|
Article Excerpt
The Attributional Style Questionnaire (ASQ; C. Peterson et al., 1982) has been the subject of some debate challenging its usefulness in research. This study examines the ASQ component structure for Central American immigrants to the United States. Research and cultural implications are discussed.
**********
Attributional style refers to the consistent manner in which one attributes causal explanation to events that occur in one's life across three dimensions: an internal-external locus dimension, a stable-unstable or stability dimension, and a specific-global or globality dimension (Ickes, 1988). Abramson, Seligman, and Teasdale (1978) suggested that people who have a consistent internal, stable, and global response to perceived negative events exhibit a depressogenic attributional style. This depressogenic attributional style, in conjunction with a stressful negative life event (diathesis-stress), may predict subsequent development of a subtype of depression notable for the role a profound sense of hopelessness plays in its etiology (Abramson, Alloy, & Metalsky, 1995; Abramson, Metalsky, & Alloy, 1989; Metalsky, Abramson, Seligman, Semmel, & Peterson, 1982; Sweeney, Anderson, & Bailey, 1986).
To measure attributional style, Peterson et al. (1982) designed the Attributional Style Questionnaire (ASQ). This 36-item, self-report, paper-and-pencil measure presents 12 hypothetical events (6 positive, 6 negative) that participants rate according to three attributional dimensions: internal-external (Locus), stability-instability (Stability), globality-specificity (Globality), and importance to self. Scores on each pole of the three dimensions yield six subscores (three negative and three positive) for a composite negative attribution style score and a composite positive attributional style score (Peterson et al., 1982). In addition, the six positive and negative events are divided into either individual achievement or affiliation/interpersonal (three each). However, this delineation appears to be both more relevant, regarding responses to positive events rather than negative events, and somewhat elusive. Critics charge that some of the "achievement" items may incorporate more "affiliation" perspectives (Corr & Gray, 1996; Reivich, 1995).
Internal, unstable, and specific attributions lie on the lower end of the scoring spectrum, whereas external, stable, and global attributions receive higher scores. The three scores for negative events may be used to measure depressogenic attributional style. Peterson et al. (1982) reported Cronbach's alpha coefficients of .21 for Locus, .56 for Stability, .66 for Globality, and .72 for the composite score for negative events. Test-retest reliability for the negative composite score (r = .64) is considered "temporally consistent" among nonclinical individuals (Tennen & Herzberger, 1985, p. 23).
The ASQ's negative-event scores have been used extensively in the literature to measure depressogenic attributional style. Higher scores on negative attributional style are correlated with general dysphoria (Corr & Gray, 1996) and, more specifically, depressive symptomatology (cf. Sweeney et al., 1986). A preliminary search on PsycINFO in October 2003 for the measure yielded 47 published articles and many more dissertations. The measure is readily available, easy to administer, and free to use. Given the measure's popularity, it is not surprising that its factor structure should come under some scrutiny (Arntz, Gerlsma, & Albersnagel, 1985; Corr & Gray, 1996; Cutrona, Russell, & Jones, 1984; Higgins, Zumbo, & Hay, 1999). The results have been inconclusive; none have offered strong support for the hypothesized dimensions. Cutrona et al. modified their structural model four times, to account for goodness-of-fit for the six negative events, and found that although two of the factors, Stability and Globality, offered evidence of cross-situational consistency (average of 20.8% and 23.5% of variance explained respectively), the factor Locus was decidedly weak (average of 8.5% variance explained).
Arntz et al. (1985) used a confirmatory factor analysis on all three scales for both positive and negative events and found that the factors accounted for very little of the hypothesized variance in the ASQ scales. Single factors explained between 3.7% and 7.4%, whereas the whole model explained between 30.1% and 34.1% of the variance. Corr and Gray (1996) conducted a principal component analysis on the total ASQ and found four factors (a combined Stability/Globality score for negative events, the composite positive attributional style on affiliation events score, a combined Stability/Globality score for positive affiliation situations, and Locus for negative events) explaining 64% of the variance.
Most research conducted on attributional style has been conducted with U.S. Americans and has not examined ethnicity and cultures as separate variables. Yet, it seems reasonable to assume that people explain behavior in every culture; how they explain that behavior, however, may vary by culture and by historical time (Anderson & Deuser, 1991). Causal attributions may be the focus of interest in Western societies where cultural norms tend to reward individuals "who 'predict and control' the events that...
|
