Background: Engineering students commonly learn domain knowledge by engaging with visual representations of it. However, at times they have trouble accessing information from these representations due to the way information is encoded in features of the representation. Purpose: To describe how students engage with representation features, we explored two research questions: (a) what is the interplay between how concepts are encoded within representations, students' use of those concepts, and how students translate between representations during their problem-solving and (b) how is the interplay described in Research Question 1 similar and different across students in statics versus those in digital logic? Design/Method: We synthesized findings from two of our prior research studies using the constant comparative method. We described the effect of representations on students' ability to access and use domain knowledge during problem-solving within and across engineering disciplines. Results: We identified three themes that describe how visual representations affect students' reasoning. First, students conflated concepts that were represented using similar features. Second, students consistently failed to use concepts that were not represented with visually salient features. Third, statics students coordinated multiple representations when translating between representations during problem-solving more often than the digital logic students. Conclusions: These themes provide possible domain-general pathways for redesigning the notation and representations that we use to teach engineering concepts and suggest future avenues of research to further explore the generality of these findings.