Maximum likelihood analysis of repeated measures.

Biased tests and non-exact F distributions are deficiencies in the classica l ANOVA approach to repeated measures in the case of non-spherical variance covariance matrices or unbalanced data. Individual differences in developm ental patterns have also been typically ignored by being shoved into the er ror term. This paper discusses two case studies that illustrate analysis op portunities in the solution of these problems by using maximum likelihood m ethods. Evidence suggests that the latter allow for more precise inference by using more accurate models of the covariance matrix. They also provide e stimators with known and favorable asymptotic properties in the case of unb alanced data. Individual differences in growth can be quantified by inclusi on of new covariance parameters, which provide some answers to Barlow and H ensen critical remarks on classical experimental psychology. Some potential problems in the use of these methods and characteristics of available soft ware are finally discussed.