Social cognitive and self-construal predictors of well-being among African college students in the US

Journal of Vocational Behavior 85 (2014) 413–421

Contents lists available at ScienceDirect

Journal of Vocational Behavior journal homepage: www.elsevier.com/locate/jvb

Social cognitive and self-construal predictors of well-being among African college students in the US☆,☆☆ Ijeoma Ezeofor ⁎, Robert W. Lent University of Maryland, College Park, USA

a r t i c l e

i n f o

Article history: Received 12 July 2014 Available online 16 September 2014 Keywords: Social cognitive career theory African students Well-being Academic satisfaction Self-construal

a b s t r a c t We examined social cognitive and cultural predictors of academic satisfaction in African students studying at American universities. In addition to predictors drawn directly from the social cognitive model of work and educational well-being (Lent, 2004; Lent & Brown, 2006, 2008), self-construal was included in the predictive model as a culture-specific variable with potential relevance to academic satisfaction. Self-construal refers to the way in which one's thoughts, behaviors, and feelings are guided by one's relationship to self and others (Markus & Kitayama, 1991). The findings indicated that the model, with some modification, fit the data well and accounted for 59% of the variance in academic satisfaction. The findings also suggested that the three indicators of self-construal (personal, relational, and collective) do not relate directly to academic satisfaction but rather operate through mediated pathways. Implications of the findings for research and practice are discussed. © 2014 Elsevier Inc. All rights reserved.

The recently developed social cognitive model of domain-specific well-being highlights the joint roles of social, cognitive, behavioral, and personality factors in promoting people's adjustment to different life roles and contexts (Lent, 2004), particularly in occupational and educational settings (Lent & Brown, 2006, 2008). Subsequent research has thus far examined the model's utility in predicting teachers' job satisfaction in three different countries, the US (Duffy & Lent, 2009), Italy (Lent et al., 2011), and the United Arab Emirates (Badri, Mohaidat, Ferrandino, & El Mourad, 2013). Studies have also used the model to predict domain satisfaction and other aspects of well-being among college students in the US (Lent, Singley, Sheu, Schmidt, & Schmidt, 2007; Lent et al., 2005), Portugal (Lent, Taveira, Sheu, & Singley, 2009), Taiwan and Singapore (Sheu, Chong, Chen, & Lin, 2014), and Angola and Mozambique (Lent et al., 2014). Studies of the social cognitive model within different countries represent a useful effort to examine the range of the model's applicability across different cultures (Sheu & Lent, 2008). However, it is also valuable to assess how well the model may account for cultural variations in well-being within countries. Several studies have applied the model to racial/ethnic minority college samples in the US, including Latino/a (Flores et al., 2014; Ojeda, Flores, & Navarro, 2011) and Asian American students (Hui, Lent, & Miller, 2013). The current study was designed to test the model's relevance to the academic experiences of African students studying at American colleges, a group that has received relatively little attention in the academic and career development literatures (Constantine, Anderson, Berkel, Caldwell, & Utsey, 2005; Kamya, 1997). We use the term “African students” to include those who were either born in an African country and migrated to the US after high school (1st generation), migrated before high school (1.5 ☆ This research was partially funded by a research grant to the first author from the Division 17 Student Affiliates of Seventeen. ☆☆ This manuscript is based on the master's thesis of the first author performed under the second author's supervision. We are grateful to Matthew J. Miller and Jonathan Mohr for their valuable input as members of the thesis committee. ⁎ Corresponding author at: Department of Counseling, Higher Education, and Special Education, University of Maryland, College Park, MD 20742, USA. E-mail address: [email protected] (I. Ezeofor).

http://dx.doi.org/10.1016/j.jvb.2014.09.003 0001-8791/© 2014 Elsevier Inc. All rights reserved.

414

I. Ezeofor, R.W. Lent / Journal of Vocational Behavior 85 (2014) 413–421

generation), or were born in the US but have at least one parent who was born in Africa (2nd generation) (Awokoya & Harushimana, 2011). African students may differ from their Black/African-American counterparts in the closeness of their ties to their African heritage and African cultural values (e.g., emphasis on respect for authority and family solidarity, education as an obligation rather than a choice) (Massey, Mooney, Torres, & Charles, 2007). The social experiences of African students in U.S. schools may also differ in several ways from their African-American peers, with implications for their sense of well-being in academic settings (Awokoya & Harushimana, 2011). Academic satisfaction, or enjoyment of the academic part of college life, plays a key role in the overall college experience for most students (Gerdes & Mallinckrodt, 1994; Pascarella & Terenzini, 1991). Though academic satisfaction is important in its own right (i.e., as a reflection of how satisfied students are with a key domain of their lives), it is also a valuable predictor of other outcomes, such as intended persistence with particular majors (Lent, Singley, Sheu, Schmidt and Schmidt, 2007; Lent et al., 2013). Therefore, it is valuable to examine the possible antecedents of academic satisfaction. Such research could inform the design of interventions to aid the college adjustment of particular groups of students. Many writers have noted the relevance of culture to subjective well-being (e.g., Cantor & Sanderson, 1999; Christopher, 1999). For example, culture can be reflected in the personal goals people set for themselves, in the values they seek to fulfill, and in the activities they voluntarily pursue. In proposing an integrative social cognitive approach to well-being, Lent (2004) noted the importance of studying both pancultural and culture-specific aspects of domain and life satisfaction. Most of the subsequent research on Lent's framework has involved participants from Western cultures. There is a need, therefore, to examine the extent to which the general model may fit the experiences of persons from non-Western cultures as well as to identify culture-specific variables that may complement the model's more pancultural elements, both in testing the model abroad and within US sub-cultures. Although Africa consists of many distinct countries and cultural groups, for many African students, particularly those who migrate to the US or who are children of African immigrants, a great emphasis is placed on the value of higher education (Amayo, 2007). African students typically learn about their roles and expectations from their family and community members, with a common emphasis on the individual assuming responsibility for the well-being, status, and security of their family (Francis, 2000). Further, as the African familial structure tends to emphasize the extended family over the nuclear family, many African students are encouraged to view their behavior as having consequences for their extended community of kinsmen, and not only for their immediate family or for themselves as individuals. Therefore, in studying the experiences of African students at US colleges, it may be valuable to assess how this interdependent orientation may be linked to students' academic satisfaction. Self-construal, the cultural variable we selected for inclusion in this study, provides a way to index students' affinity for individually-oriented as well as for more collective (e.g., family, community) values.

