“…However, we estimate the population average marginal conditional effect (pAMCE) proposed in de la Cuesta, Egami, and Imai (2022) using the theoretical -rather than empirical -target distributions Specifically, Figure 7 9 Though, one could easily adapt the definitions of AMCE and pAMCE to allow for more general distributions by replacing probability function, P r[t ijk,−l , t i,−j,k ] to a density function and replacing the summation over profiles to an integral. 10 It is not terribly surprising that controlling for background factors randomized independently of each other does not change the findings. We thank the author for pointing this out.…”