Open AccessAn introduction to the partial least squares approach to structural equation modelling: a method for exploratory psychiatric research
Author(s) -
Riou Julien,
Guyon Hervé,
Falissard Bruno
Publication year - 2016
Publication title -
international journal of methods in psychiatric research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.275
H-Index - 73
eISSN - 1557-0657
pISSN - 1049-8931
DOI - 10.1002/mpr.1497
Subject(s) - structural equation modeling , partial least squares regression , psychology , sample size determination , exploratory research , population , sample (material) , clinical psychology , applied psychology , psychiatry , mathematics , statistics , medicine , social science , sociology , environmental health , chemistry , chromatography
In psychiatry and psychology, relationship patterns connecting disorders and risk factors are always complex and intricate. Advanced statistical methods have been developed to overcome this issue, the most common being structural equation modelling (SEM). The main approach to SEM (CB‐SEM for covariance‐based SEM) has been widely used by psychiatry and psychology researchers to test whether a comprehensive theoretical model is compatible with observed data. While the validity of this approach method has been demonstrated, its application is limited in some situations, such as early‐stage exploratory studies using small sample sizes. The partial least squares approach to SEM (PLS‐SEM) has risen in many scientific fields as an alternative method that is especially useful when sample size restricts the use of CB‐SEM. In this article, we aim to provide a comprehensive introduction to PLS‐SEM intended to CB‐SEM users in psychiatric and psychological fields, with an illustration using data on suicidality among prisoners. Researchers in these fields could benefit from PLS‐SEM, a promising exploratory technique well adapted to studies on infrequent diseases or specific population subsets. Copyright © 2015 John Wiley & Sons, Ltd.