Abstract
The factor structure of the Edinburgh Postnatal Depression scale (EPDS) and similar instruments have received little attention in the literature. The researchers set out to investigate the construct validity and reliability of the EPDS amongst impoverished South African women. The EPDS was translated into isiXhosa (using Brislin’s back translation method)and administered by trained interviewers to 147 women in Khayelitsha, South Africa. Responses
were subjected to maximum likelihood confirmatory factor analysis. A single factor
structure was found, consistent with the theory on which the EPDS was based. Internal
consistency was satisfactory (a = 0.89).