Transliteral equivalence and reliability of the Turkish version of the Piers-Harris children's self-concept scale

Abstract

The present study constituted the first step ln the development of a standardized adaptation of a self-concept scale for Turkish students. It was concerned with two issues: (1) to translate and adapt the Piers-Harris Children's Self-Concept Scale into Turkish, and (2) to conduct research on the reliability of the Turkish version of the scale. The study was realized ln two phases. In the initial phase, the English version was translated into Turkish and checked by back-translations. The final translated form was tested for its transliteral equivalence. For this purpose, both the English and the Turkish versions of the ~cale were administered to a sample of 242 (66 boys, 176 girls) bilingual university and secondary school students with approximately two weeks interval. Analysis of variance, !-tests and Pearson Product Moment Correlations generally supported the expectation that the Turkish translation was transliterally equivalent to the English version. In the second phase, research was conducted on the reliabilty of the Turkish Piers-Harris scale with a sample of 447 (247 boys, 200 girls) students from grades four through eight. Test-retest reliability coefficients showed high stability of scores over time intervals ranging from one to seven days. Analysis using the Kuder-Richardson Formula 20 revealed the high homogeneity of the scale. Internal consistency of the items was also investigated using itemtotal (point biserial) correlations which varied greatly (ranging from zero to mediocre correlation) for different items in different sample groups. The varying low item-total correlations were interpreted to indicate the multifactorial nature of the scale. This supported the theoretical conceptualization of self-concept as a multidimensional phenomenon. Further research is needed on this topic.