Adapting the emotional dysregulation questionnaire for South Africa : methods of adaptation and psychometric properties of the South African emotional dysregulation scale

Abstract

PURPOSE : Emotional dysregulation is an important aspect of psychopathology, especially borderline personality disorder, and is prevalent in populations exposed to multiple traumas, like South Africa. We describe adaptation of the emotional dysregulation questionnaire (EDQ) for South Africa and its psychometric performance. METHODS : We enrolled 58 ethnically-diverse women, who had experienced severe intimate partner violence, in eight group discussions held in three Provinces. They were asked what they did when they were upset, and related questions. During the groups, cognitive interviews focused on the EDQ. We adapted the EDQ using these findings, removing very similar or poorly understood items, and drafted new items for matters raised by the women. The resulting 23 items were tested with 236 women in a pilot study. RESULTS : Women described a range of behaviours when they were upset that indicated loss of control such as drinking heavily, being abusive, breaking things and beating their children. In the pilot, two items were highly skewed and three, measuring restricted emotional expression, were also poorly correlated with the rest of the scale and performance was poorer with them included. The remaining 18 items performed well with Cronbach's alpha = 0.862, acceptable item-rest correlations, and KMO = 0.865 (all items were above 0.80). On confirmatory factor analysis, the 18-item measure had the best fit. The measure was positively associated with depression, borderline personality disorder, PTSD and complex PTSD. CONCLUSION : The adapted emotional dysregulation questionnaire had robust psychometric properties and is positioned to be a useful scale with traumatised populations in South Africa.

HIGHLIGHTS • The population of South Africa is exposed to multiple traumas and emotional dysregulation is common. • We asked trauma-exposed women to explore what they did when they were upset and their views on a standard measure. • We used information provided by South African trauma-exposed women to adapt and test the measure. • Our measure performed well in tests and was correlated with mental health problems in the way we expected. • We concluded that an 18-item measure of emotional dysregulation can be useful for mental health research in South Africa.