Embedded Files
It is a measurement model specification, which assumes that the construct is caused by the assigned indicators.
Changes in the indicators directly cause changes in the latent constructs.
Indicators are not expected to be correlated.
When all indicators are high, the construct score will be high. Coversely, when all indicators are low, the construct score will be low.
To assess construct validity of a formative construct; convergent validity, collinearity, and significance of outer weight should be assessed.
step 1: assess Convergent Validity
step 1: assess Convergent Validity
Convergent validity is the extent to which a measure correlates positively with other measures (indicators) of the same construct.
When evaluating formative measurement models, we have to test whether the formatively measured construct is highly correlated with a reflective measure of the same construct. This type of analysis is also known as redundancy analysis (Chin, 1998).
The term redundancy analysis stems from the information in the model being redundant in the sense that it is included in the formative construct and again in the reflective one.
redundancy analysis
redundancy analysis
Technique 1: Relating Formative construct to its reflective constructs
Technique 1: Relating Formative construct to its reflective constructs
Specifically, one has to use the formatively measured construct as an exogenous latent variable predicting an endogenous latent variable operationalized through one or more reflective indicators.
Technique 2: Global Item
Technique 2: Global Item
An alternative is to use a global item that summarizes the essence of the construct the formative indicators purport to measure (Sarstedt et al., 2013).
Example: For the PLS-SEM example on corporate reputation in Chapter 3 (Book by Hair et al., 2017), an additional statement, “Please assess the extent to which [the company] acts in socially conscious ways,” was developed and measured on a scale of 0 (not at all) to 10 (definitely). This question can be used as an endogenous single-item construct to validate the formative measurement of corporate social responsibility (CSOR).
A magnitude of 0.90 or at least 0.80 and above is desired for the relationship between Formative construct and its global item.
step 2: assess Collinearity
step 2: assess Collinearity
Collinearity is high correlations between two formative indicators. Thus, a variable can be explained by the other variables in the analysis.
It can prove problematic from a methodological and interpretational standpoint:
Collinearity boosts the standard errors and thus reduces the ability to demonstrate that the estimated weights are significantly different from zero.
High collinearity can result in the weights being incorrectly estimated, as well as in their signs being reversed.
How to test for collinearity issue?
How to test for collinearity issue?
Compute the tolerance representing the amount of variance of one formative indicator not explained by the other indicators in the same block.
A related measure of collinearity is the variance inflation factor (VIF defied as the reciprocal of the tolerance (i.e., VIF =1/TOL).
Tolerance < 0.20 or VIF > 5 indicates potential collinearity issue (Hair et al., 2011).
Tolerance < 0.30 or VIF > 3.3 indicates potential collinearity issue (Diamantopoulos & Siguaw, 2006)
Remedial Methods
Remedial Methods
Removing one of the corresponding indicator(s). Here, researcher must consider whether or not the remaining indicators still sufficiently capture the construct's content from a theoretical perspective.
Constructing higher-order constructs.
Combining the collinear indicators into a single (new) composite indicator (i.e., an index)-for example, by using their average values, their weighted average value, or their factor scores).
step 3: Assess the Significance & Relevance of Formative Indicators
step 3: Assess the Significance & Relevance of Formative Indicators
Assess the significance of outer weight (relative contribution).
Non-significant indicator weights should not automatically be interpreted as indicative of poor measurement model quality. Rather researchers should also consider a formative indicator's absolute contribution to (or absolute importance for) its construct-that is, the information an indicator provides without considering any other indicators.
The absolute contribution is given by the formative indicator's outer loading, which is always provided along with the indicator weights.
When an indicator's outer weight is non-significant but its outer loading is high (i.e., above 0.50), the indicator should be interpreted as absolutely important but not as relatively important. In this situation, the indicator would generally be retained.
When an indicator has a non-significant weight and the outer loading is below 0.50, the researcher should decide whether to retain or delete the indicator by examining its theoretical relevance and potential content overlap with other indicators of the same construct.
Decision-Making Process for Keeping or Deleting Formative Indicators
Decision-Making Process for Keeping or Deleting Formative Indicators
Summary
Summary
Rules of Thumb for evaluating Formative Measurement Indicators
Rules of Thumb for evaluating Formative Measurement Indicators
Please refer to Gholami et al., (2013) for an example of reporting the formative measurement model.
ExerciseS
ExerciseS
Exercise 1
Exercise 1
Download the dataset here, and assess:
Convergent validity.
Collinearity.
Significance of outer weight.
Otherwise, you may download the project HERE.
Exercise 2
Exercise 2
Download the dataset here, and assess:
Convergent validity.
Collinearity.
Significance of outer weight.
Exercise 3
Exercise 3
By using the data in Excercise 2, assess the following model
Further Reading
Further Reading
Gholami, R., Sulaiman, A. B., Ramayah, T., & Molla, A. (2013). Senior managers’ perception on green information systems (IS) adoption and environmental performance: Results from a field survey. Information & Management, 50(7), 431-438. Click here.
Mikulić, J., & Prebežac, D. (2011). What drives passenger loyalty to traditional and low-cost airlines? A formative partial least squares approach. Journal of Air Transport Management, 17(4), 237-240. Click here.
Coltman, T., Devinney, T. M., Midgley, D. F., & Venaik, S. (2008). Formative versus reflective measurement models: Two applications of formative measurement. Journal of Business Research, 61(12), 1250-1262. Click here.
Nawanir, G., Lim, K. T., Lee, K. L., Okfalisa, Moshood, T. D., & Ahmad, A. N. A. (2020). Less for More: The Structural Effects of Lean Manufacturing Practices on Sustainability of Manufacturing SMEs in Malaysia. International Journal of Supply Chain Management, 9(2), 961-975. Click here.
Cheah, J.-H., Sarstedt, M., Ringle, C. M., Ramayah, T., & Ting, H. (2018). Convergent validity assessment of formatively measured constructs in PLS-SEM. International Journal of Contemporary Hospitality Management, 30(11), 3192-3210. Click here.
Google Sites
Report abuse