In Miyamoto et al. (2015, this issue) the authors looked to substantiate the presence of the spacing effect, referenced from the psychology literature, in several MOOCs. Their secondary analyses constituted a robust, empirical finding on the correspondence between session distribution and certification but with only a coarse, analogous relationship to the theory of distributed practice. This article underscores the difficulty of validating theory and inferring causation from data produced in the wild but also the resultant rigour necessitated by this difficulty. [For "Beyond Time-on-Task: The Relationship between Spaced Study and Certification in MOOCs," see EJ1126940.]