PLS-SEM with SmartPLS 4: From Zero to Hero

STATISTICAL REMEDY

Marker Variable

Marker variable is the variable that is not theoretically related with any of endogenous variables under study.

Marker variable MUST be highly reliable.

In the testing, the marker variable must be pointing into all endogenous variables (dependent and mediating variables).

Step 1

Relate all variables involved in the study, except marker variable.

Look at the R square.

Step 2

Introduce the marker variable into the model.

Relate marker variables with all endogenous variable.

Look at the R square.

The relationship between the marker variables and the endogenous variables MUST NOT be significant. In other words, the R square must not significantly increase.

If the relationships are significant and the R square increases by 10%, then CMV exists in the model.

Examples of Marker Variables

Social Desirability Scale (Fischer & Fick, 1993)

I like to gossip at times.
There have been occasions where I took advantage of someone.
I'm always willing to admit it when I made a mistake.
I sometimes try to get even rather than forgive and forget.
At times I have really insisted on having things my own way.
I have never been irked when people expressed ideas very different from my own.
I have never deliberately said something that hurt someone's feeling

Cognitive Rigidity (Oreg, 2003)

Once I’ve come to a conclusion, I’m not likely to change my mind.
I don’t change my mind easily.
My views are very consistent over time.

Attitudes towards the Color Blue (Miller & Simmering, 2008)

I prefer blue to other colors.
I like the color blue.
I like blue clothes.
I hope my next car is blue

Attitudes towards Private Label Brands (Burton, Lichtenstein, Netemeyer, & Garretson, 1998)

Buying private label brands makes me feel good.
I love it when private label brands are available for the product categories I purchase.
For most product categories, the best buy is usually the private label brand.
Considering value for the money, I prefer private label brands to national brands.
When I buy a private label brand, I always feel that I am getting a good idea.

Full Collinearity Testing

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 doublechecked 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 variancebased structural equation modeling is proposed and demonstrated in the context of the illustrative analysis.

Step 1: Create a dummy variable.

Step 2: Assign the dummy variable as an endogenous variables. Relate all variables of study with the dummy variable.

Step 3: Look at the collinearity diagnostic. If the VIF values are greater than 3.3 (Diamantopoulos & Siguaw, 2006) or 5 (Hair et al., 2011), CMV exists in the model.

Exercise

Exercise 1

Please download the data HERE.

Step 1

Step 2

Exercise 2

Please use same data with Exercise 1, do the full multicollinearity testing

Reference Video

Assessing CMV/CMB using Full Collinearity Testing (Random Dependent Variable)

Google Sites

Report abuse