Is My Model Right? Model Quality and Model Misspecification in PLS -- Recommendations for IS Research

68 Pages Posted: 10 Oct 2013

See all articles by Joerg Evermann

Joerg Evermann

Memorial University of Newfoundland

Date Written: October 8, 2013

Abstract

Partial Least Squares (PLS) is a statistical technique that is widely used in the Information Systems discipline to estimate statistical models with structural equations and latent variables. While PLS does not provide a statistical test of model fit to data, PLS users apply a set of heuristics to evaluate the quality of estimated models. In this paper, we investigate to what extent these heuristics are able to identify misspecified models and may be used in place of a statistical test of model fit. The results of our analysis and simulation study indicate that existing heuristics are unable to reliably detect typical model misspecifications. We discuss the implications of this result for theory testing, prediction, and exploratory analysis in Information Systems research.

Keywords: partial least squares, misspecification, simulation study, structural equation modeling, data analysis, research methods

Suggested Citation

Evermann, Joerg, Is My Model Right? Model Quality and Model Misspecification in PLS -- Recommendations for IS Research (October 8, 2013). Available at SSRN: https://ssrn.com/abstract=2337699 or http://dx.doi.org/10.2139/ssrn.2337699

Joerg Evermann (Contact Author)

Memorial University of Newfoundland ( email )

St. John's, Newfoundland A1B 3X5
Canada

Register to save articles to
your library

Register

Paper statistics

Downloads
59
rank
340,117
Abstract Views
272
PlumX Metrics