Convergent validity assessment in PLS-SEM: A loadings-driven approach

The article below explains how one can conduct a convergent validity assessment in the context of structural equation modeling via partial least squares (PLS-SEM).

Amora, J. T. (2021). Convergent validity assessment in PLS-SEM: A loadings-driven approach. Data Analysis Perspectives Journal, 2(3), 1-6.

Link to PDF file for this and other DAPJ articles:

https://scriptwarp.com/dapj/#Published_Articles

Abstract:

Assessment of convergent validity of latent variables is one of the steps in conducting structural equation modeling via partial least squares (PLS-SEM). In this paper, we illustrate such an assessment using a loadings-driven approach. The analysis employs WarpPLS, a leading PLS-SEM software tool.

Enjoy!

WarpPLS 7.0 upgraded to stable

Dear colleagues:

Version 7.0 of WarpPLS is now available as a stable version.

Please use this link () to go to the WarpPLS blog post describing this version’s new features.

You can download and install it for a free trial of from warppls.com ().

Enjoy!

WarpPLS 7.0 beta now available: HTMT, new moderating effects features, and more!

Dear colleagues:

Version 7.0 of WarpPLS is now available, as a beta version. You can download and install it for a free trial from:

https://warppls.com

Below is a summary of this version’s new features.

HTMT ratios. The sub-option “'Discriminant validity coefficients (extended set)”, under the menu option “Explore additional coefficients and indices”, now allows you to inspect the heterotrait-monotrait (HTMT) ratios calculated by the software. These ratios have been proposed for discriminant validity assessment, particularly in the context of composite-based SEM via classic PLS algorithms; as opposed to factor-based SEM via modern algorithms that estimate factors (which have been available from this software for quite some time now). Discriminant validity is a measure of the quality of a measurement instrument; the instrument itself is typically a set of question-statements. A measurement instrument has good discriminant validity if the question-statements (or other measures) associated with each latent variable are not confused by the respondents, in terms of their meaning, with the question-statements associated with other latent variables.

Discriminant validity coefficients (extended set). The HTMT ratios are provided along with other coefficients that are useful for discriminant validity assessment, in one single combined set of outputs. These other coefficients are correlations among latent variables and square roots of AVEs, structure loadings and cross-loadings, and full collinearity VIFs. For the HTMT ratios, the following coefficients are also provided: P values, and 90% confidence intervals.

There is a short video that illustrates the discriminant validity coefficients (extended set) ().

Graphical user interface optimization. Several elements of the graphical user interface, such as screens and warning messages, have been optimized so that users can perform SEM analysis tasks with only a few clicks – and in a straightforward fashion. For example, automatic re-analyses are now conducted whenever any of the several SEM analysis settings are changed, with the results becoming immediately available to users. Also, new menu options are now available to facilitate tasks; e.g., users can now open or create a project through the “Open or create project (Step 1)” menu option available under the “Project” menu option. This new “Open or create project (Step 1)” option allows users to open or create a project file, providing an alternative path for executing Step 1.

Fractional splits for 2D moderating effects graphs. Users can now set “fractional” splits for 2D moderating effects graphs, through a new “Split” menu option. The default is 0.5, which splits the sample in the middle when drawing the lines for the effects and the “low” and “high” values of the moderating variable. For instance, if you set the fractional split to 0.1, the software splits the sample in 10% (of the sample) to the left and 90% to the right, respectively for the “low” and “high” values of the moderating variable. This new and powerful fractional split feature enables users to significantly expand their options for illustrating moderating effects in 2D graphs. In the previous version of the software the split was set at 0.5 (i.e., in the middle), with no way to modify it.

There is a short video that illustrates the fractional splits for 2D moderating effects graphs ().

New moderating effects calculation options. Users can now choose among three options for moderating effects calculation: “Two Stages”, “Variable Orthogonalization”, and “Indicator Products”. This is done through the new “View or change moderating effects settings” menu option, under the “Settings” menu option on the software’s main window. The default moderating effects calculation option is “Two Stages”, whereby latent variable scores are calculated first and then used in a second stage for the creation of the interaction variable that implements the moderating effect. The “Variable Orthogonalization” option implements a similar procedure, but stochastically departs from a random variable, which is fully orthogonal to the latent variables in the model, for the creation of the interaction variable that implements the moderating effect. The “Indicator Products” option employs indicator products for the creation of the interaction variable that implements the moderating effect; this was the only option available in the previous version of the software.

There is a short video that illustrates the new moderating effects calculation options ().

Incremental code optimization. This is conducted in each new version of this software. At several points the code was optimized for speed, stability, and coefficient estimation precision. This led to incremental gains in speed even as a number of new features were added. Several of these new features required new and complex calculations, mostly to generate coefficients that were not available before.

Enjoy!

Lateral collinearity and misleading results in variance-based SEM: An illustration and recommendations

A new article discussing methodological issues based on WarpPLS is available. The article is titled “Lateral collinearity and misleading results in variance-based SEM: An illustration and recommendations”. It has been recently published in the Journal of the Association for Information Systems. A full text version of the article is available here as a PDF file. Below is the abstract of the article.

Variance-based structural equation modeling is extensively used in information systems research, and many related findings may have been distorted by hidden collinearity. This is a problem that may extend to multivariate analyses, in general, in the field of information systems as well as in many other fields. In multivariate analyses, collinearity is usually assessed as a predictor-predictor relationship phenomenon, where two or more predictors are checked for redundancy. This type of assessment addresses vertical, or “classic”, collinearity. However, another type of collinearity may also exist, here called “lateral” collinearity. It refers to predictor-criterion collinearity. Lateral collinearity problems are exemplified based on an illustrative variance-based structural equation modeling analysis. The analysis employs WarpPLS 2.0, with the results double-checked with other statistical analysis software tools. It is shown that standard validity and reliability tests do not properly capture lateral collinearity. A new approach for the assessment of both vertical and lateral collinearity in variance-based structural equation modeling is proposed and demonstrated in the context of the illustrative analysis.

Exploring free questionnaire data with anchor variables: An illustration based on a study of IT in healthcare

A new article discussing methodological issues based on WarpPLS is available. The article is titled “Exploring free questionnaire data with anchor variables: An illustration based on a study of IT in healthcare”. It has been recently published in the International Journal of Healthcare Information Systems and Informatics. A full text version of the article is available here as a PDF file. Below is the abstract of the article.

This paper makes an important methodological contribution regarding the use of free questionnaires, illustrated through a study that shows that a healthcare professional’s propensity to use electronic communication technologies creates opportunities for interaction with other professionals, which would not otherwise be possible only via face-to-face interaction. This in turn appears to increase mutual trust, and eventually improve the quality of group outcomes. Free questionnaires are often used by healthcare information management researchers. They yield datasets without clear associations between constructs and related indicators. If such associations exist, they must first be uncovered so that indicators can be grouped within latent variables referring to constructs, and structural equation modeling analyses be conducted. A novel methodological contribution is made here through the proposal of an anchor variable approach to the analysis of free questionnaires. Unlike exploratory factor analyses, the approach relies on the researcher’s semantic knowledge about the variables stemming from a free questionnaire. The use of the approach is demonstrated using the multivariate statistical analysis software WarpPLS 2.0. The study leads to a measurement model that passes comprehensive validity, reliability, and collinearity tests. It also appears to yield practically relevant and meaningful results.