I am analysing paired choice-based conjoint data (full sample of possible profiles shown plenty of times). Following the Hainmueller et al. (2014) and Leeper et al (2020) approach I estimated with the cregg package in R AMCEs based on LPM and then derived descriptive marginal means from the AMCEs (https://thomasleeper.com/cregg/).
The conjoint consists of 8 attributes each having 2 levels (logic: high / low). My dependent variable is dummy choice (1/0).
My questions is:
I want to compare the combined effect of attributes 1-5 (set 1) to the effect of attributes 6-8 (set 2) to know which of the effects explains choice better? Which test can I apply here? how do I do it in R?