Assessing Cross-Cultural Reliability of a Posttraumatic Growth Scale Among Black and White Survivors of Hurricane Katrina

Background and purpose: This study assessed cross-cultural reliability and validity of a 7-item scale measuring Post Traumatic Growth (PTG) among black and white Hurricane Katrina survivors. In contrast to deficit models of psychopathology that served social work in the past, posttraumatic growth (PTG) theory is in line with a renewed focus on a strengths perspective (NASW, 2009). This involves a conceptual shift from focusing exclusively on the negative impacts of trauma to the simultaneous positive changes that can occur (e.g., strengthened relationships, improved outlook on life). Given this renewed focus on a strengths perspective, it is important to understand the positive experiences of disaster survivors. Recent research has focused on PTG resulting from a range of traumatic events across different populations (see Calhoun & Tedeschi, 2006). However, little attention has been given to the potential cross-cultural variability in the measurement of PTG. To study cultural or racial variability in predictors or outcomes of PTG, researchers must address the question of whether the measure of PTG has any psychometric similarities across populations. Currently, there is little research on the cross-cultural applicability of these measures.

Methods: Data comes from the Baseline Survey of the Hurricane Katrina Community Advisory Group Study consisting of 270 black and 707 white adults was used (Kessler, P.I., 2006). This is a multi-frame random sample using telephone interviews for data collection. Internal consistency analysis and multi-group confirmatory factor analysis (CFA) via LISREL were used to ascertain the psychometric equivalence between black and white survivor adults including factor pattern, factor loadings, measurement errors, and factor variance and covariance between the two racial groups.

Results: Overall, the results supported the similarities in psychometric properties of this seven item PTG scale between black and white adults. This scale had similar internal consistency reliability between the black and white samples (white sample: Cronbach's alpha=.789, black sample: Cronbach's alpha= .792). In addition, the correlation between individual scale items and their respective overall scale were equally strong in both black and white samples. Within group CFA showed adequate fit according to several goodness of fit indicators for both groups (NNFI=white: .96, black: .95; GFI=white: .97, black: .96; AGFI=white: .95, black: .92). Maximum likelihood factor loadings across groups showed variation in strength of predictors and measurement errors, but differences were small on most items. Tests of measurement invariance hypotheses showed mixed results, but results of RMSEA, NNFI, and CFI indicated that differences in factor pattern, factor loading, error variance, and factor variance and covariance were negligible between groups. The changes in CFI from the nested models also suggested similarity in measurement between the two groups.

Implications: The results of this study warrant further racial comparisons of predictors and outcomes of PTG among black and white adults. Future studies also should examine cross-cultural reliability and validity of this scale among age, gender and other ethnic or racial groups.