Aim: This paper is a report of an assessment of the construct validity of the Chinese version of the Edinburgh Feeding Evaluation in Dementia scale.
Background: The Edinburgh Feeding Evaluation in Dementia scale has previously been translated and back-translated and to determine the equality of the Chinese and English versions. However, the construct validity of the Chinese version has not been investigated.
Methods: Participants (n = 477) were selected from residents with dementia in licensed long-term care facilities in Taiwan in 2006-2007. Data collectors received training before observing feeding for 2 days during lunch and dinner. Data were analysed using principal component analysis, Mokken scaling and correlation.
Findings: A two-factor structure was demonstrated for the 11 items of the Chinese version of the scale and six items conformed to a Mokken scale. The Chinese version of the scale correlated with external constructs such as weight and body mass index, as predicted by the unmet needs model.
Conclusion: A confirmatory factor analysis is needed to confirm that the Chinese version measures the same qualities, in the same proportions, and with the same factor structure as the original Edinburgh Feeding Evaluation in Dementia scale.