In this game, every round you will be presented ten sets of research results. For half of the rounds, the null hypothesis will be true for all ten sample means. This means that any differences between the ten sample means and the population mean is purely due to sampling error (i.e., from having used random selection). For the other half of the rounds, the research hypothesis will be true, but only for one sample mean. That particular sample mean will be the result of both sampling error and the treatment effect. In each round, your goal is to either pick the sample where the research hypothesis is true, or to state that none of the samples were affected. To put you in the power seat, you get to decide both the sample size (5, 25, 100, or 1000) and the size of the treatment effect (small, medium, or large). Play as many rounds as you like. Can you score ten correct decisions? What combination of sample size and effect size makes your decisions easy? What combinations are hard? Which ones are fairly doable? Play and decide. To get started, click the 'Draw Samples' button. You will then see ten samples, drawn randomly from the population. A summary of the results will be provided. After reviewing the results, make your decision.
Controls
Generates ten independent random samples at the chosen size.
Sample size equals 25 and effect size is Large.
Summary
There are no samples to summarize.
The population histogram appears bell-shaped, with a slightly longer tail on the left side.
The scores represent college students' overall life satisfaction (on the SWLS assessment).
Each score ranges from 5 (extremely dissatisfied) to 35 (extremely satisfied).
The mean for the population is 24.
Legend:
Purple solid = Normal curve (μ=24, σ=6),
Red dashed = Population mean (μ=24),
Green solid = Sample mean (x̄).
Population (M = 24, SD = 6)
Population chart coming in Phase 2
The SWLS scores are bounded at 5 and 35, and demonstrate a negative skew to the left. The X-axis ticks are at 5, 10, 15, …, 35. Bin width = 1.
Toggle the blue theoretical normal curve overlay on all charts.
Samples (10, All generated at once)
Each sample will display a histogram on the same axes as the population. Below each chart a text summary will report min, max, and mean.