I am currently self-learning Structural Equation Modeling. I have studied some theory and also a bit about the software implementation, but not enough to practically conduct an SEM study.

I wanted to have a more detailed reading of the full practical implementation of Structural Equation Modeling.

For example, I have a theoretical model hypothesised for a phenomenon (with the latent and respective observed variables and their relations and the directions). How do I go next. What should be the sample size (10/observed variable?), how to practically make sure that I have chosen the right observable variables (which software, any rules of thumb that I must remember etc.), testing reliability and validity, any research design subtleties I need to take care of that might not be obvious from the theory, and then deriving the covariance matrix and hypothesis testing, what should be the practical steps to ensure their validity etc.

Any lecture/ video series, books, notes etc. or a list of resources that are comprehensive enough to be self-sufficient for a basic SEM study would be very helpful, especially those that are useful for self study.

  • $\begingroup$ Could you pleas state what sort of structural modeling you are interested in? $\endgroup$ – 1muflon1 Dec 28 '20 at 18:20
  • $\begingroup$ I was actually asking for a broad guide for SEMs, but I understand that might be too broad. Given the research question I currently have in mind, I guess I am thinking of Partial Least Squares Path modeling, because the question is a cause and effect one, but I might be wrong as a I am only newly exposed to SEM $\endgroup$ – Ishan Kashyap Hazarika Dec 28 '20 at 18:49
  • $\begingroup$ but SEM will be done differently in macro, micro, marketing etc. this is not like OLS, SEM is an umbrella term for various models some SEM models that you will find in growth might not be found in any other field of economics. I don’t think there is a general guide that would cover every field of economics. $\endgroup$ – 1muflon1 Dec 28 '20 at 19:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.