SCCT Research in the International Context

Asia

International SCCT research conducted in Asia has covered all four SCCT outcome variables (interest, choice, performance, and satisfaction) and involved 14 independent samples derived from 12 studies. Sheu and his colleagues (Sheu, Chong, Chen, & Lin, 2014; Sheu, Liu, & Li, 2017) gathered data from college students in Taiwan, Singapore, and China using different languages (traditional/simplified Mandarin, English) to test Lent and Brown’s (2006b, 2008) satisfaction model. Independent and interdependent self-construals were added to the model for the purpose of examining the utility of individualistic and collectivistic cultural orientations in predicting satisfaction outcomes. In spite of cross-national differences on a few paths (e.g., academic goal progress → academic satisfaction, independence → academic stress), data gathered from these Asian countries provided evidence for most of Lent and Brown’s (2006b, 2008) hypotheses, and the satisfaction model explained 49–67%, 47–66%, and 46–57% of the variance, respectively, in academic satisfaction, academic stress, and life satisfaction.

Specifically, the aforementioned two studies showed that the relations of personality traits (extraversion and emotional stability) and interdependent self-construal to satisfaction outcomes were partially mediated by academic supports, academic self-efficacy, and academic goal progress. That is, extraverted, emotionally stable, and interdependent individuals reported higher levels of academic satisfaction and lower levels of academic stress due, in part, to their perceptions of supports for, confidence in, and progress made toward academic goals. Additionally, self-construals were found to predict academic satisfaction and stress after controlling for personality traits, indicating the benefit of adding one’s cultural orientations to the satisfaction model. However, the predictive utility of academic outcome expectations was not supported in these three Asian samples.

The SCCT interest and choice models have also been tested in several studies with Asian samples. Jiang and Zhang (2012) found that in a sample of engineering students in China, social supports and barriers predicted major choice goals directly and indirectly via self-efficacy and outcome expectations and that interest fully mediated the relations of self-efficacy and outcome expectations to choice goals. A study conducted by Song and Chon (2012) showed that interest was predictive of choice goals in tourism and hospitality jobs after controlling for general self-efficacy and perceived person-job fit among Chinese college students. However, Sam, Othman, and Nordin (2005) demonstrated that time spent on using the Internet (i.e., choice action) was mildly and positively correlated with computer anxiety but unrelated to computer self-efficacy in a Malaysian college student sample.

Other studies have found that interest in science or math was positively associated with prior performance (i.e., mastery experience) and negatively associated with science and math anxiety among elementary school students and middle school students in Taiwan (Chiu, 2012; Jen, Lee, Chien, Hsu, & Chen, 2012). Interest was also positively correlated with outcome expectations and prior performance and negatively correlated with math anxiety in a sample of Singaporean pretertiary school students (average age = 17.8 years; Lim & Chapman, 2013). Finally, among Japanese college students, outcome expectations were predictive of career interest after controlling for self-efficacy across all six Holland (1997) career domains; on the other hand, self-efficacy offered incremental utility in predicting interest above and beyond outcome expectations only in the Investigative and Artistic domains (Adachi, 2004). Together, the above studies offer support for SCCT hypotheses regarding the relations among self-efficacy, outcome expectations, interest, and choice goals across different career domains and student populations (i.e., elementary students to college students). A few of these studies also demonstrated theoretically expected relations between interest and sources of efficacy beliefs (mastery experience, anxiety) in younger student groups.

A few studies have examined a mix of immediate (e.g., self-efficacy) and more distal (e.g., interest and supports) predictors of performance attainments. Essentially, these studies tested modified versions of the performance model by including predictor or outcome variables from other SCCT models. For example, Stephen (2008) studied the relations between performance (end of semester grades), interest, self-efficacy, and sources of self-efficacy in three subject areas (math, science, and English) among high school students in India. Her findings showed strong support for a self-efficacy → interest → performance pathway in math but partial support for the same pathway in Science and English. Furthermore, her findings regarding prior mastery experience, emotional arousal, social persuasion, and vicarious learning as predictors of self-efficacy did not reveal a consistent pattern across these three subject areas. In another study with Korean middle school students, Lee, Lee, and Bong (2014) found positive correlations among self-efficacy, interest, and end of semester grades across four academic subjects (Korean, English, science, and math). However, Yang’s (2004) study showed that self-efficacy and social support did not predict academic achievement among vocational college students in Taiwan.