1. Social cognitive model of domain satisfaction The social cognitive model posits that domain and life satisfaction can be predicted by environmental supports and resources, selfefficacy expectations, outcome expectations, goal progress, as well as by personality traits and affective dispositions (Lent, 2004). Based partly on Bandura's (1986, 1997) general social cognitive theory, the satisfaction model incorporates the assumption that “… by engaging in self-efficacy-building and valued life activities, setting and making progress at meaningful personal goals, or seeking out needed social resources, individuals have the opportunity to partly steer themselves toward happiness” (Sheu & Lent, 2008, p. 49). Self-efficacy refers to confidence in one's ability to successfully perform particular behaviors or courses of action. Outcome expectations involve individuals' beliefs about the consequences of pursuing a specific course of action. Goal pursuit refers to people's

Fig. 1. Social cognitive model of normative well-being. From Lent, R. W. (2004). Toward a unifying theoretical and practical perspective on well-being and psychosocial adjustment. Journal of Counseling Psychology, 51 (4), 482–509.

I. Ezeofor, R.W. Lent / Journal of Vocational Behavior 85 (2014) 413–421

415

intentions to perform particular activities or to achieve performances at a particular level of proficiency. Environmental supports and resources are contextual factors that enable people to pursue their goals or to build their self-efficacy. Personality and affective traits are person variables that predispose one toward the general experience of pleasant or unpleasant emotions. Each of these variables is seen as contributing to satisfaction within particular life domains. Life domain satisfaction can, in turn, promote overall satisfaction with life, or general happiness. The current study is concerned with academic domain satisfaction, which Lent and Brown (2008) described as the positive emotion experienced when appraising one's academic conditions. Fig. 1 displays how the social cognitive variables are hypothesized to relate to one another and to domain and life satisfaction. Since its introduction, a number of studies have applied the model to the prediction of academic satisfaction. For example, the basic model has been found to fit the data well overall in predicting academic satisfaction or other aspects of academic functioning in US (Lent et al., 2005), European (Lent, Taveira, & Lobo, 2012), and African countries (Lent et al., 2014). A few studies have also begun to examine how culture-specific variables may interrelate with the model's basic predictors. For instance, enculturation and acculturation were found to relate to academic satisfaction indirectly, rather than directly, via particular social cognitive variables in Mexican American (Ojeda et al., 2011) and Asian American students (Hui et al., 2013).

2. Self-construal Self-construal is a multi-dimensional construct reflecting the “constellation of thoughts, feelings, and actions concerning one's relationship to others, and the self as distinct from others” (Singelis, 1994, p. 581). It has been conceptualized as a personality trait that shapes the cognitions, emotions, motivations, and behaviors of individuals. Markus and Kitayama (1991) identified independent and interdependent dimensions of self-construal. Individuals with strong independent self-construals hold a view of the self as an autonomous entity and tend to value individual uniqueness and self-expression (Christopher, D'Souza, Peraza, & Dhaliwal, 2010). By contrast, individuals with strong interdependent self-construals hold a view of the self that is defined primarily by relationships with others (Christopher et al., 2010). It is possible that African students in the US often seek to balance the collectivistic values of their upbringing and the individualistic norms of the society in which they reside. Consequently, they may hold both interdependent and independent self-construals to varying degrees. Extant research has found that independent and interdependent self-construals predict various indicators of psychological and emotional functioning (e.g., coping, stress, life satisfaction) (Cross, 1995; Suh, Diener, & Updegraff, 2008). Sheu et al. (2014) examined self-construal relative to the social cognitive predictors of academic satisfaction in Taiwanese and Singaporean college students. They found that, in both samples, ratings of interdependence were associated with greater social support, which in turn predicted academic satisfaction. However, independent self-construal was not a significant direct or indirect predictor of academic satisfaction in either sample. Despite its promise as a predictor of well-being-related outcomes, there has been debate about the measurement of self-construal, with some research questioning the psychometric adequacy (cf., Levine et al., 2003) and dimensionality (Brewer & Gardner, 1996; Hardin, Leong, & Bhagwat, 2004) of common measures of the construct. For example, in a test of Singelis' (1994) self-construal scale, Hardin et al. found that a six-factor model provided better fit to the data than did the original two-factor model. To assess self-construal in the present study, we used Harb and Smith's (2008) Sixfold Self-Construal Scale (SSCS), which measures self-construal across six dimensions and has yielded adequate reliability and validity estimates in initial research (Harb & Smith, 2008). The personal self-construal dimension of the SSCS is analogous to Markus and Kitayama's (1991) independent self-construal. Relational and collective self-construals are similar to Markus and Kitayama's interdependent self-construal. However, relational self-construal refers to a view of the self that is linked to one's dyadic or small group relationships (e.g., friends, family) and collective self-construal refers to a view of the self as defined by one's membership in a larger, more distal group unified by shared norms and values (e.g., cultural group). Harb and Smith (2008) extended the conceptualization of relational and collective self-construal by differentiating between vertical (i.e., hierarchical) relationships and horizontal (i.e., egalitarian) relationships. The sixth dimension of self-construal is humanity-bound, which reflects the extent to which individuals' thoughts, feelings, and behaviors are shaped by their views of themselves as part of a universal human culture (Harb & Smith, 2008).

