Identifying contingencies between constructs in a multi-faceted learning progression (LP) is a challenging task. Often, there is not enough evidence in the literature to support connections, and once identified, they are difficult to empirically test. Here, we use causal model search to evaluate how connections between ideas in a genetics LP change over time in the context of an introductory biology course. We identify primary and secondary "hub" ideas and connections between concepts before and after instruction to illustrate how students moved from a phenotypic grounding of genetics knowledge to a more genotypic grounding of their genetics knowledge after instruction. We discuss our results in light of conceptual change and illustrate the importance of understanding students' idea structures within a domain.