[Method of simulation and choice of factors in the analysis of principal components]

There are many methods to determine how many components should be retained in principal components analysis. This choice can be made on the basis of arbitrary (Kaiser) or subjective (Interpretable factors) criteria. This work presents the simulation criteria of Lébart e Dreyfus. The method create a matrix of randomized numbers and a principal component analysis is performed on the basis of this matrix. The components extracted from this data represent the cut off values. Those that exceed this cut off value should be retained. As an example, a principal component analysis is performed with the Hamilton depression rating scale (17 items) on a sample of 130 subjects.

The Simulation method is compared with the Kaiser method and is shown that the Simulation method maintains the components clinically significant.