3. Purpose of the study The current study employed the social cognitive model of domain satisfaction to predict the academic satisfaction of African students attending US universities and colleges. Results from several studies have supported the model's utility in explaining domain and life satisfaction cross-culturally. To examine possible cultural variations in academic satisfaction owing to independent and interdependent views of self, three dimensions of self-construal—personal self-construal, relational self-construal, and collective self-construal—were added to the basic social cognitive model. Similar to Hui et al.'s (2013) addition of acculturation and enculturation to the social cognitive model with an Asian American student sample, we examined the possibility that different types of self-construal would produce direct paths to academic satisfaction as well as indirect paths through perceived environmental support for one's academic behavior. Life satisfaction was not included in the study since the primary goal was to explore the predictors of academic domain satisfaction.

416

I. Ezeofor, R.W. Lent / Journal of Vocational Behavior 85 (2014) 413–421

4. Method 4.1. Participants and procedures The population of interest for this study included first, 1.5, and second generation African undergraduate students attending 4-year colleges and universities in the US. A total of 174 African undergraduate students participated in the study. Two-hundred-two persons responded to the survey, but 28 of them were deemed ineligible and removed from the data set, yielding a sample size of 174 (18 students were removed because they had already graduated or were in graduate school; 8 students were not in a 4-year university program; 1 student was attending school in Europe; and 1 student did not complete the demographic questions). A majority of the participants were female (81%), of West African descent (73%), and in their junior or senior years of college (61%). Fifty-five percent of the respondents were born outside of the US and 26% were the first member of their family to attend college in the US. The mean age was 21.1 years, SD = 1.9 years. Data were collected via an online survey link. After completing an informed consent form, participants were presented with the measures of academic self-efficacy, support, goal progress, outcome expectations, and satisfaction; the three subscales of the SSCS; and demographic questions. Students were recruited via email, social media, flyers, and informal contacts with a variety of African student associations, both at the authors' university and nationally. Students were also recruited at a national gathering of Nigerian students living in the US, via African-serving churches in the local campus area, and through an email list of African international students at the authors' university. Those receiving information about the study were asked to forward the study link on to other African students at US universities. Students who completed the survey were invited to participate in a raffle to win one of twenty gift cards from an online retailer. Those who completed the survey via psychology courses at the authors' university were offered the option of receiving extra course credit. Given the variety and nature of recruitment methods (e.g., snowball technique), it was not possible to calculate a response rate. 4.2. Instruments 4.2.1. Academic self-efficacy Academic self-efficacy was assessed with a 12-item scale measuring self-efficacy for general academic achievements (5 items; e.g., “How confident are you in your ability to excel in your intended major over the next two semesters”) and self-efficacy for coping with academic challenges (7 items; e.g., “How confident are you in your ability to complete a degree despite financial pressures”; Lent et al., 2005). Participants indicated their confidence in completing all items on a 10-point scale ranging from 0 (no confidence) to 9 (complete confidence). Lent et al. (2005) found evidence of discriminant validity for the academic self-efficacy scale, observing a small correlation between this measure and a measure of social domain self-efficacy (r = .24). Criterion-related validity was evidenced by correlations between academic self-efficacy and measures of academic adjustment and intended persistence (Lent et al., 2005). Internal consistency reliability estimates were above .85 (Lent et al., 2005). The Cronbach α for the full academic self-efficacy measure in the present study was .89. 4.2.2. Academic support The 9-item academic support scale assessed participants' perceptions of their academic domain-specific support (e.g., “At the present time, I feel support from important people in my life (e.g., teachers) for pursuing my intended major”; Lent et al., 2005). All nine items were measured on a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree). Criterion-related validity was estimated through significant correlations between this measure and measures of academic adjustment and intended persistence (Lent et al., 2005). Lent et al. (2005) found a moderate correlation between this measure and a measure of social domain support (r = .41), suggesting that this measure reflects perceptions that are partly domain-specific in nature. Lent et al. (2005) calculated an internal consistency reliability estimate of α = .81. In this study, the reliability estimate for the scale was .83. 4.2.3. Academic outcome expectations The 10 items of the academic outcome expectations scale capture individuals' beliefs about potential outcomes that could result from their academic pursuits (e.g., “A college education will allow me to obtain a well-paying job”; Lent et al., 2005). The items were measured on a 10-point scale from 1 (strongly disagree) to 10 (strongly agree). Lent et al. (2005) found a moderate correlation between this measure and a measure of social outcome expectations (.34) as evidence of discriminant validity. Correlations of this measure with measures of academic interest and choice provided evidence of criterion-related validity (Lent et al., 2007). Lent et al. (2005) reported a reliability estimate of .91. The internal consistency reliability estimate was .94 in the present study. 4.2.4. Academic goal progress The academic goal progress scale (7 items) assessed the extent of perceived progress towards academic goals (e.g., “completing academic requirements of your major satisfactorily”) (Lent et al., 2005). Participants responded on a 5-point scale ranging from 1 (no progress at all) to 5 (excellent progress). This scale correlated modestly with a measure of social goal progress (r = .16) but more substantially with measures of academic adjustment and intended persistence, offering evidence related to discriminant and criterion-related validity, respectively (Lent et al., 2005). Lent et al. (2005) reported an internal consistency reliability estimate of .86. The reliability estimate in the present study was .93.