Overall, SCCT research conducted in Asia provides evidence for supports, self-efficacy, and goal progress as mediators of the relations of personality and cultural variables to satisfaction outcomes among Asian college students. Existing evidence is also consistent with the hypothesized relations among self-efficacy, outcome expectations, interest, and choice goals. However, the performance model has received limited attention with inconsistent findings. Data of these studies were analyzed using different statistical techniques (e.g., structural equation modeling [SEM], regression, and correlation), and the majority of them included college students and middle school/high school students. It should be noted that most of these studies recruited samples from East Asian countries, which suggests that more empirical attention should be given to Southeast Asian or South Asian countries (e.g., Philippines, Vietnam, Indonesia, and Pakistan). Future research should also be extended to adult samples from this part of the world.

Europe

Fourteen studies (involving 15 samples) were conducted in European countries, with most of them testing the choice or satisfaction models (see Table 1). Consistent with predictions of the satisfaction model, Lent, Taveira, and Lobo’s (2012) cross-sectional findings showed that positive affect was predictive of academic satisfaction, academic stress, and life satisfaction both directly and indirectly via the pathways formed primarily by academic supports and self-efficacy among Portuguese undergraduate students. Academic self-efficacy partially mediated the relation of academic supports to academic satisfaction and stress, while academic goal progress partially channeled the effect of academic supports on academic stress. In addition to college students, Lent et al. (2011) tested the satisfaction model in a sample of Italian teachers. This study showed that efficacy-relevant supports and favorable work conditions partially mediated the effect of positive affect on teachers’ job satisfaction and life satisfaction; however, contrary to hypotheses, self-efficacy and/or goal progress were not part of the direct pathways to job satisfaction among these Italian teachers.

Using longitudinal data (in which Time 1 and Time 2 data points were separated by 15 weeks) from two Portuguese undergraduate student samples, Lent and colleagues (Lent et al., 2012; Lent, Taveira, Sheu, & Singley, 2009) produced findings that were partially consistent with the temporal ordering hypothesized in the satisfaction model. For example, while Time 1 supports and self-efficacy were predictive of Time 2 goal progress and satisfaction outcomes (e.g., academic satisfaction and stress), Time 1 goal progress did not predict Time 2 satisfaction outcomes as posited. Additionally, these two samples demonstrated reciprocal relations of positive affect to both self-efficacy and supports.

Another longitudinal study (6 months between Time 1 and Time 2 data points) conducted with a sample of employed adults in Belgium showed that Time 2 perceived barriers, goal progress, and self-efficacy mediated the relations of Time 1 personality traits (extraversion and emotional stability) and self-efficacy to Time 2 career satisfaction (Verbruggen & Sels, 2010). Interestingly, Time 1 self-efficacy was found to negatively predict Time 2 perceived barriers, which suggests that more confident adults were likely to perceive fewer barriers to their career goals over time. However, the authors did not test SCCT’s prediction that barriers at Time 1 would predict self-efficacy at Time 2. Abele and Spurk (2009) tracked a sample of graduate students in Germany for 7 years into their adult lives as full-time employees and found that occupational self-efficacy predicted career satisfaction 7 years later after controlling for other variables, such as salary and job status.

Existing cross-sectional and longitudinal research conducted in Europe did not produce consistent findings regarding the mediating roles of self-efficacy and goal progress in the satisfaction model. Also, outcome expectations or perceived work conditions were excluded in most of these studies; thus, they provided only a partial test of the satisfaction model. Additional research is needed to examine the functions of self-efficacy and goal progress and to test a fuller set of variables. Longitudinal designs with more than two data points will also be helpful in testing temporal hypotheses in the satisfaction model.

Another set of studies was devoted to testing the core portion of the choice model, including choice goals, interest, self-efficacy, outcome expectations, and proximal contextual supports and barriers. For example, Blanco (2011) found that prior performance in math and statistics was predictive of Spanish undergraduate students’ intention to use statistics in academic or professional activities, both directly and indirectly via statistics self-efficacy, outcome expectations, and interests. In another Spanish college student sample, Inda, Rodríguez, and Peña (2013) demonstrated that self-efficacy, outcome expectations, and interest in engineering partially channeled the effects of contextual supports and barriers on female and male students’ intention to pursue engineering majors.

