Year-to-Year Correlations of Mental Tests: A Review of the Hofstaetter Analysis

Hofstaetter's factor analysis of mental test scores from infancy through adolescence is reconsidered. A series of factors, each prominent during 1 age period, is to be expected when data on growth are treated in the conventional manner. The locations of the factors with regard to age give no fundamental information on stages of development, since arbitrary decisions regarding the beginning and ending of the series produce radical shifts in the ages where new factors become prominent. The Thurstone concept of simple structure does not provide an appropriate model for explaining longitudinal changes in performance.