Statistical Significance Testing at CHI PLAY: Challenges and Opportunities for More Transparency
summary
A systematic audit of 119 CHI PLAY papers shows concrete problems: over half lack explicit hypotheses, many omit sample size and test details, and selective decisions during analysis can push nominal α≈5% to as high as 60%. The paper quantifies reporting gaps, notes that few authors share data, and provides a reusable template and OSF.io materials to distinguish exploratory from confirmatory work and to report stopping rules, outlier handling, and dropped measures.
AI-written summary · about