Conventional versus confirmatory factor analysis : Methods for validating the structure of existing scales
Paul C. Burnett, Barry C. Dart · QUT ePrints (Queensland University of Technology) · 1997
Educational researchers are often required to validate measurement instruments on their research samples because they differ in characteristics from the original norming sample. Two factor analytic approaches can be used to evaluate an instrument's construct validity, namely, Conventional or Exploratory Factor Analysis and Confirmatory Factor Analysis. In an endeavour to investigate the degree of variation in factor structures produced using different factor analytic techniques two large data sets were analysed using four factor analytic approaches. The four methods used were Theoretical Conventional, Theoretical Conventional followed by Confirmatory, Modified Confirmatory, and Pure Confirmatory. The variation in the factor structures produced by each of the factor analytic approaches was found to be small. Recommendations for the use of factor analytic methods as a scale validation tool for existing instruments are presented