Is There a Reliable Scale for Assessing Attitudes and Preferences among Hispanic and Non-Hispanic Consumers?

Abstract

A great concern for marketing researchers is the development of equivalent and comparable instruments that can be useful in assessing attitudes and preferences across cultures. One important issue in the construction of scales is the determination of the optimum number of response categories that will let us discriminate between rated items and that will give us consistent and reliable responses. National studies related to this issue have indicated that reliability and consistency are independent of the number of scale points (Boole 1981; Matell and Jacoby 1972), but cross-cultural studies indicate that the same scale may have different reliabilities in different countries. In this regard, Parameswaran and Yaprak (1987) have stated that the same research instrument used in a cross-national survey may lead to different levels of response reliabilities among various country samples due to difference in knowledge, perceptions, familiarity with the research instrument and the national propensity for certain response style. The present paper explores this issue by comparing the consistency of responses to different Likert scale formats and different Graphic rating scale formats among Hispanics in the United States, Puerto Ricans and non-Hispanics. The study also examines the effect of level of acculturation on the reliability of responses among Hispanics in the United States.