XLSTAT: Analysis of moderating effects with PLS-PM

 

 

Background

Moderator effects in PLSPM make it possible to study how a variable (qualitative or quantitative) influences the strength of the relationship between two other latent variables. With XLSTAT, these effects can be analyzed using either the direct (product-indicator) approach, which automatically creates interaction variables, or the two-step approach based on latent scores.

Let A, B, and C be three latent variables. We want to understand the moderating effect of B on the relation between A and C. We will have the following model:

 

The latent variables are A, B, C and A*B which is the interaction term. It is used to understand the strength of the moderating effect.

 

Dataset for Example

250 customers of a mobile phone operator were surveyed to answer a large number of questions to model their loyalty to the operator. The PLS-PM model is based on the European Customer Satisfaction Index (ECSI).

This tutorial will show you how to investigate moderating effects in PLS-PM context in XLSTAT.

 

Setting-Up

• Select the expert display mode in the XLSTAT-PLSPM options.

 

• This dialog box appears:

• Select the expert display and save your settings.

• As in the general tutorial, build a model including three variables, Image, Satisfaction, and Loyalty, using the PLSPMGraph sheet and the Path Modeling toolbar.

Once the model has been drawn, two approaches are available to study moderating effects.

 

The product indicator approach

• Add a new latent variable to the model.

• Click on the button for defining manifest variables on the toolbar available in the PLSPMGraph sheet.

In the General tab:

• Select the Interaction option.

In the Interaction tab:

• Select the two variables to be used to create the interaction (image and satisfaction) and the standardized option for treatment of the manifest variables.

 

• Click on the OK button. A new latent variable appears on the diagram with 9 manifest variables being the products of the manifest variables associated to satisfaction and image.

• Click on the run button to run the analysis. The run dialog box appears.

In the General tab:

• Select the option to process standardized manifest variables, weights on standardized VM, and initialize weights by taking the values of the first eigenvector.

• Be sure to select the PLSPM method.

In the Options tab:

• Select the Centroid internal weight estimation scheme and select bootstrap calculation of confidence intervals.

In the Outputs tab:

• Activate latent variable scores with the Standardized option.

• Click on the Ok button.

• The most important result can be found in the path coefficients table.

We see that image and satisfaction have a positive effect on loyalty and that the interaction term has a significant negative effect. That means that the moderating effect of image on the relation between satisfaction and loyalty is significant.

 

The two-stage approach

• If you already have applied the product indicator approach, first delete the interaction latent variable and launch XLSTAT-PLSPM using the run button of the toolbar located on the PLSPMGraph sheet.

• Configure the dialog box as for the product-indicator approach.

We are only interested in the latent variables scores for the simple model. In the PLSPM1 sheet, we compute the product of the latent variables Satisfaction and Image using Microsoft Excel formulas next to the latent variable scores table.

 

• Change the definition of the latent variable, replacing the manifest variables with the latent variable scores that can be selected in the PLSPM1 sheet.

• Add a latent variable with only one manifest variable being the new column.

The new model has the following form:

 

 

• Run the calculations with the same options as in the previous calculations (see above) and study the results.

The path coefficients are the most important results.

 

We see that image and satisfaction have a positive effect on loyalty and that the interaction term has a non-significant negative effect. That means that the moderating effect of image on the relation between satisfaction and loyalty is not significant.