Psychometric evaluation of the Spiritual Coping Strategies scale in English, Maltese, back-translation and bilingual versions

Background. The little research that has been done on the topic considers spiritual coping as consisting mainly of religious coping strategies. This limits spiritual coping solely to believers. However, it is argued that spiritual coping should address both believers and non-believers. The development of the new Spiritual Coping Strategies (SCS) scale, which consists of both religious and non-religious coping strategies attempts to fill this research gap. Aim. The aim of this article is to discuss the psychometric assessment of the SCS scale designed in four versions, namely English, Maltese, back-translation and bilingual. Methods. The Likert-type SCS scale comprises 20 items, each represented by the frequency and helpfulness of both religious and non-religious coping strategies. Test-retest of the four versions was conducted between October 1999 and January 2000, with 55 final year nursing students, mean age of 24.5 years, recruited from the same class to facilitate supervision and prevent contamination of results. Test-retest reliability of the scale was investigated by cross-tabulations, Kappa (kappa) measures and Spearman's correlation coefficient. Internal consistency was tested by Cronbach's alpha. The correlation structure was examined by the multivariate techniques of R-type cluster analysis and direct oblimin factor analysis. Results. Cross-tabulations between tests and retests gave highly significant values of chi-squared (P less than or equal to 0.0001) for most items in all versions of the scale. These cross-tabulations also yielded reasonably good values for kappa. Additionally, satisfactory values for Spearman's coefficient of correlation resulted between tests and retests. Twelve frequency variables and 13 helpfulness variables had test-retest correlations in the range 0.5 less than or equal to r (s) less than or equal to 0.79 in at least three of the versions. Optimum values of Cronbach's alpha were observed for the helpfulness variable in retests namely, 0.81 for the English version, 0.73 for the Maltese version, 0.79 for the back-translation, and 0.82 for the bilingual version. These values were all above the acceptable minimum value of 0.7. The correlation structure identified two factors, namely religious and non-religious coping strategies. Although the two factors together explained only between 33% and 40% of the variance in the different versions, the religious factor had good reliability, with 0.77 less than or equal to alpha less than or equal to 0.82 for the four versions, whilst thenon-religious factor also had alpha greater than or equal to 0.71 for all versions except the Maltese one, for which alpha = 0.59. Conclusion. Therefore, psychometric assessment suggests that the SCS scale, in any of the four versions, is a reliable tool which can be used in future studies on SCS in Maltese patients.