The author of this commentary argues that physical scientists are attempting to advance knowledge in the so-called hard sciences, whereas education researchers are laboring to increase knowledge and understanding in an "extremely hard" but softer domain. Drawing on the work of Popper and Dewey, this commentary highlights the relative similarities between hard sciences and education research in their rhetorical nature, while acknowledging the divergent paths of these two fields of inquiry with regard to prediction and generalizability. The author suggests that given the highly contextualized nature of educational processes, embedded in shifting complex social settings, and the relevance of all variables, very little education research is able to pursue predictive power.