Log-Linear Analysis of Frequency Tables Introductory Overview - Automatic Model Fitting

When analyzing four-way or higher tables, finding the best fitting model can become increasingly difficult. Log-Linear Analysis contains automatic model fitting options to facilitate the search for a "good model" that fits the data. The general logic of this algorithm is as follows.

First, STATISTICA will fit a model with no relationships between factors; if that model does not fit (i.e., the respective Chi-square statistic is significant), it will fit a model with all two-way interactions. If that model does not fit either, STATISTICA will fit all three-way interactions, and so on.

Let's assume that this process found the model with all two-way interactions to fit the data. STATISTICA will then proceed to eliminate all two-way interactions that are not statistically significant. The resulting model will be the one that includes the least number of interactions necessary to fit the observed table.