Unit 2 · Describe your data
How to check normality in SPSS with Explore
Run Shapiro-Wilk and Kolmogorov-Smirnov tests in SPSS, check skewness and kurtosis z-scores, and read Q-Q plots.
Quick answer: How do you check normality in SPSS with Explore?
- Click Analyze › Descriptive Statistics › Explore.
- Move the outcome into Dependent List (and the grouping variable into Factor List if you have groups).
- Under Plots… tick Normality plots with tests, then click OK.
- Read the Shapiro-Wilk Sig.: above .05 means no significant departure from normality.
Report it in APA style: Scores did not differ significantly from a normal distribution, W(100) = .99, p = .384.
What is normality, and why check it?
A normal distribution is the symmetrical, bell-shaped curve that many tests (t-tests, ANOVA, Pearson correlation, regression) assume, either for the scores or for the residuals. Checking normality tells you whether these parametric tests are appropriate or whether a non-parametric alternative is safer.
When to use it
- Before t-tests and ANOVA: check the outcome separately in each group.
- Before Pearson correlation: check each variable.
- In regression, check the residuals rather than the raw scores.
How it works
Use three kinds of evidence together. Statistics: skewness (lopsidedness) and kurtosis (peakedness) divided by their standard errors give z-scores; values within ±1.96 suggest no serious problem. Tests: Shapiro-Wilk compares your data with a normal curve; p > .05 means no significant departure. Plots: on a Q-Q plot, points near the diagonal mean the data are close to normal. In large samples the tests flag trivial departures, and the central limit theorem makes t-tests and ANOVA fairly robust, so rely more on plots.
Worked example
Exam scores: skewness z = 1.02, kurtosis z = −0.82, Shapiro-Wilk W(100) = .99, p = .384, and the Q-Q plot hugs the line. Normality is reasonable, so a t-test is fine.
Key terms
- Skewness
- Asymmetry. Positive skew has a long tail to the right.
- Kurtosis
- How heavy the tails are compared with a normal curve.
- Shapiro-Wilk
- Normality test recommended for small and medium samples.
- Q-Q plot
- Plots observed against expected values; a straight line means normal.
Common mistakes
- Checking normality of the whole sample instead of each group.
- Treating p = .049 on Shapiro-Wilk as a disaster in a sample of 500.
- Switching to a non-parametric test without saying why in your method.
How to do it in SPSS
The SPSS syntax
Prefer typing to clicking? Open File › New › Syntax, paste this in, select it and choose Run › Selection. Every dialog's Paste button writes the same kind of code.
EXAMINE VARIABLES=score
/PLOT BOXPLOT HISTOGRAM NPPLOT
/STATISTICS DESCRIPTIVES
/MISSING LISTWISE
/NOTOTAL.Practice data
Download the file used in this lesson and run it yourself in SPSS. Your results will closely match the screens above, though a few numbers may differ slightly.
- exam_scores.savExam scores for checking normality (N = 100)
Now practise it interactively →
Reading is only half of it. Tap through the real SPSS menus, read the output and pick the right write-up in 7 quick questions, with instant feedback on every answer.
Practice for this lesson is part of Pro, with every unit, unlimited hearts and saved progress.