Rachel Phillips1, Yin Bun Cheung2,3, Simon Lowes Collinson4,5, May-Li Lim4,6, Audrey Ling4,6, Lei Feng4,6, Tze-Pin Ng4,6. 1. a Biostatistics Department , Singapore Clinical Research Institute , Singapore , Singapore. 2. b Centre for Quantitative Medicine, Academia , Duke-National University of Singapore Graduate Medical School , Singapore , Singapore. 3. c Department of International Health , University of Tampere , Tampere , Finland. 4. d Gerontological Research Programme, Faculty of Medicine , National University of Singapore , Singapore , Singapore. 5. e Department of Psychology , National University of Singapore , Singapore , Singapore. 6. f Department of Psychological Medicine , National University of Singapore , Singapore , Singapore.
Abstract
OBJECTIVE: Chinese is the most commonly spoken language in the world. The availability of Chinese translations of assessment scales is useful for research in multi-ethnic and multinational studies. This study aimed to establish whether each of the Chinese translations (Mandarin, Hokkien, Teochew, and Cantonese) of the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) achieved measurement equivalence to the English version. METHOD: Participants included 1856 ethnic Chinese, older adults. The RBANS was administered in the language/dialect according to the participants' preference by interviewers who were fluent in that language/dialect. Multiple regression analysis was used to adjust for demographic and clinical differences between participants who spoke different languages/dialects. Equivalence (practical equivalence) was declared if the 90% confidence interval for the adjusted mean difference fell entirely within the pre-specified equivalence margin, ±.2 (±.4) standard deviations. RESULTS: The delayed memory index was at least practically equivalent across languages. The Mandarin, Hokkien, and Teochew versions of the immediate memory, language, and total scale score were practically equivalent to the English version; the Cantonese version showed small differences from the English version. Equivalence was not established for the Hokkien and Teochew versions of the visuospatial/constructional index. The attention index was different across languages. CONCLUSIONS: Data from the English and Chinese versions for the total scale score, language, delayed, and immediate memory indexes may be pooled for analysis. However, analysis of the attention and visuospatial/constructional indexes from the English and Chinese versions should include a covariate that represents the version in the statistical adjustment.
OBJECTIVE: Chinese is the most commonly spoken language in the world. The availability of Chinese translations of assessment scales is useful for research in multi-ethnic and multinational studies. This study aimed to establish whether each of the Chinese translations (Mandarin, Hokkien, Teochew, and Cantonese) of the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) achieved measurement equivalence to the English version. METHOD:Participants included 1856 ethnic Chinese, older adults. The RBANS was administered in the language/dialect according to the participants' preference by interviewers who were fluent in that language/dialect. Multiple regression analysis was used to adjust for demographic and clinical differences between participants who spoke different languages/dialects. Equivalence (practical equivalence) was declared if the 90% confidence interval for the adjusted mean difference fell entirely within the pre-specified equivalence margin, ±.2 (±.4) standard deviations. RESULTS: The delayed memory index was at least practically equivalent across languages. The Mandarin, Hokkien, and Teochew versions of the immediate memory, language, and total scale score were practically equivalent to the English version; the Cantonese version showed small differences from the English version. Equivalence was not established for the Hokkien and Teochew versions of the visuospatial/constructional index. The attention index was different across languages. CONCLUSIONS: Data from the English and Chinese versions for the total scale score, language, delayed, and immediate memory indexes may be pooled for analysis. However, analysis of the attention and visuospatial/constructional indexes from the English and Chinese versions should include a covariate that represents the version in the statistical adjustment.