Career Attitudes 14
Running head: A LONGITUDINAL EXAMINATION OF SCCT
Career Attitudes of Declared Psychology Majors: A Longitudinal Examination of Social Cognitive Career Theory
Todd A. Brahler and Linda Mezydlo Subich
The University of Akron
Abstract
The purpose of this study was to examine the nature of the temporal relationships between the Social Cognitive Career Theory (SCCT; Lent, Brown, & Hackett, 1994) predictor and criterion variables. Psychology students’ (N = 74) concurrent levels of self-efficacy, outcome expectations, interests, and choice intentions were measured four times during a 15-week semester. Results of the correlational analyses indicated that the correlations between (a) self-efficacy and interests, (b) outcome expectations and interests, (c) outcome expectations and choice intentions, and (d) interests and choice intentions tended to increase over time. Hierarchical regression analyses confirmed the additive nature self-efficacy and outcome expectations in predicting interests, but when predicting choice intentions, self-efficacy and outcome expectations did not function in a manner that was consistent with the theory. The results suggest that relations between SCCT predictor and criterion variables strengthen over time.
Career Attitudes of Declared Psychology Majors: A Longitudinal Examination of Social Cognitive Career Theory
Social Cognitive Career Theory (SCCT; Lent, Brown, & Hackett, 1994) extended Bandura’s (1986) Social Cognitive Theory to the vocational domain by specifying a model of the long-term effects of learning experiences on an individual’s academic and career pursuits. According to SCCT, an individual’s self-efficacy and outcome expectations—both of which are influenced by learning experiences consisting of “educational and occupationally relevant activities” (Lent et al., 1994, p. 104)—underlie his or her interests in potentially rewarding activities, and these interests influence subsequent choice intentions and behaviors (e.g., declaring a major and enrolling in an academic program). Ensuing performance attainments in these learning situations may enhance or diminish existing levels of self-efficacy and outcome expectations which, in turn, may influence subsequent academic interests and choices. Thus, the theoretical model developed by Lent et al. (1994)—consisting of an integrated network of causal pathways—is “cyclical in nature and longitudinal in scope” (Nauta & Epperson, 2003, p. 449). In short, SCCT epitomizes a process that is not stagnant but is continuously in operation.
Lent et al. (1994) positioned self-efficacy and outcome expectations at the heart of their model, but considered self-efficacy to be the dominant construct due to its hypothetical direct influence on outcome expectations. Strong self-efficacy beliefs are purported to be especially important when quality performance underlies positive outcomes (Bandura, 1986) and may even sustain an individual’s efforts over time regardless of unforeseen consequences or repeated failures (Bandura, 1989). On the other hand, an individual with low self-efficacy may hesitate to initiate a course of action, exert less effort, or give up entirely when faced with obstacles, even if outcomes are positive and highly desirable (Bandura, 1986). Research has supported the importance of self-efficacy within the framework of SCCT by identifying it as a significant predictor of career and academic interests and choice intentions (Gainor & Lent, 1998; Lent, Lopez, & Bieschke, 1993; Lent, Brown, et al., 2003; Lent et al., 2005).
Although past research has tended to focus more on self-efficacy (Diegelman & Subich, 2001), outcome expectations are also considered to be an important component of SCCT (Lent et al., 1994). Defined by Lent et al. (1994) as “imagined consequences of performing particular behaviors” (p. 83), outcome expectations may be especially important in vocational behavior when outcomes are not dependent on the quality of one’s performance (Bandura, 1989; Lent et al., 1994). Within the SCCT model, outcome expectations are posited to have pathways to both interests and choice intentions (Lent et al., 1994), with relations to the criterion variables that are similar in magnitude to that of self-efficacy (cf. Bishop & Bieschke, 1998; Diegelman & Subich, 2001; Lent, Lopez, & Bieschke, 1991; Lent et al., 1993). Furthermore, past research has established outcome expectations as a significant predictor of interests and choice intentions (Diegelman & Subich, 2001).