I. Ezeofor, R.W. Lent / Journal of Vocational Behavior 85 (2014) 413–421

417

4.2.5. Academic satisfaction scale Academic satisfaction was measured with a 7-item scale that asked participants to rate their satisfaction with various aspects of their academic experience (e.g., “I enjoy the level of intellectual stimulation in my courses”). Items were measured on a 5-point scale, from 1 (strongly disagree) to 5 (strongly agree). The measure correlated only modestly (r = .25) with a measure of social domain satisfaction in prior research, offering evidence of discriminant validity (Lent et al., 2005). In terms of criterion-related validity, this scale correlated significantly with measures of academic adjustment and intended persistence (Lent et al., 2005). The scale yielded internal consistency reliability estimates of .87 and .90 in the Lent et al. (2005) and current samples, respectively. 4.2.6. Sixfold Self-Construal Scale (SSCS) The personal, relational, and collective self-construal subscales of the SSCS (Harb & Smith, 2008) were used in the present study. Each subscale contained five items. Participants rated the extent to which they agreed with each item on a 7-point scale, from 1 (to a very small extent) to 7 (to a very large extent). Sample items were as follows: “I think of myself as connected (linked) to myself (I am a unique person separate from others)” (personal self-construal); “I control my behavior to accommodate the wishes (interests) of my family” (relational self-construal); and “I am affected by events that concern (relate to) the ethnic community” (collective self-construal). Note that the term, family, was used to represent relational self-construal and participants' specific African ethnic community (e.g., Nigerian, Ethiopian) was used as the referent for collective self-construal. Participants indicated their African ethnicity at the start of the survey and their selection was then automatically placed within the collective self-construal item stems. Harb and Smith (2008) factor analyzed the SSCS in samples of British, Lebanese, Syrian, and Jordanian students. In each sample, a six-factor model provided good fit to the data. The SSCS scales were also found to correlate appropriately with measures of conceptually related constructs (Harb & Smith, 2008). For example, among the self-construal scales, personal self-construal correlated most consistently with values such as self-direction, achievement, and power, and relational self-construal correlated most strongly with conformity values. The scores from each of the subscales yielded adequate Cronbach's α reliabilities (α = .68–.92) in Harb and Smith's four samples. The internal consistency reliability estimates for the personal, relational, and collective-level self-construal subscales in this study were .63, .73, and .86, respectively. (For the purposes of this study, we did not include the humanity-bound self-construal dimension and did not deem it necessary to differentiate between horizontal and vertical relationship aspects of relational and collective self-construal.) 5. Results We ran a series of t-tests to determine whether foreign-born and U.S.-born African students differed significantly on the eight variables under study (see Table 1). The only statistically significant difference between the two groups was on personal self-construal, with the U.S.-born students reporting higher scores. Because the effect size of this difference was relatively small (d = .31) and the groups did not differ on the dependent variable of interest (academic satisfaction), the data were combined over the groups in the subsequent analyses. Examination of the score distributions revealed no substantial deviations from univariate normality. However, because the value of Mardia's coefficient was large, reflecting multivariate non-normality, robust maximum likelihood estimation was used in the structural equation analyses. Table 2 presents descriptive statistics and correlations among the variables. There were no missing data as the online survey was set up to require a response for each item; participants could close their browsers if they did not wish to complete the full survey. The covariance matrices of the observed variables were subjected to path analysis, using LISREL 9.10 (Jöreskog & Sörbom, 1996). The model included the four social cognitive predictors (support, self-efficacy, outcome expectations, and goal progress), the three types of self-construal (personal, relational, and collective), and the dependent variable, academic satisfaction. The three self-construal variables were allowed to covary, and each was modeled as being linked to academic satisfaction indirectly, through academic support. That is, it was assumed that those with a tendency toward relational and collective self-construal, reflecting views of the self as interdependent with others, might be more likely to rely on others for support in negotiating the academic environment. It was also assumed that relatively greater levels of personal self-construal, reflecting beliefs that one's needs and Table 1 Comparison of U.S.-born and foreign-born Africans on the study variables. M

Personal self-construal Relational self-construal Collective self-construal Academic self-efficacy Academic social support Academic outcome expectations Academic goal progress Academic satisfaction Note. NUS = 78; NNon-U.S. = 96. † p = .050.

