Elizabeth Lee, PhD
“Achieving Semantic Equivalence Between the Chinese and English Versions of the Postpartum Depression Screening Scale”
Postpartum depression is a major depressive mood disorder that occurs within four weeks after delivery and can last as long as a year. The prevalence of postpartum depression is 10% to 20% of all postnatal women in the United States. When postpartum depression is undetected and untreated, adverse outcomes affect not only the postpartum woman but also other family members. For Chinese immigrant postpartum women in the United States language is a barrier to detecting postpartum depression and accessing care. The purpose of the study was to translate the English version of the Postpartum Depression Screening Scale (PDSS) into Chinese, and to achieve and assess for semantic equivalence using Brislin’s multi-translation method in cross-cultural research. The application of Brislin’s multi-translation method included translation, back translation, the use of a bilingual panel of experts, a committee of bilingual Chinese women, pilot testing the Chinese version of the Postpartum Depression Screening Scale (PDSS) in six Chinese postpartum mothers, and administering the instrument to a sample population of Chinese postpartum women (N = 31). Data analysis and psychometric testing were used to assess semantic equivalence. The paired t-test findings showed no significant differences between the Chinese and English PDSS versions of the total scores M = 56.65, SD = 19.94 and M = 56.29, SD = 18.36 respectively, t (30) = .23, p = .82. The total score and dimension-level scores of the Chinese and English PDSS versions were significantly correlated, (r = .75 to .91, p = .001). The coefficient of determination (r2) indicated that the Chinese and English versions shared 56% to 83% of the variance. The Cronbach’s alpha of reliability for the Chinese PDSS version was .97 and for the English PDSS version was .96. The kappa statistic was large (Κ = .728, p = .001). The Chinese and English PDSS versions appear to be semantically equivalent.