Longitudinal Aspects of SCCT Constructs
Historically, most SCCT researchers have used cross-sectional designs and correlational analyses to support Lent et al.’s (1994) hypothetical relationships between self-efficacy, outcome expectations, and the criterion variables (Nauta & Epperson, 2003). Although such studies constitute an impressive body of knowledge on the theory, one particular aspect has yet to be fully investigated: temporal influences on predictor-criterion variables’ relations. Lent et al. (1994) conceptualized SCCT as a dynamic process and proposed that “at any given point in time (or for particular individuals over time), certain variables will carry differential causal weight” (p. 88), which suggests that the relationships between the predictor and criterion variables may change over time. The observation and documentation of this phenomenon is all but impossible via traditional cross-sectional research methodologies and, to date, longitudinal studies are underrepresented in the SCCT literature.
At least one study of SCCT, however, has given some indication that the relations between the predictor and criterion variables can change over time. In a longitudinal study spanning two semesters, Nauta, Kahn, Angell, and Cantarelli (2002) investigated the bidirectional relationship between self-efficacy and interests. Using the six General Occupational Theme scores of the Strong Interest Inventory, they measured students’ vocational interests in each of Holland’s RIASEC domains and reported (a) six significant efficacy-to-interest paths after a 3-month lag (i.e., January, 2001 to April, 2001), (b) one significant efficacy-to-interest path after a 4-month lag (i.e., September, 2000 to January, 2001), (c) and four significant efficacy-to-interest paths after a 7-month lag (i.e., September, 2000 to April, 2001). In light of what appeared to be a strengthening relationship in the efficacy-to-interest link from the 4-month to 7-month period, they concluded that “SCCT becomes a more accurate reflection of the relationship between these two variables [self-efficacy and interests] as one considers them over longer periods of time” (Nauta et al., 2002, p. 298).
Although the Nauta et al. (2002) study was sufficient to provide evidence of time-induced changes in the self-efficacy-interest relation, it suffered from several shortcomings which they acknowledged. First, the method by which data were collected may have introduced extraneous variance into students’ perceived levels of self-efficacy and/or interests. The 4-month and 7-month periods spanned two semesters, and this included the holiday break. Nauta et al. (2002) stated that they could not control for any learning experiences encountered by the students during that time. Second, their study focused on the relation between self-efficacy and interests, but did not take into account the role of outcome expectations. This is especially important since past research has provided evidence that outcome expectations partially mediate the effect of self-efficacy on interests (Bishop & Bieschke, 1998). Third, Nauta et al. (2002) did not consider how goals (i.e., choice intentions) could influence the future relations between self-efficacy and interests. In their words, “a person’s goals and actions are determined, in part, by their interests. The experiences that a person obtains as a result of their actions then provide learning experiences that feed back into the SCCT model in the form of self-efficacy sources” (p. 299). Nauta et al.’s (2002) implications were clear. Choice intentions could impact choice behaviors which, in turn, have an effect on the types of learning experiences in which the students engage. Eventually, these learning experiences not only may affect subsequent levels of self-efficacy, but outcome expectations as well. As a consequent, the existing relations between these constructs and the criterion variables could change over an extended period of time.
Purpose of the Current Study
The purpose of this study was to examine further the nature of the relationships between the SCCT predictor and criterion variables over an extended period of time. The conception of SCCT as a dynamic process, in conjunction with the findings by Nauta et al. (2002), suggests that the relations between the predictor and criterion variables are open to temporal influences. Because of the potential impact that these relations could have on students’ academic and vocational exploratory behavior, this phenomenon warrants further investigation.
Due to the cyclical nature of the SCCT model, it was determined that a longitudinal study was the best way to investigate this phenomenon. In this manner, students’ perceived levels of SCCT constructs could be measured repeatedly over a particular time period. To circumvent the shortcomings of the Nauta et al. (2002) study, the following procedures were implemented. First, in order to limit the effect of extraneous variance on student’s career attitudes, the current study was conducted over a 15-week period (i.e., one academic semester) in a setting that offered an occupationally relevant learning environment: a psychology careers class. In addition, only those students who were enrolled in the class were invited to participate in the study; a procedure which Nauta et al. (2002) suggested could facilitate “maximize participant completion of multiple data waves” (p. 299).