Lent and colleagues (Lent, Brown, Nota, & Soresi, 2003; Lent, Paixão, da Silva, & Leitão, 2010) tested the choice model across Holland’s six occupational themes (R, I, A, S, E, and C) among high school students in Italy and Portugal. Findings of these two cross-sectional studies provided evidence for the applicability of the choice model across these career domains and showed that self-efficacy and outcome expectations were jointly predictive of career interest, which in turn predicted choice consideration. However, social supports were associated with occupational consideration indirectly through other variables (e.g., self-efficacy) rather than directly, and the indirect effect of social barriers on occupational consideration was observed in some career domains (R, A, and C) but not in others (I, S, and E). Two other studies provided inconsistent support for a few choice-related hypotheses. Primé et al. (2010) reported no relation between anticipated occupational choice and interest among Italian children, whereas Dickhäuser, Reuter, and Hilling (2005) found that self-reported grades (prior mastery experience) in Biology and Chemistry were positively correlated with selection of courses (choice action) in these two subjects among seventh grade students in Germany.

In addition to the above research that included choice goals or choice action as the dependent variables, a few other European studies focused on interest or performance as the outcome variables. Math performance was positively related to interest in math among sixth graders in Germany (Preckel, Goetz, Pekrun, & Kleine, 2008), and self-efficacy in math was also found to be positively correlated with interest in math in a sample of ninth graders in the same country (Schukajlow et al., 2012). Finally, Macher, Paechter, Papousek, and Ruggeri’s (2012) study offered some evidence related to the performance model, showing that lower anxiety in statistics was associated with higher interest and less procrastination (choice action) in carrying out statistics-related tasks, and lower anxiety and higher interest were related to better performance on a statistics examination among undergraduate students in Austria.

Most of the empirical inquiries conducted in Europe were devoted to testing the choice/interest and satisfaction model. Compared with the research conducted in Asia, a few longitudinal European studies added to the literature by offering evidence for some temporal hypotheses (e.g., self-efficacy at Time 1 → goal progress at Time 2) in the satisfaction model. On balance, evidence gathered in Europe supports the function of self-efficacy in the satisfaction model and those of self-efficacy and outcome expectations in the choice model in mediating the relations of proximal contextual factors and/or personality traits to academic/work satisfaction or choice intentions. While both self-efficacy and outcome expectations were included in the majority of the studies that tested the choice model, outcome expectations were omitted in most of the studies that examined the satisfaction model. Furthermore, performance accomplishments as the outcome variable have been understudied in European samples. Existing studies covered a wide range of age groups but focused primarily on college student and high school/middle school student populations. Our literature search did not identify any studies from Northern European (e.g., Norway) or Eastern European (e.g., Czech Republic) countries, which suggests the need for future research testing the generalizability of the SCCT models in these European sectors.

Other Regions

In addition to studies carried out in Asia and Europe, samples have been recruited from other parts of the world to test SCCT models. In total, there are 11 studies (12 samples) conducted in Canada, Australia, and countries in Africa and the Middle East. Lent et al. (2014) gathered data from college students in Angola and Mozambique and demonstrated that the satisfaction model was applicable to both of these African student samples, and there were no differences by gender or by country. Specifically, positive affect was found to predict academic satisfaction and life satisfaction both directly and indirectly through academic supports, self-efficacy, and goal progress. Similar results were also found in a large sample of teachers in Abu Dhabi, the United Arab Emirates (Badri, Mohaidat, Ferrandino, & El Mourad, 2013). As hypothesized by SCCT, several pathways, which consisted of perceived organizational supports, self-efficacy, favorable work conditions (i.e., person-organization fit and needs-supplies fit), and progress at work-related tasks, were found to partially mediate the effect of positive affect on these teachers’ work satisfaction. These two studies offer preliminary evidence for the validity of the satisfaction model for college students and adults in Africa and the Middle East.

Researchers in Australia, Canada, Lebanon, and Turkey have also explored variables included in the choice model. Martin and colleagues (Martin, Anderson, Bobis, Vellar, & Way, 2012; Martin, Bobis, Anderson, Way, & Vellar, 2011) conducted two multilevel studies with middle school students in Australia and found that, at the student level, choice goals and choice actions in math were positively correlated with self-efficacy and outcome expectations and negatively correlated with anxiety about math. Positive correlations among choice goals or choice actions, interest, and outcome expectations were also observed in the data gathered from a large sample of middle school students in Canada (Areepattamannil, Freeman, & Klinger, 2011) and among college students in Lebanon (Naccache, 2012). Furthermore, Sahin’s (2008) path analysis showed that interest completely mediated the effect of self-efficacy, and partially mediated the effect of outcome expectations, on university faculty members’ intention to use educational technology in Turkey. However, in a Turkish high school student sample (Özyürek, 2005), intentions to choose math-related majors were found to be negatively correlated with self-efficacy in solving math problems and positively correlated with vicarious learning experience in math, which is partially consistent with the predictions in the choice model.

