Anecdotal evidence is regarded highly by actors in the educational sector. While sometimes informative, misinterpretation of its evidential value can further the spread of misinformation and lead to undue conclusions. Therefore, it is important to understand how laypeople assess anecdotal compared to empirical evidence. We hypothesized that texts containing empirical evidence are rated as more credible and more appropriate for supporting action- and policy-oriented measures than texts containing anecdotal evidence. We also assumed that this advantage of empirical over anecdotal evidence increases with more sophisticated epistemological beliefs of participants. We examined these hypotheses in two preregistered experiments (N1 = 427, N2 = 473). Although a manipulation check showed that participants correctly recognized the type of evidence presented, our hypotheses were not supported. Neither effects of evidence on message credibility, perceived appropriate support for resulting measures, nor a moderation effect of epistemological beliefs were registered. However, there was a significant main effect of epistemological beliefs on the dependent variables. The more sophisticated people’s epistemological beliefs, the lower they rated message credibility and perceived appropriate support—independent of the type of evidence provided. Although participants with more sophisticated epistemological beliefs appear to rate evidence in science communication about educational insights more critically, they did not regard the type of evidence differently. This non-discrimination of evidential value between anecdotal and empirical evidence is critical for decision processes, which should not be based on weak anecdotal evidence. This makes the need for interventions promoting evidence-based thinking and action even more apparent.