Examining the psychometric properties of comprehensive knowledge about HIV/AIDS prevention scale in the Uganda demographic and health survey tool

Abstract

Background: Limited information publicly exists on the functioning of the items intended to measure comprehensive knowledge about HIV/AIDS in the Uganda Demographic and Health Survey (UDHS) tool. The five items in the scale include: Can people reduce their chance of getting the AIDS virus by using a condom every time they have sex? Can people reduce their chance of getting the AIDS virus by having just one uninfected sex partner who has no other sex partners? Can people get the AIDS virus from mosquito bites? Can people get the AIDS virus by sharing food with a person who has AIDS? Is it possible for a healthy-looking person to have the AIDS virus? Materials and Methods: This study therefore intended to examine the psychometric properties of those five items. The study was based on cross-sectional UDHS data of 2011 for both men and women where by the analyses involved: use of confirmatory factor analysis in the framework of structural equation modelling to assess how the items are associated with the construct; use of a two-parameter logistic item response theory model to establish the items’ difficulty and discrimination parameters; and use of a hybrid logistic regression model to examine differential functioning of items between men and women subgroups. Results: The results show that all the five items were significantly associated with the construct of comprehensive knowledge about HIV/AIDS prevention but do not adequately account for the scale. The difficulty parameters depicted the items to be easy and thus favouring only low ability respondents while the discrimination parameters depicted the items to be good at discriminating between high and low ability respondents. On the other hand, only 1 out of 5 items that is to say: Can people get the AIDS virus from mosquito bites?, functioned favourably between men and women groups. The rest of the items exhibited situations where they favoured one group over the other, when persons’ abilities are matched. Conclusion: The results of this study therefore indicate a need to include more items in the UDHS tool to adequately account for the construct of comprehensive knowledge about HIV/AIDS prevention. Furthermore, difficult items also need to be included in the scale so as to be able to cater for all levels of the respondents. Lastly, items that function favourably between men and women groups also need to be included in the scale to allow for a credible comparison between these groups.