Randomized controlled trials (RCTs) are experiments conducted in specific experimental settings by researchers. In educational research, RCT results obtained in experimental settings are often extrapolated/applied to other settings by educational practitioners. RCTs are recognized as a “gold standard” for obtaining reliable causal claims in education. However, RCTs face a serious problem of extrapolating RCT results, i.e., when practitioners extrapolate the results from RCT studies to their local settings, the causal claims sometimes do not hold in the new settings. This problem, according to Nancy Cartwright and her colleagues, can be explained by the input-output logic of RCT studies. This means that RCT studies work in a black box with no explanation of why and how the results are established, and therefore practitioners may blindly apply the results to local contexts. To help address this problem, Cartwright and her colleagues suggest scientists turn their attention beyond RCT studies. They argue that local practitioners can use ethnography to gather local information and use the information to localize the implementation of RCT results to different contexts. I agree that RCT results can be applied more effectively if practitioners apply the results based on their investigations into local environments. To build upon their proposals, I argue that attention should not be moved away immediately from RCT studies because researchers can also help address the problem by recording data on different influencing factors and organizing them into useful causal information. In this way, researchers can reduce the burden on practitioners and provide practitioners with more resources for extrapolating the results.