The analyses
The rail lists thirteen analyses in the order of the course. Pick by the question you are asking:
| Your question | Analysis | Module |
|---|---|---|
| What do my numbers look like — centre, spread, shape, outliers? | Descriptive statistics | 2 |
| Given some probabilities of two events, what is another? Bayes? How many arrangements? | Probability rules | 3 |
| What is the chance of exactly / at most / at least k successes or events? | Discrete distribution | 4 |
| What is the chance a measurement is below / above / between values? Which value is exceeded 5 % of the time? | Continuous distribution | 5 |
| How likely is a sample mean or proportion to land in a range? | Sampling distribution | 6 |
| How big is it, and how sure am I? | Confidence interval | 7 |
| How many observations do I need? | Sample size | 7 |
| Is one group's mean, proportion or variance equal to a claimed value? | One-sample test | 8–9 |
| Do two groups differ? | Two-sample test | 9–10 |
| How does y change with x? Predict y at a new x. | Simple linear regression | 11, 13 |
| How does y depend on several factors at once? | Multiple regression | 12–13 |
| Do three or more groups share a mean? Which pairs differ? | One-way ANOVA | 14 |
| Do two factors matter, and do they interact? | Two-way ANOVA | 14 |
How every analysis works
- Click it in the rail. It opens with a worked example from the course already filled in, so you can see what each result looks like. That example is not your data — replace it.
- Make the choices at the top of the form (the test, the distribution, Data from, the alternative hypothesis). The form shows only the fields your choice needs.
- Point it at your data: pick a column (see Data tables), or type summary numbers such as x̄, s and n.
- Read the RESULTS column. Click any result for its meaning in the help panel.
- Open Plots, Report and Code for the figures, the write-up with sources, and the Python and R.
A field left blank means "not known", not zero. A result that cannot be computed shows "—" and a red note says what is missing.
The formulas follow the course and Montgomery and Runger, Applied Statistics and Probability for Engineers, 7th ed.; some cite the NIST/SEMATECH e-Handbook of Statistical Methods. The exact source of every method appears in the help panel, the Report tab and the Instructions tab.