Path analysis of empirically evaluated model and sample size

Question

I am new to Path Analysis and ESM. I have been doing reading on this method, but still feel very much confused.

I have a empirically evaluated failure model having 45 variables connected with each other in a cause and effect model. In my models, there are total 40 failure paths. The longest paths have as many as 7 variables.

After evaluating the failure models, I have measured each variable through a questionnaire survey and have received a total of 50 replies. Now, I am wondering if this sample size is enough to perform Path Analysis to evaluate the "strongest failure paths"? or to perform SEM if I am attaching 5 latent variables to my model? One of my friends suggested to use PLS techniques by using software 'SmartPLS" as it is less data sensitive.

Answer 1

The required sample size depends on the size of the model. You said you have 5 latent variables, but how many equations do you have? How many indicators are measuring the latent variables? In general: The bigger you model, the larger your sample should be. There are no commonly accepted minimum sample size criteria, only rule of thumbs. As for my opinion I would not perform PLS SEM with a sample size less than 200.

I also reccommend reading chapter 2.3.2. in this paper on page 291 :