|
Abstract. Lent and Brown (2006) suggest guidelines for creating and adapting assessment tools based on the Social Cognitive Career Theory (SCCT, Lent, Brown, & Hackett, 1994). In the last years, this theory has been the subject of substantial research, both basic and theoretical. The authors indicate that any assessment of SCCT should (a) contextualize measures to make sure they are grounded in a particular, domain-specific context; (b) be reasonably comprehensive in sampling the domain, designing multifaceted measures when the criterion is correspondingly complex; and (c) ensure compatibility between predictors and criteria along key dimensions, including content, context, temporal orientation, and level of specificity. Additionally, it is important to utilize reliable and valid tools. Without sound measures, it is difficult, if not impossible, to establish whether theory-discrepant findings are attributable to problems with the theory, flaws in operationalizing it, or both. In Argentina, Cupani and Gnavi (2007) assessed a model of academic performance in mathematics, based on the SCCT. Towards this aim, the authors adapted the subscales for mathematics outcome expectancies and performance goals of the Middle School Self-Efficacy Scale (Fouad, Smith, & Enochs, 1997). They found that the goals subscale has a simple factorial structure with adequate internal consistence, although it did not predict academic performance in mathematics. Moreover, the subscale for mathematics outcome expectancies exhibited low internal consistence and some of its items did not transfer well to our cultural setting. Therefore, two follow-up studies were conducted to improve the psychometric properties of both scales. The first study employed two focus groups (n per group ≥ 8) and aimed at generating ideas on the student’s expectations of results and goals on academic achievement in our cultural setting. The information gathered was used to write 7 new items for the goals subscale and 12 items for the outcome expectancies. These items were then tested for clarity and understandability in a sample of adolescents. Language amendments were also made, yielding a new version of the scales that measures goals (11 items) and the scale on outcome expectancies (13 items). On the second study, these scales were administered to a sample of 420 adolescents (M = 13.84, DT = .76). The internal structure of the scales was examined through exploratory and confirmatory factor analysis and their internal consistency by Cronbach's alpha. The predictive validity for academic achievement in mathematics was also analyzed. The scale of logical-mathematical self-efficacy from the (revised) Self-efficacy Inventory of Multiple Intelligences (Perez & Cupani, 2008) was also administered. Exploratory and confirmatory factor analysis revealed that a single-factor structure for the scale of performance goals (GFI .92; CFI .95, RMSEA .08) and the scale for mathematics outcome expectancies (GFI .95; CFI .96, RMSEA .06) is the most appropriate model for the data gathered. Both scales had optimal Cronbach's alpha values (α = 86 and 85; for performance goals and outcome expectancies, respectively). The study on predictive validity also showed that logic-mathematic self-efficacy beliefs and achievement goals in mathematics explain 32% of the variance of math school performance. The results show that academic achievement in mathematics is partially explained by the model. In summary, both scales allow a contextualized measurement of outcome expectancies and performance goals on mathematics of teenagers from our cultural area. These scales apparently possess adequate psychometric properties, with a clear internal structure, and adequate internal consistence. Future application of path analysis will allow a more precise identification of the interrelations between variables and their direct and indirect effects upon academic achievement in mathematics. Key words: outcome expectancies - performance goals – Academic Achievement
|