SD

t

US

Non-U.S.

US

Non-U.S.

6.11 6.13 4.35 7.49 3.92 7.18 3.75 3.89

5.89 6.14 4.56 7.54 3.81 7.25 3.81 4.01

.71 .75 1.44 1.22 .77 1.56 .91 .80

.72 .79 1.33 1.07 .71 1.35 .78 .74

1.974† −0.110 −1.006 −0.308 0.958 −0.430 −0.305 −1.015

418

I. Ezeofor, R.W. Lent / Journal of Vocational Behavior 85 (2014) 413–421

Table 2 Descriptive statistics and correlations on the study variables.

1. Personal self-construal 2. Relational self-construal 3. Collective self-construal 4. Academic self-efficacy 5. Academic social support 6. Academic outcome expectations 7. Academic goal progress 8. Academic satisfaction

M

SD

1

6.00 6.14 4.47 7.52 3.86 7.22

0.72 0.77 1.38 1.13 0.74 1.45

-

3.78 3.96

0.84 0.77

2

.184⁎⁎ .127⁎ .333⁎⁎ .201⁎⁎ .464⁎⁎

-

.256⁎⁎ .298⁎⁎

3

.465⁎⁎ .171⁎ .233⁎⁎ .132⁎ .166⁎ .174⁎

4

5

6

7

8

.053 .143⁎ .164⁎

-

.377⁎⁎ .361⁎⁎ .705⁎⁎ .628⁎⁎

−.015 .063

-

.394⁎⁎ .341⁎⁎ .388⁎⁎

.374⁎⁎ .697⁎⁎

-

.514⁎⁎

-

α

Range

Skewness

Kurtosis

.63 .73 .86 .89 .83 .94

3.60–7.00 3.40–7.00 1.20–7.00 1.58–9.00 1.89–5.00 1.40–9.00

−0.779 −1.199 −0.259 −1.395 −0.452 −1.187

0.368 1.070 −0.672 3.932 −0.404 2.130

.93 .90

1.00–5.00 1.00–5.00

−0.783 −0.786

0.695 0.988

⁎⁎ p b .01. ⁎ p b .05.

desires should be prioritized, might enable students to assert themselves in gaining access to needed academic supports. In each case, the linkage to social support was conceived as parallel to the enculturation and acculturation orientation variables studied by Hui et al. (2013) in their test of the social cognitive model with Asian American students. We employed three primary indicators of the adequacy of model-data fit: the comparative fit index (CFI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMR). Hu and Bentler (1999) have suggested that good model fit is indicated by CFI values of .95 and above, SRMR values of .08 and below, and RMSEA values of .06 and below. The target model did not produce adequate fit to the data, S-B χ2(12) = 74.82, p b .01; CFI = .89; SRMR = .10; RMSEA = .14. Modification indices suggested that fit could be improved by adding a path from personal self-construal to both self-efficacy and outcome expectations. These additions seemed reasonable because, conceptually, the personal self-construal items may reflect a trait sense of personal or independent agency, and persons with a greater sense of agency may be likely to perceive themselves as more efficacious (Lent, 2004) and more likely to receive positive outcomes from their academic efforts (Sheu et al., 2014). Adding these paths (shown with dashed lines in Fig. 2), resulted in a much-improved model fit, S-B χ2(10) = 15.64, p = 0.11; CFI = .99; RMSEA = .00; and SRMR = .03. The paths from personal self-construal to academic self-efficacy and outcome expectations each produced significant parameter estimates and statistical comparison between the original and modified models, using the S-B χ2 difference test (Bryant & Satorra, 2012; Satorra & Bentler, 2001), indicated that the modified model fit the data significantly better than did the original model, ΔS-B χ2(2) = 34.11 p b .001. While the modified model produced good fit to the data, we also tested an alternative model exploring the possibility that the self-construal variables might relate to academic satisfaction directly as well as indirectly through the social cognitive variables. This alternative model included all of the paths in the modified model in addition to a direct path from each of the self-construal variables to academic satisfaction. Results indicated that this model variant fit the data well, S-B χ2(7) = 13.77, p = .06; RMSEA = .03; CFI = .99; and SRMR = .02. However, none of the self-construal to satisfaction path coefficients was significant, and

Collective SelfConstrual .04

.47**

Relational SelfConstrual .19*

.16

.13 .18*

Personal SelfConstrual

.27** .36**

Academic Self-Efficacy

.32**

.21** .64**

.14

Academic Goal Progress

.05

Academic Support

.27**

Academic Outcome Expectations

.12* .25**

.06 Fig. 2. Parameter estimates of the alternative model. *p b .05, **p b .01.

