How to optimize statistical approach in terms of reducing number of statistical tests used?

by st4co4   Last Updated July 12, 2019 09:19 AM - source

I need to analyse second step care among very heterogenic patient population whose second step care is also very diverse. I can do this using multiple models and tests (up to 10), but this is just too much for an scientific article. Nobody would accept this for publishing. I need to optimize my approach.

I made a scheme summary of my data with possible statistical solutions:

What would be the most optimal solution to analyse such data?

Related Questions

regression models and dummy variables

Updated July 27, 2019 11:19 AM

predict function and categorical variables in R

Updated June 11, 2015 03:08 AM

Fitting Com- Poisson and obtaining expected frequencies

Updated September 10, 2018 15:19 PM