disproportionate number of female students in our sample, the study focused on female college students.

To our knowledge, the current study is the first investigation of the indirect effect of emotional intelligence in the negative association between secure attachment and career indecision. Consistent with previous bivariate correlational findings, it was predicted that secure attachment would be directly associated to emotional intelligence and to the three dimensions of career decision-making difficulties. It was also expected that emotional intelligence would mediate the relation of secure attachment to the three career indecision dimensions. While attachment orientation, including attachment security, tends to be stable across the life span (Cozzarelli et al., 2003), capacities associated with emotional intelligence seem to be malleable through intervention (Pool & Qualter, 2012). Therefore, in addition to providing a better understanding of the mechanisms that may help explain the observed negative relation of secure attachment to career indecision dimensions, results of this study may provide support for the inclusion of emotional intelligence training in career-decision-making interventions.

Method

Results

Data exploration procedures indicated low levels of missing data; 96–99% of cases had complete data in the measures of attachment and career indecision dimensions; 92% of cases had complete data in the emotional intelligence scale. None of the scales had more than 10% of the items missing; in most (82%) instances, cases were missing values for only 1 item per scale. Results of Little’s (1988) missing completely at random χ² test was not statistically significant (χ² = 89.983, df = 105, p = .85), which suggested that values were missing at random. In a study that compared various methods of handling missing data, Parent (2013) concluded that in data sets with low levels of item-level missingness, mean substitution and available item analyses (AIA) procedures yielded similar results (e.g., means, standard errors, and αs) compared to more complex multiple imputation procedures. Mean substitution refers to replacing a missing value in a scale with the mean of the respective participant’s nonmissing items in the said scale; AIA refers to using the mean of the available items on the scale to calculate a scale score. Given the characteristics of our data described above, we followed Parent’s (2013) recommendations in addressing missing values, which included a slight preference for AIA methods for scores based on the mean of the scale’s items. Because scores in the CDDQ (Gati & Saka, 2001) are obtained by averaging the items’ values in each subscale, the means of available items were used to calculate participants’ scores in each of the three career difficulties subscales. However, instructions for the calculation of scores for the attachment (Armsden & Greenberg, 1987) and emotional intelligence (Schutte et al., 2009) measures call for summing participants’ responses to each item. Therefore, mean substitution procedures were implemented before calculating the sum scores for the attachment and emotional intelligence measures.

A path model was tested using Mplus Version 7.3 (Muthén & Muthén, 1998–2015) to investigate (a) the direct association of attachment to the three dimensions of career indecision (lack of readiness, lack of information, and inconsistent information) and (b) whether emotional intelligence mediated the relation between attachment and each of the three dimensions of career indecision. As shown in Table 1, age was negatively related to career indecision; therefore, it was controlled for in the path model.

OPEN IN VIEWER

Table 1.

Correlations and Descriptive Statistics.

Variables	1	2	3	4	5	Mean	SD
1. Age						21.72	4.7
2. Attachment	.02					145.11	30.6
3. Emotional intelligence	.06	.39***				123.15	7.7
4. Lack of readiness	−.21**	−.28**	−.19**			4.59	1.23
5. Lack of information	−.16**	−.23**	−.29**	.52**		4.01	2.04
6. Inconsistent information	−.14*	−.22**	−.28**	.49**	.82**	3.62	1.88

Note. N = 331. Range of scores: attachment = 80–140, emotional intelligence = 33–165, readiness = 1.3–8.1, lack of information = 1–9, and inconsistent information = 1–9.

*p < .05. **p < .01. ***p < .001.

The following criteria were used to assess model data fit as recommended by Hu and Bentler (1999): comparative fit index (CFI), Tucker–Lewis index (TLI), standardized root mean square residual (SRMR), and root mean square of error approximations (RMSEA) along with its associated 90% confidence interval (CI). CFI greater than 0.95, TLI greater than .95, SRMR below .08, and RMSEA less than or equal to .06 indicate strong model data fit, and CFI close to 0.90, TLI close to .90, SRMR close to .10, and RMSEA close to .08 indicate acceptable model data fit. Following the examination of model data fit and direct effects within the path model, we examined the mediation hypotheses using bootstrapped CIs to test the indirect effects with 1,000 samples.

