The marker index: A new method of selection of marker variables in factor analysis

A new method for selecting marker variables in factor analysis is introduced, called the Marker Index. The Marker Index evaluates the usefulness of a variable in representing a factor by weighting both primary and secondary loadings, and by maximizing the similarity between the marker and the underlying factor. The method is compared with other four methods of selections, based on factor loadings or factor weights. Theoretical and empirical comparisons of the performance of the different methods are performed and the advantages of the Marker Index are illustrated. The comparisons demonstrate that the use of the Marker Index as a selection method can improve factorial simplicity, scale validity, and reliability of the composite scales composed by variables selected as markers. © 2007 Cises.