Extending social cognitive theory to counselor training: problems and prospects. (response to articles by Lisa M. Larson and Jeffrey A. Daniels in this issue, pp. 179, 219)

Larson and Daniels (1998 [this issue]) did a highly commendable job in reviewing the literature on counseling self-efficacy (CSE), and Larson's (1998 [this issue]) theory-building article was both ambitious and heuristic. We applaud both of these efforts, particularly the attempt to articulate the theoretical undergirding of the CSE literature and to detail implications for research and practice. Nevertheless, there are a number of areas in which we believe this work might be strengthened. Our major reactions concern the scope of Larson's theory-building effort and various accompanying definitional and measurement issues. We offer the following cautions, clarifications, and recommendations constructively, in the hope that they may augment and extend this line of theory and research.

ON CAFETERIA-STYLE THEORIZING

We had a deja vu experience in reading Larson. In the process of developing our social cognitive career theory (Lent, Brown, & Hackett, 1994), we struggled with and debated the extent to which we should draw primarily or exclusively from Bandura's (1986, 1997) social cognitive theory (SCT) versus incorporating relevant aspects of diverse theories. After much discussion, we went to the source Bandura's input was most instructive about the dangers of mixing constructs from different theories in a cafeteria-style manner. To quote from a letter Bandura sent us (A. Bandura, personal communication, March 1, 1993),

The only justification for tacking on the variables from [other] theories ...

would be if, after the sociocognitive variables are entered first into the

analyses, the variables from the other theories account for additional unique

variance. In short, you don't start with cafeteria theorizing. Rather, you go

the hybrid route only if evidence indicates the necessity to do so. Greater

scientific progress is achieved by applying more aspects of a unified

theory ... than by stringing together constructs from divergent theories.

In addressing Bandura's sage advice, we acknowledge that we did not entirely forgo elements of other theories. Because part of our intent was to encourage a unifying perspective on career development, we found it necessary to consider possible connections between social cognitive career theory and preexisting approaches to understanding career behavior. It was also necessary to clarify how certain theoretical constructs (e.g., goals, outcome expectations) were similar to, or different from, constructs of other theories. Finally, to fine-tune the theory's account of certain career-relevant processes (e.g., the means by which sociocognitive variables operate in tandem with person factors, such as gender and social-contextual variables), we found it necessary to fashion some novel predictions, sometimes involving concepts from other approaches that we viewed as theoretically compatible.

Because we are fellow travelers, we can empathize with Larson's desire to be comprehensive and creative in her theory-extension effort, while trying to achieve parsimony and solidarity with the overarching precepts of SCT. This is no easy struggle. Our own effort to balance these objectives resulted in a sort of layered approach to theorizing, starting off with the sociocognitive core of the theory, and then building on a second layer of theoretical analysis, focusing on "additional person, contextual, and experiential" influences on career development. In doing so, we tried to distinguish between newer and older, and more and less established, theory elements and predictions. Although it is difficult for us to assess the degree to which we succeeded in this balancing act, we did feel that Bandura's comments were very well taken and that the creative tension between theoretical breadth and precision was an extremely important, if challenging, aspect of our quest.

As one good turn deserves another, we would like to pass Bandura's advice on to Larson, believing that it might also benefit her theorizing. Although her effort to draw on numerous literatures, theoretical perspectives, and constructs is impressive in its breadth, the danger is that the theoretical coherence Larson is trying to achieve may become obscured under a torrent of eclectic elements. In essence, cafeteria-style theorizing may yield an approach that is both difficult to grasp conceptually and to implement in research and practice. It may also, ultimately, divert attention from the sociocognitive variables that Larson holds as central to counselor and supervision effectiveness. Hence, we suggest that future efforts to explicate the sociocognitive underpinnings of counselor effectiveness adhere more closely to basic SCT. In particular, it may …