A thematic evaluation of data journalism courses resulted in a typology that parses the field and offers guidance to educators. At the center is pattern detection, preceded by data acquisition and cleaning, and followed by data representation. The typology advances academic understanding by offering a precise conceptualization that distinguishes data journalism from peripheral technologies and identifies coding as a supportive skill. It also enables a fresh definition of data journalism as the primary reliance on numerical evidence as a journalistic tool in detecting patterns, or the visual representation of numerical evidence to enable audiences to discern patterns.