Combining Parallel and Exploratory Factor Analysis in Identifying Relationship Scales in Secondary Data.

Abstract

Common methods used in the literature to identify factors within exploratory factor analysis has been shown to be potentially problematic. This brief report illustrates a state of the art approach in identifying factor structure by adding parallel analysis prior to exploratory factor analysis. Parallel analysis enables researchers to have a high degree of confidence of the number of factors to extract prior to exploratory factor analysis. The procedure is illustrated by using items from the National Survey of Families and Households (NSFH) that were used to identify relationship scales.