Finite mixture regression is a powerful technique that lets you explore possible unobserved heterogeneity in your regression model. This way you can check whether your regression results are the same for your whole sample or whether there are latent (that is unobserved) subgroups with different regression results. In a way, you could see this as a moderation analysis without a moderator variable.
Link to the R code for this tutorial:
https://www.regorz-statistik.de/en/r_fmr.html