The causal paths proposed by Lent et al. (1994) imply that relations exist among all the variables in the model; not just the two (i.e., self-efficacy and interests) that were investigated by Nauta et al. (2002). Thus, in order to fully assess the nature of the temporal relations between SCCT predictor and criterion variables, this study measured students’ perceived levels of outcome expectations and choice intentions, in addition to self-efficacy and interests. Correlation coefficients were then compared at distinct points in time to determine whether or not the relations between the predictor and criterion variables changed over the course of the semester. Since past research has amply supported their hypothetical relations, it was expected that the concurrent correlations between self-efficacy, outcome expectations, interests, and choice intentions would be positive and significant at each data collection point. However, the lack of empirical evidence on how and when these relationships may change prevented the formulation of any formal hypothesis regarding temporal patterns of correlations.
This study also addressed the nature of predictor-criterion relations by analyzing hierarchical regression models. According to SCCT, self-efficacy and outcome expectations would combine in an additive manner in the prediction of interests (Lent et al., 1994). Furthermore, the effects of self-efficacy and outcome expectations on choice intentions are channeled through interests (Lent et al., 1994). However, it is possible that changes in the relations between SCCT constructs could affect how the predictors function in a regression model. Therefore, full regression models were built and compared in the current study to determine if the functioning of the predictor variables at every data collection point was in accordance with the theory.
Methods
Participants
The participants for this study were recruited from a one-credit hour psychology careers class at a large Midwestern university. One hundred eighteen students completed the first set of data, 117 completed the second set, and 79 completed the third set and fourth set. Only 74 students completed all four sets of data. This final participant pool (N = 74) consisted of 18 men and 56 women. There were 47 freshman, 12 sophomores, 6 juniors, 8 seniors, and 1 postbaccalaureate student. Most were psychology majors (92.3%), and the remaining 7.6% were either non-psychology majors or were undeclared. Their ages ranged from 18 to 53 (M = 21.49 years, SD = 5.92). They were primarily Caucasian (81.1%), but 13.5% were African American, 4.1% were Hispanic American, and the remaining 1.4% were Asian and Indian Americans. All of the students received extra credit for their participation.
Measures
The construction and content of the current measure was inspired by the one used by Diegelman and Subich (2001). Those researchers changed scales used by Fouad and Smith (1997) and Riggs, Warka, Babasa, Betancourt, and Hooker (1994) in order to assess students’ self-efficacy and outcome expectations for a bachelor’s degree in psychology (as cited by Diegelman & Subich, 2001). In a similar fashion, the survey in the current study was developed in order to assess students’ self-efficacy, outcome expectations, interests, and choice intentions regarding a career in the field of psychology.
Individual items addressed specific academic and career issues relevant to the field of psychology. For each item, participants were required to indicate their level of agreement—using a 7-point scale ranging from strongly disagree (1) to strongly agree (7)—by writing the appropriate number on a blank line located at the end of the statement. Item scores were summed and divided by the total number of items, yielding a total score that could range from 1 to 7, with a higher score indicating a higher level of a particular construct. In addition, the order of the items in each subtest was changed at each data collection point in order to reduce familiarity effects on participants’ responses.
Psychology Career Self-efficacy. Lent et al. (1994) defined self-efficacy as “personal convictions about one’s generative capabilities to negotiate specific task or situational challenges” (p. 86). In accordance with their definition, psychology career self-efficacy was measured using a scale consisting of 13 items which assessed participants’ confidence in their ability to (a) perform specific tasks relevant to a career in psychology (e.g., “I am able to master the statistical skills necessary to analyze data”), (b) engage in certain activities relevant to a career in psychology (e.g., “I am confident that I can give an oral presentations of a psychology-related topic to a group of people), (c) achieve career-relevant milestones (e.g., “I am confident that I can apply for and be accepted into a psychology graduate program”), and (d) be successful in the field of psychology (e.g., “I believe I can be successful in a psychology career”). Three of the items were negatively worded and were reverse-scored.
Psychology Career Outcome Expectations. Lent et al. (1994) defined outcome expectations as “personal beliefs about probable response outcomes . . . [involving] the imagined consequences of performing particular behaviors” (p. 83). Psychology career outcome expectations were measured via 13 items that reflect possible outcomes of a career in psychology. These items were based on Bandura’s (1986) three classes of outcome expectations: (a) physical (e.g., “A psychology career would make me financially secure”, (b) social (e.g., “Friends and family members whose opinions I value will have more respect for me if I pursue a career in psychology”), and (c) self-evaluative (e.g., “A career in psychology would provide the intellectual stimulation that I desire”). Two of the items were negatively worded and reverse-scored.