The Dirty Laundry of ESL Survey Research

Abstract

For the novice researcher, the processes of design and implementation of a normed survey instrument for an ESL population hold many opportunities for error, and so for misleading reporting of results. This article traces the formulation and norming of one such questionnaire, constructed specifically for the self-reporting of ESL students on their preferred perceptual learning styles. This retrospective report provides description and comments on the problems of reliability and validity in survey design. It is presented here so that beginning researchers may gain some insight into the actual life of a research project, and to assist others working to design survey instruments for an ESL population. Several years ago, I encountered almost by accident the CITE (Center for Innovative Teaching Experiences) Learning Styles Inventory (Babich, Burdine, Allbright, & Randol, 1975), a selfreporting instrument used by Kansas public schools to assist students-native speakers (NSs) of English-in identifying their preferred perceptual learning styles. It was a simply stated survey, normed on elementary and secondary NSs. My knowledge of learning styles was limited, but the idea of an instrument for such measuring, and for self-discovery, interested me. During the next several months, I investigated the available literature about learning styles; I found several additional survey instruments, and a wealth of information.