An objective counterpart to the visual scree test for factor analysis: The standard error scree

Abstract

Many methods have been suggested for determining the number of factors in factor analysis solutions, most of them trying to match visual solutions. Presented here is a linear regression approach providing objectivity while producing the same results as the visual scree test. The standard error scree is based on the standard error of estimate for a set of points in the plot of eigenvalues. The number of factors to retain corresponds to the point where the standard error exceeds 1/m, where m is the numbers of variables in the analysis. Experience with the method to date indicates that the results are consistent with visual solutions.