Identifying indicators to guide adaptive scaffolding in games

In game-based learning, adaptive scaffolding can enhance the learning of domain-specific skills, known as firstorder scaffolding, and self-regulatory skills, known as second-order scaffolding. To design adaptive scaffolding, we need indicators that identify learning opportunities. Therefore we investigated how indicators of performance and self-regulation relate to overall game performance in a medical emergency simulation game. These indicators have the potential to guide the design of adaptive first-order and second-order scaffolding, respectively. Twenty-six fourth-year medical students played 116 game sessions. Using a multilevel model, we investigated the relationship between overall game performance and a range of online and offline measures. For first-order scaffolding, accuracy, systematicity and thoroughness were found to be valid indicators; for second-order scaffolding, high global self-regulatory scores and frequent monitoring were found to be valid indicators. These indicators can be included in future algorithms for adaptive scaffolding in game-based learning.