Results indicated a very good fit for the proposed path model: χ²(1, n = 362) = 1.377, p > .05, CFI = 1.000; TLI = .992; SRMR = .011; RMSEA = .032 with CI [.000, .148]. Figure 1 displays a visual representation of the path model including parameters for the direct effects. In terms of direct paths, as expected, attachment positively and directly predicted emotional intelligence (β = .431, p < .001). In turn, students’ emotional intelligence negatively and directly predicted lack of information (β = −.191, p < .01) and inconsistent information (β = −.196., p < .01). In addition, attachment negatively and directly predicted lack of readiness (β = −.222, p < . 001) and lack of information (β = −.119., p < . 05). However, the direct paths from attachment to inconsistent information (β = −.109, p < .10) and from emotional intelligence to lack of readiness were not statistically significantly (β = −.054, ns).

OPEN IN VIEWER

Figure 1.

Accepted model with standardized path coefficients. +p < .10. *p < .05. **p < .01. *p < .001.

The results also provided support for the positive relations among the three dimensions of career indecision. Lack of readiness was positively related to lack of information (β = .481, p < .001) and inconsistent information (β = .449, p < .001). Lack of information was also positively related to inconsistent information (β = .800, p < .001). Students’ age was a statistically significant control variable. Age was negatively related to lack of readiness (β = −.168, p < .05), lack of information (β = −.139, p < .01), and inconsistent information (β = −.117, p < .01) dimensions. In the accepted model, attachment accounted for 18.6% of the variance in emotional intelligence (R ² = .186). In addition, 9.8% of the variance in lack of readiness (R ² = .098), 9.4% of the variance in lack of information (R ² = .094), and 8.6% of the variance in inconsistent information (R ² = .086) were accounted for by attachment and emotional intelligence while controlling for students’ age.

Mediation tests using bootstrapped standard errors yielded statistically significant indirect effects which indicated that emotional intelligence mediated the relationship between attachment and lack of information (β = −.082, p < .01), 99% bootstrapped CI [−0.010, −0.002]; emotional intelligence also mediated the relation between attachment and inconsistent information (β = −.084, p < .01), 99% bootstrapped CI [−0.010, −0.003]. In contrast, the indirect effect tests did not provide support for the role of emotional intelligence as a mediator of the associations of attachment to lack of readiness (β = −.023, ns, 99% CI [−0.004, 0.001]).

While the proposed model provided a good fit to the data, other mediation models may also fit the data (Frazier, Tix, & Barron, 2004). Given the study’s cross-sectional design, the relation between emotional intelligence and career indecision dimensions may be reciprocal. To test this possibility, we examined an alternative path model in which the three career indecision factors mediated the relation of attachment security to emotional intelligence, while controlling for age. Based on the fit indices, this alternative model, χ²(6, n = 362) = 506.97, p < .001, CFI = 0.217; TLI = −.856; SRMR = .202; RMSEA = .48 with CI [0.445, 0.516], did not meet the criteria for good fit. These findings provide support for the hypothesized indirect effect of emotional intelligence in the relation of attachment security to career indecision difficulties among college women.

Discussion

The purpose of this study was to investigate to what extent emotional intelligence mediated the relation of secure attachment to career indecision among college women. While previous findings suggest that secure attachment is negatively linked to career indecision (Bowlby, 1982; Mikulincer & Florian, 2001), the extent to which other variables explain the relation of attachment to dimensions of career indecision has been less studied. The proposed mediation path model demonstrated very good fit. With two exceptions, results provided support for the study’s hypothesis regarding the direct and mediated links in the model; all statistically significant paths were in the expected direction.

When controlling for all other variables in the model, attachment security was directly linked to emotional intelligence and to lack of career decision-making readiness; however, emotional intelligence was not directly linked to lack of readiness. Conceptually, the lack of readiness dimension approximates the construct of career indecisiveness, which captures difficulties that thwart students’ engagement in the career decision-making process, including lack of motivation, general indecisiveness, and irrational expectations. Researchers have proposed that career indecisiveness is linked to stable personality characteristics, while career indecision is a normative developmental stage that individuals typically navigate successfully by acquiring information about the self, the world of work, and the career decision-making process (Gati, Gatassi, et al., 2010; Tokar et al., 2003). In support of this perspective, personality factors (extroversion, neuroticism, and trait anxiety; Di Fabio et al., 2013; Mojgan et al., 2013) have been more strongly correlated to career indecisiveness than to career indecision. From this perspective, it makes sense that in the hypothesized model, attachment security, which is a relatively stable personality disposition, was directly linked to lack of career decision-making readiness while emotional intelligence, a more malleable disposition, was not. As expected by the results of the direct paths examined above, tests of indirect effects did not provide support for the role of emotional intelligences as a mediator of the relation of attachment security to lack of career decision-making readiness.

