Robustness of PLS for estimating latent structure models

Commission of the European Communities

Latent variable structural models and the PLS (partial least squares) estimation procedure have found increased interest since they have been used in the context of customer satisfaction measurement. The well-known property that the estimates of the inner structure model are inconsistent implies biased estimates for finite sample sizes. A simplified version of the structural model that is used for the SCSI system has been used to generate simulated data and to study the PLS algorithm in the presence of three inadequacies: (i) skew instead of symmetric distributions for manifest variables, (ii) multicollinearity within blocks of manifest and between latent variables, and (iii) misspecification of the structural model (omission of regressors). The simulation results show that the PLS method is quite robust against the mentioned inadequacies. The bias that is caused by the inconsistency of PLS estimates is substantially increased only for extremely skewed distributions and for the erroneous omission of a highly relevant latent regressor variable. The estimated scores of the latent variables are always in a very good agreement with the true values and seem to be unaffected by the inadequacies under investigation.