Mathematics is a discipline that exists within a social context, a context that influences how the subject is taught and learned. Women's performance in mathematics has been shown to be vulnerable to stereotype threat, and psychologists have developed interventions aimed at reducing the effect of stereotype threat on women's mathematics performance. We applied interventions, shown to be effective at reducing gender stereotype threat in mathematics in a lab setting, in undergraduate calculus classrooms. We found that these interventions had no detectable effect in the classroom setting, leading to two important cautions about the application of lab-based studies in authentic settings. First, lab settings differ from classroom settings in fundamental ways, limiting the ability to generalize inferences from lab studies to classroom settings, a caution that is more well-known amongst researchers seeking to replicate studies than it is amongst practitioners seeking to apply studies in practice. Second, when seeking to apply lab-based studies in classroom settings, researchers must be aware that relying on research constructs (such as sample size and significance) masks the practical measure of a meaningful sample size, which is roughly the number of students in a particular instructor's course at a given time.