The Effects of Sampling Model on Inference with Coefficient Alpha

Abstract

The distinction between Type I and Type 12 sampling in connection with measurement data is discussed, and a method is presented for simulating data arising from Type 12 sampling. Next, an empirical study is described in which the perfornance of inferential procedures for coefficient alpha was examined when these procedures were applied with data arising from both kinds of sampling. The results indicate that as long as the data are drawn from a population exhibiting strict essential parallelism among the conditions (an assumption underlying the procedures), the methods maintain precise confidence level control under Type 12 sampling. For data departing from essential parallelism, however, these procedures are not robust under Type 12 sampling. In contrast, actual confidence coefficient values were close to nominal levels under Type I sampling. Implications are noted for practice and for future research on inference with coefficient alpha.