#592: Adaptive Support for Collaborative Learning: Effects of Design Features Across Task Types

This meta-analysis examines how adaptive instructional support facilitates collaborative learning across task types and design features, based on 80 studies. Cognitive-model approaches were most effective for factual knowledge learning, whereas data science approaches showed the strongest effects in problem-solving. Human-involved adaptation yielded smaller effects. These findings suggest adaptive instructional support is most effective when its design features align with task demands, with data science approaches and system control being particularly effective for problem-solving tasks.