.44**

Academic Satisfaction

I. Ezeofor, R.W. Lent / Journal of Vocational Behavior 85 (2014) 413–421

419

the direct effects model did not improve on the fit of the target model, ΔS-B χ2(3) = .68, p N .10. Thus, it may be inferred that the more parsimonious (indirect effects-only) model is preferable. Examining the path coefficients in Fig. 2, it may be seen that most of the hypothesized paths among the social cognitive variables, and from these variables to academic satisfaction, yielded significant parameter estimates. Academic satisfaction was well-predicted by the combination of academic self-efficacy, academic goal progress, and academic outcome expectations; collectively, the model explained 59% of the variation in academic satisfaction. However, the path from academic self-efficacy to academic outcome expectations was non-significant, as were the paths from academic support to academic goal progress and academic satisfaction. Thus, academic support appeared to be linked to academic satisfaction only indirectly, via its relations to academic self-efficacy and academic outcome expectations, rather than directly or through academic goal progress. Most of the paths involving the self-construal variables did not conform to expectations. Only relational self-construal yielded a significant direct path to academic support, suggesting that those who are closely connected to others in their social circle are likely to receive more support for their academic efforts. Collective self-construal, while covarying significantly with relational self-construal, did not relate uniquely to the other predictors. Personal self-construal, likewise, did not produce a significant path, as expected, to support. It did, however, predict satisfaction indirectly through its relations both to academic self-efficacy and academic outcome expectations. None of the self-construal variables produced significant, direct paths to academic satisfaction. Using the joint test of significance (Mallinckrodt, Abraham, Wei, & Russell, 2006), it may be inferred that the relations of the self-construal variables to satisfaction were mediated by other variables in the model. 6. Discussion The present findings indicated that the basic social cognitive variables explained nearly 60% of the variance in academic satisfaction. These findings replicate, with a sample of African college students living in the US, the general pattern of earlier findings with samples representing other cultural groups in the US (e.g., Mexican American, Asian Americans) and abroad (e.g., Italians, Taiwanese). Thus, the findings add useful information about the potential cross-cultural applicability of the model's core variables. In particular, academic self-efficacy, outcome expectations, and goal progress each yielded significant direct paths to academic satisfaction. While prior research has yielded mixed findings regarding outcome expectations as a unique predictor of academic satisfaction (e.g., Lent et al., 2005), similar to Ojeda et al. (2011), we found that holding favorable beliefs about the outcomes of one's academic efforts complemented the explanatory utility of self-efficacy and goal progress in our sample. Despite the predictive utility of the social cognitive variables, a few hypothesized paths did not receive support. Contrary to expectations, academic support did not produce significant direct paths either to academic goal progress or satisfaction. However, academic support was linked to academic satisfaction indirectly via its relations to self-efficacy and outcome expectations. In other words, support for students' academic efforts may help to bolster their self-efficacy and outcome expectations which, in turn, may promote goal progress and academic domain satisfaction. The pathway from support to academic satisfaction has been found to be at least partly indirect in studies involving Asian American (Hui et al., 2013), Taiwanese (Sheu et al., 2014), and African students (Lent et al., 2014). A unique feature of our study was the addition of three self-construal variables to capture aspects of African students' cultural experiences in negotiating the academic environment in the US educational context. We had anticipated that the three forms of self-construal – two reflecting a group-focused cultural identification and the third representing a more independent cultural stance – would shed light on the interplay between pancultural social cognitive and culture-specific variables relevant to African students' academic adjustment. Our findings indicated that the self-construal variables did not produce significant direct paths to academic satisfaction, over and above the social cognitive predictors. Rather, in keeping with other studies that have added culture-specific variables to the SCCT satisfaction model (Hui et al., 2013; Ojeda et al., 2011; Sheu et al., 2014), our findings suggested that certain aspects of self-construal may relate to academic satisfaction indirectly through the social cognitive elements. In particular, consistent with expectations, relational self-construal was linked to academic satisfaction through academic support, suggesting that students adhering to relational norms were more likely to seek social support as part of their academic adjustment. Sheu et al. (2014) also found that interdependent self-construal, which was conceptually similar to our relational self-construal measure, was linked to academic satisfaction indirectly via academic support. Collective self-construal, by contrast, did not produce a unique link to academic support in our study, perhaps because its predictive value was subsumed by relational self-construal. That is, these two aspects of self-construal both reflected adherence to collective cultural values and were highly interrelated. While identification with one's larger cultural community held an important place in many students' lives, it is possible that relational selfconstrual, reflecting strong ties to one's family, provided a more tangible source of resources (e.g., emotional assistance, cognitive guidance) and motivations (e.g., to contribute to the welfare of one's family) for managing the academic demands of college life. Personal self-construal functioned somewhat differently than relational self-construal in the model test. Unlike relational selfconstrual, it did not yield a significant path to academic support but did produce significant paths both to self-efficacy and outcome expectations. The item content of the personal self-construal scale suggested that it may reflect a global sense of agency or personal volition. For instance, the scale's items emphasize one's ability to “act as an independent person” or to be aware of one's own values and goals. This global or trait notion of independent agency may have helped to bolster academic domain-specific self-efficacy and outcome expectations. That is, at a trait level, those who generally see themselves as capable of exercising self-direction may be more likely to hold favorable beliefs about their efficacy and the results of their actions within particular life domains. Interpretation and generalizability of our findings are limited by several considerations. First, our results may have been affected by participant self-selection. It was not possible to calculate an accurate response rate given our use of snowball sampling and other

