I am doing research with secondary data using SEM SMART-PLS, I need references to reinforce that secondary data tested with SMART-PLS requires reliability and validity tests because these tests are generally tested for primary data.
Hi! Maybe this paper could help you: Ringle, C.M./ Sarstedt, M./ Sinkovics, N./ Sinkovics, R.R.: A Perspective on Using Partial Least Squares Structural Equation Modelling in Data Articles, Data in Brief, Volume 48 (2023), p. 109074.
However, you'll not find a specific reference that says that secondary data in pls-sem requires reliability and validity tests, because this is not the objective of this statistic method. For the measurement of the latent constructs that we include in our PLS-SEM model, we can use any type of data, including primary and secondary data. Of course, we have to report the values of the measures for assessing the reliability of the reflective or formative constructs. This is not related with the origins of our data.
That is why your requirement is a little bit strange. But, if something that a reviewer that is not familiar with the method you can cite other studies that have used secondary data for measuring the variables of their proposed PLS-SEM models.
Please keep in mind that reliability and validity are found for a particular population and does not apply in general. If you are applying the results to a new population, the data's reliability and validity would need to be re-established for that population.
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology