Gravitational wave observations can now strongly differentiate between assumptions for how binary compact objects form. Different models for compact binary formation can be ranked by their similarity to GW observations, as a marginal likelihood. In this work, we show how to carefully interpolate this marginal likelihood between model parameters, enabling posterior distributions for these model parameters. Using the StarTrack binary evolution code, we compare one- and three-dimensional models to the compact binary mergers reported in GWTC-1. Consistent with prior work, with our one- dimensional models we infer that modest natal kicks are more consistent with the observed merger rates and mass distributions.