420

I. Ezeofor, R.W. Lent / Journal of Vocational Behavior 85 (2014) 413–421

methods that relied on third parties to distribute the study invitation. However, it should be noted that many of the students in this study were members of African affinity clubs. Thus, the students who responded may differ from those who did not respond in their level of social connectedness and identification with their African culture. A second limitation was the cross-sectional nature of this study, which precludes drawing causal inferences from the findings. Third, the SSCS is a relatively new measure that produced marginal reliability estimates for two of the three self-construal variables in our study; this may have attenuated their relations to the other variables in the model test. Fourth, we relied on modification indices to improve the fit of the model by adding links from the personal self-construal variable to self-efficacy and outcome expectations. Although these two additional paths could be justified on conceptual grounds, it would be useful to examine the modified model in a new sample. Several additional directions for future research might be noted. For example, we found that the self-construal variables did not explain unique variance in academic satisfaction, beyond the social cognitive variables. One reason may be that students rated their self-construal in a nonspecific context whereas the social cognitive variables, as well as the satisfaction ratings, were all linked to the academic life domain. Future research might examine whether domain-specific ratings of self-construal are capable of explaining unique variance in domain-specific outcomes. A second direction for research would be to examine the validity of the social cognitive model of well-being in African students' other life domains, such as work or social relationships. Third, to address some of the limitations of the current study, it may be valuable to test the hypothesized directional relations among the predictors using longitudinal designs and to further assess the psychometric adequacy of the self-construal scales with African student samples. Fourth, because Africa includes many distinct ethnic affiliations, it would be useful to study social cognitive, self-construal, and other predictors of well-being within students who identify with particular African cultural groups. The results of this study offer tentative practical implications for efforts to promote the academic experiences of African college students in the US. In particular, the findings involving the linkage of relational self-construal to academic support in our sample suggest the potential value of social support-based interventions, for example, involving mentoring to facilitate academic adjustment and success. In addition, methods might be designed to enable students to access needed supports from family members and others within their larger social circles. The four theoretical sources of efficacy information might also be harnessed in culturally relevant ways. For example, students might be exposed to coping role models who share their cultural backgrounds and values and who can demonstrate methods for negotiating the college environment while maintaining connections to one's home culture and family. In sum, this study provides the first empirical test of the SCCT satisfaction model (Lent & Brown, 2008) with African students attending US universities. The results contribute to an understanding of factors that may facilitate African college students' adjustment to the college environment in the US. Specifically, academic self-efficacy, outcome expectations, and goal progress were found to be directly associated with academic satisfaction in our sample. We also found that both personal and relational self-construal related to academic satisfaction, albeit through different indirect pathways. These findings may inform counselors about targets for preventive and developmental academic interventions with African college students (e.g., promoting academic self-efficacy and goal progress) and, hopefully, stimulate further research on this understudied population. References Amayo, O. F. (2007). Evaluation of Nigerian immigrant parent practices in preparing their children for college. (Doctoral dissertation). (Retrieved from ProQuest Dissertations and Theses database. (UMI No. 3314129)). Awokoya, J. T., & Harushimana, I. (2011). Family-school mediation: Righting the mental self-perceptions of 1.5 and 2.0 African youth in US schools. In T. L. Mann, N. E. Hill, & H. E. Fitzgerald (Eds.), African American children and mental health. Prevention and social policy, Vol. 2. (pp. 211–238). Santa Barbara, CA: ABC-CLIO, LLC. Badri, M.A., Mohaidat, J., Ferrandino, V., & El Mourad, T. (2013). The social cognitive model of job satisfaction among teachers: Testing and validation. International Journal of Educational Research, 57, 12–24. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman. Brewer, M. B., & Gardner, W. (1996). Who is this “we”? Levels of collective identity and self representations. Journal of Personality and Social Psychology, 71(1), 83–93. Bryant, F. B., & Satorra, A. (2012). Principles and practice of scaled difference chi-square testing. Structural Equation Modeling: A Multidisciplinary Journal, 19(3), 372–398. Cantor, N., & Sanderson, C. A. (1999). Life task participation and wellbeing: The importance of taking part in daily life. In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: The foundations of hedonic psychology (pp. 230–243). New York: Russell Sage Foundation. Christopher, J. C. (1999). Situating psychological well‐being: Exploring the cultural roots of its theory and research. Journal of Counseling and Development, 77(2), 141–152. Christopher, M. S., D'Souza, J. B., Peraza, J., & Dhaliwal, S. (2010). A test of the personality–culture clash hypothesis among college students in an individualistic and collectivistic culture. International Journal of Culture and Mental Health, 3(2), 107–116. Constantine, M. G., Anderson, G. M., Berkel, L. A., Caldwell, L. D., & Utsey, S. O. (2005). Examining the cultural adjustment experiences of African international college students: A qualitative analysis. Journal of Counseling Psychology, 52(1), 57–66. Cross, S. E. (1995). Self-construals, coping, and stress in cross-cultural adaptations. Journal of Cross-Cultural Psychology, 26(6), 673–697. Duffy, R. D., & Lent, R. W. (2009). Test of a social cognitive model of work satisfaction in teachers. Journal of Vocational Behavior, 75(2), 212–223. Flores, L. Y., Navarro, R. L., Lee, H. S., Addae, D. A., Gonzalez, R., Luna, L. L., et al. (2014). Academic satisfaction among Latino/a and White men and women engineering students. Journal of Counseling Psychology, 61(1), 81–92. Francis, E. A. (2000). Social work practice with African descent immigrants. In P. R. Balgopal (Ed.), Social work practice with immigrants and refugees (pp. 127–166). New York, NY: Columbia University Press. Gerdes, H., & Mallinckrodt, B. (1994). Emotional, social, and academic adjustment of college students: A longitudinal study of retention. Journal of Counseling and Development, 72, 281–288. Harb, C., & Smith, P. B. (2008). Self-construals across cultures: Beyond independence-interdependence. Journal of Cross-Cultural Psychology, 39(2), 178–198. Hardin, E. E., Leong, F. T. L., & Bhagwat, A. A. (2004). Factor structure of the self-construal scale revisited: Implications for the multidimensionality of self-construal. Journal of Cross-Cultural Psychology, 35(3), 327–345. Hu, L. -T., & Bentler, P.M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. Hui, K., Lent, R. W., & Miller, M. J. (2013). Social cognitive and cultural orientation predictors of well-being in Asian American college students. Journal of Career Assessment, 21(4), 587–598.