Finally, research conducted in Israel demonstrated the positive relations between interest and self-efficacy in science and between interest in science, technology, engineering, and mathematics (STEM)-related fields and prior performance in math among middle school and high school students (Chachashvili-Bolotin, Milner-Bolotin, & Lissitsa, 2016; Sasson & Cohen, 2013). Ratelle, Larose, Guay, and Senecal’s (2005) longitudinal study showed that current science performance (mastery experience) was associated with persistence in science in a sample of Canadian high school students.

Results of SCCT research conducted in regions other than Asia and Europe offer initial evidence for key assumptions (e.g., self-efficacy and goal progress as mediators) in the satisfaction model. They also demonstrate some of the bivariate relations (e.g., between self-efficacy and choice goals, outcome expectations and choice goals, self-efficacy and interest, mastery experience, and persistence) posited in other SCCT models. However, these findings should be interpreted with caution because they were based on samples residing in widely varying cultural contexts. How different cultural norms might influence the applicability of SCCT models in these countries is largely unknown and warrants empirical attention.

Expanded Study of Person Input, Cultural, and Contextual Variables

Empirical evidence reviewed in this article has focused primarily on the core portions of the SCCT models, including self-efficacy, outcome expectations, interest, choice goals or goal progress, satisfaction, and proximal supports and barriers. Limited attention has been given to person input variables or to cultural and more distal contextual factors. Although these variables are positioned at the periphery of some models (e.g., choice) or less explicitly explicated in SCCT, they merit more empirical attention because they could potentially improve SCCT’s applicability to people in different countries. We will discuss study of these variables in the context of future SCCT research in different cultures.

In terms of personality traits or affective predispositions as person input variables, findings of several studies consistently showed that positive affect was directly and indirectly, through person-cognitive variables (e.g., self-efficacy), predictive of academic satisfaction among college students, and job satisfaction among adults in some European and African countries (Lent et al., 2009, 2014). Extraversion and emotional stability have also been found to produce direct and indirect relations to academic satisfaction and stress among Asian college students as hypothesized in the satisfaction model (Sheu et al., 2014; Sheu et al., 2017). In addition to these Big Five personality traits and positive affect, international researchers should explore how other trait-like variables, such as optimism (Carver, Scheier, Miller, & Fulford, 2009) and locus of control (Rotter, 1966), may function as person input variables in SCCT models.

For example, locus of control has been found to differentially predict human behaviors in different cultural contexts. Côte, Mizokami, Roberts, and Nakama (2016) found that college students in the United States had a higher internal locus of control than their counterparts in Japan, and internal locus of control was a predictor of identity development only for American but not for Japanese students. Furthermore, locus of control has been found to be related to vocational outcome expectations among undergraduate students in Turkey (Işik, 2013) and job satisfaction among adults in Australia (Wu, Griffin, & Parker, 2015) and China (Guan et al., 2012). Inclusion of additional trait variables can provide the opportunity to explore how social cognitive predictors mediate effects of these variables on different vocational outcomes within the same country or across different cultural contexts.

Although SCCT has attracted attention from researchers in different parts of the world, the majority of studies identified in this review did not include cultural or social/contextual (e.g., opportunity structure) variables in testing SCCT hypotheses. To capture individual differences in individualism and collectivism, Sheu and his colleagues (Sheu et al., 2014; Sheu et al., 2017) included independent and interdependent self-construals in testing the satisfaction model. Their findings showed that these variables contributed differentially to the prediction of academic satisfaction, academic stress, and life satisfaction above and beyond personality traits and other SCCT predictors among Chinese, Taiwanese, and Singaporean college students. Future international SCCT research should continue to explore how cultural orientations may interact with social cognitive variables in forecasting different vocational outcomes.

In addition to factors that tap broad, general cultural orientations, variables relevant to a specific cultural context should be considered when conducting SCCT research. For example, in some Asian societies, “loss of face” is a core cultural value and involvement of family members in the choice-making process may be considered as culturally appropriate. Both of these variables (saving face and family involvement) could affect vocational behaviors and satisfaction as individuals in these societies may be more satisfied with selecting a career that fits the needs of the family or improves the family’s reputation than on the basis of personal interest. Therefore, it is possible that inclusion of these specific cultural variables could increase the predictive utility of SCCT models in particular Asian cultures.

