Allfactor - Exploring Factors in Large Datasets
Application software download free Windows: Allfactor is a powerful, fast and versatile software tool for exploring factors in large datasets. It offers a number of advanced features to help users better understand and interpret their data and factor solutions, as well as to make informed decisions about how best to proceed with their analyses.
A common source of difficulty when using tetrachoric/polychoric correlations for exploratory factor analysis is that the correlation matrix might turn out to be not positive definite (i.e., it has one or more negative eigenvalues). The new release of FACTOR includes two methods to correct this problem: one uses a smoothing procedure that attempts to change the relative weight of diagonal elements of the correlation matrix in order to correct the not positive definiteness; the other is based on Pratt’s importance measures. Both methods are available in the “Other specifications of factor models” menu.
It is not uncommon for a researcher to analyze a sample that includes several different groups of participants, such as women and men. The new release of FACTOR allows users to manage these groups when doing an exploratory factor analysis.
The new release of FACTOR supports an exploratory bifactor model based on PEBI. This method allows researchers to split the original sample into a set of randomly selected subsamples and to run a fully exploratory factor analysis in each of these subsamples with a group-specific rotation criterion. The results of these analyses can then be used to construct a bifactor model for the whole sample.