It is also useful to treat the problem as one of analysis of variance using dummy variable regression methods. Standard analysis of variance provides statistical tests for the presence of some of the effects, provides residuals that can be examined for autocorrelation, and generates an efficient estimator of a that can be compared to that obtained by the variance components method. The primary challenge in performing analysis of variance for the model is the problem's size: estimating the model with sample B amounts to performing a regression with 3533 independent (dummy) variables! By recasting the problem in terms of least-squares, it was possible to develop an algorithm for iteratively improving the fit of a set of coefficients until the least-squares criterion is met. The program, ITROVA, is described in Appendix 2.