How to Make a Q-Q Plot on Mac & Windows

A Q-Q plot is the fastest visual check of whether your data is normal — a question that decides which statistical tests you can trust. Here's how to read one on Mac & Windows.

Many tests assume roughly normal data. A quantile–quantile (Q-Q) plot checks that assumption visually: it plots your data's quantiles against the quantiles of a normal distribution. If the data is normal, the points fall on a straight line.

Make it

Open your data and choose the Q-Q plot (set it to Normal Quantiles).
Pick the numeric variable to check.
Read how closely the points follow the reference line.

A normal Q-Q plot in Stratum — Points on the line mean approximately normal; departures reveal the shape.

Read the departures

The way the points leave the line tells you how the data is non-normal:

An S-curve — heavy or light tails.
Points curving up at both ends — right skew; curving down — left skew.
A few points flying off the end — outliers.

Plot first, test second. A Q-Q plot shows you the kind of departure; a formal test gives you a number. Stratum's Summary report includes an Anderson–Darling normality test — but with large samples even trivial departures turn “significant,” so trust the plot for judgment and the test for a figure to cite.

Why it matters

Normality underpins t-tests, ANOVA, and linear regression's inference. Checking it takes ten seconds and saves you from trusting a p-value you shouldn't. Full walk-through in Lesson 8.

Download Stratum

Frequently asked questions

What does a Q-Q plot tell you?

Whether your data follows a normal distribution. If the points lie on the reference line, the data is approximately normal; systematic departures reveal skew, heavy tails, or outliers.

How do I check for normality on Mac & Windows?

Make a normal Q-Q plot in Stratum and read how closely the points follow the line, then back it up with the Anderson–Darling normality test in the Summary report.

What does an S-shaped Q-Q plot mean?

Heavy or light tails relative to a normal distribution — the data has more (or fewer) extreme values than normal. Curvature at both ends pointing the same way indicates skew.

Should I trust a normality test or the Q-Q plot?

Use both, but lean on the plot for judgment — with large samples, formal tests flag trivial, harmless departures as significant. The plot shows whether the departure actually matters.