Examining Measurement Invariance With Multilevel Behavioral Rasch Models: Does Gender Affect International Food Insecurity Measures?

• In Event: 43.061 - Applications of Rasch Models in Educational Assessments

Abstract

Food security is an essential component of human well-being and quality of life around the world. The Food and Agriculture Organization (FAO) of United Nations developed the Food Insecurity Experience Scale to quantitatively measure the severity of food insecurity at the level of individual experiences, and it is collected through Gallup World Poll. A multilevel behavioral Rasch model is proposed to examine differential item functioning (DIF) related to the gender of each respondent across different countries. The effects of several covariates are further investigated on latent measures of food insecurity and DIF. Preliminary analyses (US and Brazil) indicated DIF across gender groups in the U.S. data but not in Brazil. Additional analyses in the final paper will include more countries.

Authors

• Jue Wang, University of Science and Technology of China

• Victoria Tanaka, University of Georgia - Athens

• George Engelhard, University of Georgia

• Matthew Rabbitt, The United States Department of Agriculture Economic Research Service