Using Item Response Theory to Understand Item-Nonresponse (Missing Data) in Ghanaian Surveys

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
After reviewing the theoretical and empirical literature on Item Response Theory (IRT) and Item-Nonresponse, this study investigates three issues: Firstly, to identify the most appropriate IRT model for understanding item-nonresponse. Secondly, to find out the reason behind ''don't know'' responses and missing data; whether respondents don‟t really know, don‟t care, or don‟t want to tell. Finally, to find out the characteristics of nonrespondents in Ghanaian surveys. Secondary analyses were done on questionnaire data collected in the 5th wave of the world values survey. All items were dichotomously scored. 0 was assigned to missing or „don‟t know‟ responses, and 1 was assigned to answered items. The data was analysed based on four item response theory models namely, the constrained Rasch model, the unconstrained Rasch model, the two parameter logistic model, and the three parameter logistic model. These models were explored to determine the most appropriate model for the data. The unconstrained Rasch model appeared as the best model for understanding item-nonresponse. It was found that, giving a „don‟t know‟ answer to an item is because of the item‟s difficulty, which means that respondents don‟t really know the answer to the item. The results also revealed that, item-nonresponse can be predicted by some item and respondent characteristics. In a typical application, politics and income related questions recorded the highest item-nonresponse rates. Female respondents and respondents with no formal education also recorded very high item-nonresponse rates.
Thesis submitted to the Department of Mathematics, Kwame Nkrumah University of Science and Technology in Partial fulfillment of the requirements for the award of Master of Philosophy Degree in Applied Mathematics.
Item-Nonresponse,, Item Response Theory, Respondent Characteristics,, Unconstrained Rasch Model, World Values Survey