In a data-driven world, the ability to critically assess empirical data is a core aspect of scientific literacy and a major goal of physics education. Yet, students often fall back on heuristics that conflict with statistical reasoning. A common example is the “the more measurements, the better the quality of the data” (MMB) heuristic, which naively equates quantity with quality, ignoring variance. According to conceptual change theory, such heuristics or misconceptions coexist with formal scientific knowledge and are automatically triggered during reasoning, resulting in conflicts and requiring active suppression. Inhibitory control, an executive function that overrides automatic responses, may be the central key to resolving these conflicts; however, its role in procedural domains, such as data literacy, is understudied. This study examines whether and how inhibitory control facilitates students to overcome the MMB heuristic in different representational formats (tabular vs pictorial). Using a modified Stroop paradigm, undergraduate physics students (N=30) evaluated data quality presented in tabular (numerical tables) and pictorial (dot plots) formats across the congruent, incongruent, and neutral stimulus types. Results showed significant accuracy reductions in incongruent trials compared to both neutral and congruent trials across both formats, indicating significant Stroop effects. For the response times, we found a significant Stroop effect (longer response time in the incongruent than congruent trials) in the pictorial format, while a similar but non-significant trend in the tabular format. These findings first indicated the involvement of inhibitory control when the heuristic activation occurred in both formats. However, the pictorial format may evoke stronger or more persistent heuristic conflict, or the potential higher cognitive processing demands in the tabular format may mask the subtle RT-based Stroop effects. These findings extend the conceptual change research by identifying inhibitory control as crucial for resolving heuristic interference in procedural reasoning, and they stress the need to consider how representational formats modulate both heuristic activation and cognitive processing demands when designing instruction in physics.