| Title | ABCmouse Mastering Math: Using Gameplay to Estimate Learning Outcomes |
| Presenter(s) | Daniel Jacobs, Kp Thai, Eric Keylor, V. Elizabeth Owen and Marie-Helene Roy |
| Session | Conference Reception, Game Exhibition, and Poster Session |
| Time | Thursday, October 11, 7:00p-10:00p |
| Location | MSU Union Ballroom |
| Format | Poster Presentation |
| Description | For educational games to be most effective, they should provide reliable measures of learning. This study evaluates how gameplay in ABCmouse Mastering Math - an adaptive educational game for early number sense skill development in preschool and kindergarten students - can estimate learning outcomes on a well-validated external assessment, Test of Early Mathematics Ability, 3rd edition (TEMA-3). To track how estimates of learning outcomes from Mastering Math change over time and how much gameplay is necessary for 'good enough' estimates, we built predictive models to estimate the predicted posttest TEMA-3 score from MM using in-game user telemetry. Results showed strong correlations between the predicted posttest score from Mastering Math and the observed TEMA-3 posttest score in several cross-validated models (R^2 ranged 0.47 - 0.59), suggesting that Mastering Math can provide reliable measures of student progress on specific number sense skills in the classroom. |
