This report focuses on the ways in which 36 Grade 4 students recognized, explained, described, and employed variation during interviews conducted 1 month after participating in STEM-based activities in which they tested, adjusted, and re-tested catapults. An inductive thematic methodology was used for analysis of the interview transcripts to capture the ways in which students discussed their analyses and justified their conclusions from the activity. The results were based on 1080 instances of variation in student responses to the interview questions, which evidenced three ways students characterized variation: contextual variation, specific variation, and general variation. Findings point to the essential nature of context in building statistical understanding in relation to both specific and general aspects of variation as well as decision-making in that context.