The purpose of this study was to further develop and explore the use of an analogical teaching model I call The Pre-Analogy Step (PAS)—A teaching method to improve learners’ familiarity of the important analogical features in a source analog prior to introducing an analogy for understanding a target domain. This study explores a sample of 24 students (13 boys, and 11 girls) and one teacher’s use of the PAS in a combined third- and fourth-grade class as they participate in an instructional unit on plate tectonics. Interactional analysis using video and statistical analysis of a pre- and post-test were used to answer the following research questions: (1) do students' demonstrate an understanding of the source analog's relational structures during the PAS phases of instruction, (2) do students use relational structures of the source analog to reason about target concepts in plate tectonics, and (3) how do students' ideas about plate tectonics change throughout the unit? Students’ understanding of the source analog’s relational structure at a level of simple relations was indicated by their use of both verbal and gestural modes of representation in their expressed models of the behavior of the source analog. The use of higher order relational structure was evident in addition to one case in which a student developed and used an analogical abstraction in order to reason between target models. Overall student performance between pre- and post-test results improved significantly with a 10% gain in mean difference (n = 23, p < 0.05). Improvements on plate tectonic specific items, in which students showed a 15% significant gain in mean difference (n = 23, p < 0.005) suggests that the educational unit had a positive influence on students’ ideas about plate tectonics. Taken together as a whole, these findings suggest PAS may be a viable way to help students become more familiar with the relational structure needed to make inferences about the target models presented in this educational unit. Future research on PAS should take into account classroom variables, such as time and pressure, to allow researchers to gain a more explanatory picture of students expressed models.