This replication study addresses the methodological challenges impacting the internal validity of research on the Facial Feedback Hypothesis (FFH), which proposes that facial expressions can influence emotional experiences. Recent studies have yielded inconsistent results and unexplained data, prompting an investigation of potential methodological explanations for the variability observed in a recent study by Coles et al. (2022). The current study partially replicated the original study with an added set of questions aimed at further understanding the unexplained data. A total of 67 participants were recruited, all of whom were tested in person, whereas the original study was conducted entirely online. The replication did not yield a statistically significant facial feedback effect and did not indicate a reduction in extreme response patterns relative to the original study. Both quantitative and qualitative measures were employed to assess differences in rating emotions while posing smiling or blank facial expressions. The current findings suggest that there are likely several confounding factors that complicate any interpretation of the original results. Some participants reported that while they were performing the mimicry task, which is expected to induce the facial feedback effect, they had been thinking about how they had felt happy in the past. Without an understanding of the extent or influence of these internal thoughts, these reports raise questions regarding the study’s internal validity. This study contributes to the ongoing dialogue regarding the replicability of the FFH and advocates for more critical thinking and rigorous investigations regarding potential confounds when measuring the effect of facial expressions on emotional experiences.