SCCT research has focused more on contextual factors that are within the person’s immediate environment or proximal to his or her choice of a career or pursuit of satisfaction. Such factors usually involve social supports (e.g., feelings of support for a decision from important people in one’s life) or barriers (e.g., receipt of negative comments about a choice from friends; Lent et al., 2010). However, contextual factors that are more removed from one’s immediate environment or social circles have been left largely unexamined. For example, according to the Statistical Office of the European Union, unemployment rates, especially among young adults in some European countries (e.g., Portugal, Italy, and Spain), have remained high since the last financial crisis (Eurostat, 2016). Additionally, issues such as legal and illegal immigrant status and the increasing size of the temporary labor force have changed the landscape of job markets in European countries.

These opportunity structure variables (e.g., job market, social mobility, and immigration policy) can facilitate or hinder individuals’ interest development and vocational choice. Although these factors are largely out of personal control, the individual’s perception of and response to these factors may promote or diminish volitional control over his or her vocational behaviors and satisfaction with the chosen occupation. Therefore, in addition to proximal social supports and barriers that are likely to be generally applicable to individuals across different countries, opportunity structure variables, which are relevant to a specific society or culture, might be studied as contextual factors in international SCCT research. For instance, in countries where financial crisis has been an ongoing struggle, how people perceive and respond to shrinking or unstable job markets or diminishing government aid to the unemployed could be included in SCCT models to explore the interactions between such contextual factors and other predictors in forecasting vocational behaviors and satisfaction.

Sampling and Methodological Considerations

Based on this review, most of the international SCCT research has tested hypotheses of the interest/choice and satisfaction models. Empirical efforts have yet to test the performance and persistence models to the same degree, perhaps because of the difficulties in obtaining prospective performance or persistence data. This literature will benefit from research that explores not only how individuals develop interest in and select an occupation but also how they perform or persist in their chosen fields. Similar to studies conducted in the United States (Sheu et al., 2010), research performed by international scholars has focused mostly on academic majors or occupations in Holland’s R and I themes. One direction for future inquiry is to extend this line of research to other, less studied domains. Adolescents and college students have also constituted the majority of samples recruited outside the United States, which limits the generalizability of these findings to other populations, such as adult workers or those who are close to retirement.

The geographic distribution of international SCCT research showed that evidence has mostly been accumulated in Asian and European countries. More empirical attention is still needed in these two regions (e.g., Southeast Asian countries and Northern European countries); however, future research should also be extended to testing SCCT models with samples from other parts of the world such as Central America, South America, and Africa. When testing SCCT hypotheses in different countries, relevant contextual and cultural factors, such as those discussed in the previous section, should be included, which may improve the explanatory utility and cultural relevance of SCCT models to individuals who reside in different countries and cultures.

In terms of research design and methodology, most findings were derived from cross-sectional data. A few studies gathered longitudinal data at two points in time to examine reciprocal relations among SCCT variables (e.g., Lent et al., 2012). Additional longitudinal data with three or more data points will shed increased light on the hypothesized temporal ordering of SCCT variables. Also, while studies have been conducted in the United States to test the effectiveness of SCCT-derived interventions (see Sheu & Lent, 2015 for a review), similar outcome research has yet to appear in other countries. Although a mix of univariate and multivariate data analyses have been used, future research should utilize techniques (e.g., path analysis) that are more suitable for testing the multivariate nature of SCCT models. Additionally, most studies have examined the relations among a subset of SCCT variables, such as those in the core portion of the models (e.g., self-efficacy, outcome expectations, and goals). Where possible, empirical efforts should be devoted to testing SCCT models with a fuller set of variables, which will allow for more accurate estimates of each variable’s unique contribution to the prediction of vocational outcomes.

Finally, cross-national research designs have started to receive attention from researchers in testing SCCT hypotheses (Lent et al., 2014; Sheu et al., 2014). Although direct cross-national comparisons entail additional layers of complexity (e.g., language translation and measurement equivalence), the findings of such studies would offer stronger evidence for the generalizability of SCCT. Multigroup SEM techniques are particularly suited for comparing data collected in different countries (e.g., China vs. the United States) or regions (e.g., Western vs. Eastern Europe) and allow for identifying which parts of the models might be universally applicable and which parts may be more culturally specific to individuals in a particular region, country, or culture. This approach can also be used to test the applicability of SCCT models to groups of individuals who share similar cultural values or backgrounds but reside in different parts of the world, such as Asians who live in Hong Kong, Canada, or the United States or Hispanics who live in Central America, South America, or Europe. Such cross-cultural or cross-national research will provide more fine-grained evidence regarding the validity of the SCCT models.