Abstract
Math learning programs were expected to revolutionize students' learning, but their effects so far have mostly been disappointing. Following the debate about why to continue research on math learning programs, we aimed to reformulate this question into how to continue this research. Investigations to date have neither considered a sufficiently wide set of outcome variables nor differentiated between performance measures (e.g., measuring addition and subtraction performance separately) and affective-motivational variables. Moreover, as students can only benefit from a program if they use it, researchers need to take practice behavior into account. Thus, we investigated whether the adaptive arithmetic learning program Math Garden fostered students' addition and subtraction performance, their math self-concept, and a reduction of their math anxiety. We also investigated how practice behavior (practiced tasks/weeks) affected these outcomes. We used a randomized pretest-posttest control group design with 376 fifth-grade students in Germany. Students in the experimental condition practiced with Math Garden for 20.7 weeks and had an increase in math self-concept. The more subtraction tasks the students practiced, the more they improved their subtraction performance. We found no effects on math anxiety. The results are discussed in terms of providing a starting point for new directions in future research.