Testing the Power of Game Lessons: The Effects of Art and Narrative on Reducing Cognitive Biases

This study examines whether an educational digital game can reduce three cognitive biases (fundamental attribution error, confirmation bias, and bias blind spot) more effectively than a professionally produced training video. Conducted with 480 participants recruited from three U.S. universities in Fall 2012, the research employed a controlled 2x2 experimental design that varied art style (detailed versus minimal) and narrative richness (rich versus light), with the game built in Flash and compared against the video as a benchmark.

The central finding is that the game outperformed the video immediately after play for fundamental attribution error and confirmation bias, and these gains partially persisted at an eight-week retention test. Bias blind spot proved considerably more resistant to reduction, showing little improvement except in one game condition. Minimal art was the only design variable significantly associated with learning, and specifically for bias blind spot at immediate post-test, though this advantage disappeared by the retention measurement.

Beyond learning outcomes, the study found that players generally preferred minimal art to detailed art. Rich narrative paired with detailed art appeared to raise player expectations, producing confusion and criticism, while minimal art was favored in the minimal-narrative condition. Narrative's effect on learning itself was inconclusive, and the authors noted that confirmation bias was especially difficult to convey consistently across different players.

Overall, the work supports the potential of well-designed games as tools for bias mitigation, while leaving open questions about how art and narrative shape both engagement and durable learning. The authors call for further research to clarify these design effects in educational games.