🧪 Educational Educational Research Simulation

This simulator runs a randomized classroom trial from the ground up: pick how many students go in each of the control and treatment groups, dial in the true effect size and the natural spread of test scores, then run the trial and watch each student's post-test score appear as a floating sphere in a 3D scene — blue for control, orange for the treatment group that received the intervention. Behind the scene, a real Welch's t-test computes the exact p-value from the incomplete beta function, the observed Cohen's d is recovered from the pooled sample standard deviation, and an a-priori power estimate shows whether the chosen sample size was ever large enough to reliably detect the true effect. It's the same statistical machinery behind published education-research trials — random assignment, sample-size planning, and the gap between a "real" effect and a "detected" one.