Lab - This is not normal! (Skewness in HCI research data)

Most HCI researchers rely on parametric tests (like t-tests) that assume a perfect "Bell Curve." But in the real world, task times, error rates, and SUS scores are notoriously skewed. If our data doesn't fit the model, our power analyses are wrong, leading to wasted resources or unreliable results.

In this lab, you will quantify the impact of skewness on statistical reliability by analyzing empirical datasets and performing Monte Carlo simulations.

(This lab is suitable for 1-2 persons)