ABSTRACT
Wiley and Wiley have made a contribution to the literature on dealing with errors of measurement by showing how to build a model employing the assumption of homogeneity of error variance in panel data. Generalizing to multiple-wave panel studies, when the assumptions of the Wiley and Wiley structural model are given, error variances, true-score variances, and un-standardized regression weights between corresponding true scores are identified for all but the first and last measures. Given at least four-wave data, suggestive, but not conclusive , evidence about which assumption is better may be obtained by comparing the estimated error variances and reliabilities for the inner measures. The chi-square fit before imposing a restriction can be compared to the chi-square fit for the more restricted model as a measure of the tenability of that restriction.
