An alternative approach to cross-cultural function assessment

Background This paper describes our work to create and validate a new method for cross-cultural and sex-specific function assessment that avoids the major problems with existing methods. Methods We used free listing to learn about tasks important to local people. Community-specific function questionnaires based on these tasks were then created and used in community-based surveys. The survey results were used to assess the questionnaires' internal reliability (Cronbach's alpha), combined test-retest and across-interviewer reliability using repeat interviews, and (in Uganda) criterion validity by comparing assessment by self to assessments by cohabiting adults. Results Field trials of this approach were conducted in rural Rwanda and Uganda. Differences between tasks identified by free listing were greater between sexes than sites. Cronbach's alphas for male and female questionnaires were respectively 0.815 and 0.822 in Rwanda and 0.886 and 0.881 in Uganda. Pearson correlations for combined test-retest and across-interviewer reliability were respectively 0.469 and 0.640 for Rwandan men and women and 0.797 and 0.871 in Uganda. Correlation between self-assessment and cohabiting adults was 0.904. Conclusions We have developed an alternative to the existing approach of adapting western function instruments to other cultures and situations. The field trials have demonstrated that this approach is rapid, feasible and can yield valid and reliable instruments. Developing instruments locally avoids the problems of limited local relevance and appropriateness associated with adapting western instruments. Although each instrument created in this way is culturally bound, they are 'cross-cultural' in the sense that each refers to the tasks most important to local people. This approach should prove useful for both researchers and aid agencies working in non-western countries.