Establishing Factor Validity Using Variable Reduction in Confirmatory Factor Analysis

Using a 21-statement attitude-type instrument, an iterative procedure for improving confirmatory model fit is demonstrated within the context of Bentler's EQS and maximum likelihood factor analysis. Each iteration systematically eliminates the poorest fitting statement as identified by a variable fit index. The statements were hypothesized to define a three-factor independent cluster solution, one factor being associated with each of three a priori scales. Initially, the statements did not define the subscales as hypothesized. Through a series of 11 iterations, 11 variables were eliminated. The resulting three-factor independent cluster model fit the data extremely well.