I. Ezeofor, R.W. Lent / Journal of Vocational Behavior 85 (2014) 413–421

421

Jöreskog, K. G., & Sörbom, D. (1996). LISREL 8: User's reference guide. Scientific Software International. Kamya, H. A. (1997). African immigrants in the United States: The challenge for research and practice. Social Work, 42, 154–166. Lent, R. W. (2004). Toward a unifying theoretical and practical perspective on well-being and psychosocial adjustment. Journal of Counseling Psychology, 51(4), 482–509. Lent, R. W., & Brown, S. D. (2006). Integrating person and situation perspectives on work satisfaction: A social–cognitive view. Journal of Vocational Behavior, 69(2), 236–247 (128). Lent, R. W., & Brown, S. (2008). Social cognitive career theory and subjective well-being in the context of work. Journal of Career Assessment, 16(1), 6–21. Lent, R. W., Miller, M. J., Smith, P. E., Watford, B.A., Lim, R. H., Hui, K., et al. (2013). Social cognitive predictors of adjustment to engineering majors across gender and race/ethnicity. Journal of Vocational Behavior, 83(1), 22–30. Lent, R. W., Nota, L., Soresi, S., Ginevra, M. C., Duffy, R. D., & Brown, S. D. (2011). Predicting the job and life satisfaction of Italian teachers: Test of a social cognitive model. Journal of Vocational Behavior, 79(1), 91–97. Lent, R. W., Singley, D., Sheu, H. -B., Gainor, K., Brenner, B., Treistman, D., et al. (2005). Social cognitive predictors of domain and life satisfaction: Exploring the theoretical precursors of subjective well-being. Journal of Counseling Psychology, 52(3), 429–442. Lent, R. W., Singley, D., Sheu, H. -B., Schmidt, J. A., & Schmidt, L. C. (2007). Relation of social–cognitive factors to academic satisfaction in engineering students. Journal of Career Assessment, 15, 87–97. Lent, R. W., Taveira, M., & Lobo, C. (2012). Two tests of the social cognitive model of well-being in Portuguese college students. Journal of Vocational Behavior, 74(2), 190–198. Lent, R. W., Taveira, M., Pinto, J. C., Silva, A.D., Blanco, Á., Faria, S., et al. (2014). Social cognitive predictors of well-being in African college students. Journal of Vocational Behavior, 84(3), 266–272. Lent, R. W., Taveira, M., Sheu, H. -B., & Singley, D. (2009). Social cognitive predictors of academic adjustment and life satisfaction in Portuguese college students: A longitudinal analysis. Journal of Vocational Behavior, 74(2), 190–198. http://dx.doi.org/10.1016/j.jvb.2008.12.006. Levine, T. R., Bresnahan, M. J., Park, H. S., Lapinski, M. K., Wittenbaum, G. M., Shearman, S. M., et al. (2003). Self-construal scales lack validity. Human Communication Research, 29, 210–252. Mallinckrodt, B., Abraham, W. T., Wei, M., & Russell, D. W. (2006). Advances in testing the statistical significance of mediation effects. Journal of Counseling Psychology, 53(3), 372–378. Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224. Massey, D. S., Mooney, M., Torres, K. C., & Charles, C. Z. (2007). Black immigrants and black natives attending selective colleges and universities in the United States. American Journal of Education, 113(2), 243–271. Ojeda, L., Flores, L. Y., & Navarro, R. L. (2011). Social cognitive predictors of Mexican American college students' academic and life satisfaction. Journal of Counseling Psychology, 58, 61–71. Pascarella, E. T., & Terenzini, P. T. (1991). How college affects students. San Francisco, CA: Jossey-Bass Publishers. Satorra, A., & Bentler, P.M. (2001). A scaled difference chi-square test statistic for moment structure analysis. Psychometrika, 66(4), 507–514. Sheu, H. -B., Chong, S. S., Chen, H. -F., & Lin, W. -C. (2014). Well-being of Taiwanese and Singaporean college students: Cross-cultural validity of a modified social cognitive model. Journal of Counseling Psychology, 61(3), 447–460. Sheu, H. -B., & Lent, R. W. (2008). A social cognitive perspective on well-being in educational and work settings: Cross-cultural considerations. International Journal for Educational and Vocational Guidance, 9(1), 45–60. Singelis, T. (1994). The measurement of independent and interdependent self-construals. Personality and Social Psychology Bulletin, 20(5), 580–591. Suh, E. M., Diener, E., & Updegraff, J. A. (2008). From culture to priming conditions: Self-construal influences on life satisfaction judgments. Journal of Cross-Cultural Psychology, 39(1), 3–15.