Exploring New Models for Seatbelt Use in Survey Data

Problem: Several approaches to analyze seatbelt use have been proposed in the literature. Two methods that has not been explored are the use of unweighted and weighted logistic regression model and the use of item response theory (IRT) or the Rasch model. Since accurate methods to predict seatbelt use behavior based upon observed data must include a built-in design method and model, and overcome computation challenges, weighted and IRT method deem to be other options for an observational survey of seat belt use in the state of Virginia. Method: The observed data from 136 sites within the Commonwealth of Virginia over two years was collected in a two stage systematic stratified proportional to size sampling plan. The data is analyzed using a weighted Rasch model. Results: A relationship between seatbelt use of drivers weighted for county aggregate population size and length of the road segment observed and the factors of vehicle type and gender standardized using a standardized scale is confirmed using logistic regression model selection and AIC analysis. IRT model was considered and was found highly significant. Practical Application: The addition of socio-economic measures, measure of road and driving difficulty, and data from other states may allow the prediction of seatbelt use with a in a new methodology: the models provide tools for policy decision-making.