Consistent with previous research findings, attachment was linked directly to the lack of information dimension, which captures lack of knowledge about the self, occupations, and the career decision-making process. However, the link between attachment and inconsistent information, which refers to confusing information about the self, occupations, and conflicts with significant others, was not statistically significant. This last finding is difficult to explain because based on previous findings, one would expect that secure attachment would facilitate the integration of contradictory information in self-exploration and career exploration (Emmanuelle, 2009; Tokar et al., 2003). However, as predicted, emotional intelligence served as a mediator in the relation of attachment to both lack of information and inconsistent information career decision-making difficulties. An alternative path analytic model that posited career indecision as a mediator in the relation of attachment to emotional intelligence was not a good fit to the data, which suggests that the relation of emotional intelligence to career indecision difficulties is not reciprocal.

In sum, findings indicated that only attachment was associated to higher levels of career indecision due to lack of motivation and indecisiveness, while emotional intelligence served as a mediator in the relation of attachment to career indecision associated to both information deficits dimensions. Results of the study provide support for the notion that different antecedents may explain career decision-making difficulties, and therefore, college women may require diverse intervention approaches (Gati, Amir, & Landman, 2010, Kelly & Lee, 2002). Interventions that focus on learning about the self, occupations, and how to manage negative emotions during the career exploration and decision-making process may be helpful to college women presenting with information-related career difficulties. In contrast, college women presenting with career indecision associated to lack of attachment security will likely require, in addition to career-related information, interventions that address how affective disposition rooted in early relationships may interfere with the career decision-making process (Downing & Nauta, 2010; Kelly & Lee, 2002). For example, an insecurely attached client may be overly anxious about meeting a career counselor and hesitate to pursue career services on campus. In addition, insecure individuals tend to have difficulty interacting with peers and faculty, many of whom can be important interpersonal resources. Therefore, it would be important for the counselor to assist the client to establish new reattachment experiences by providing psychoeducation on interpersonal skills and allowing the client to practice these skills in a safe environment.

Empirical findings suggest that emotional intelligence can be improved through training (Pool & Qualter, 2012). Therefore, career interventions that incorporate emotional intelligence skill building may be effective in helping clients resolve career indecision difficulties not rooted in indecisiveness or lack of motivation. One of the first steps in strengthening emotional intelligence is to become aware of one’s own feelings. The use of self-awareness homework exercises, in which clients keep a log of their feelings, may be a first step in helping clients recognize and regulate their emotions. Once clients are better able to manage their anxiety related to making a major life decision, they will be more effective in focusing on the cognitive skills required in career decision-making. Furthermore, interventions such as biofeedback and deep breathing exercises may increase emotional regulation and decrease stress and anxiety related to career indecision. In sum, a better understanding of the direct and indirect relation of attachment security and emotional intelligence to career indecision difficulties may help counselors identify the most appropriate interventions in working with undecided college students (Diener, Oishi, & Lucas, 2002; Erdogan, Bauer, Truxillo, & Mansfield, 2012).

Conclusions

Results of the present study help to elucidate the unique association of attachment and emotional intelligence to dimensions of career decision-making difficulties. Attachment was directly linked to career indecisiveness (lack of readiness), while emotional intelligence was not. In contrast, attachment was indirectly linked to both career indecision dimensions (lack of information and inconsistent information) through emotional intelligence. The identification of antecedents of career difficulty dimensions is of interest because career indecisiveness is typically more resistant to interventions than career indecision (Amir & Gati, 2006), which has implications for the career counseling process. The development of emotional intelligence, which seems to include malleable skills, may be useful in interventions designed to either prevent or address career decision-making difficulties among college women (Di Fabio & Kenny, 2015).