Quick answer: In jamovi: Exploration → Descriptives, move variables into the Variables box, tick the statistics and plots you want, and optionally split everything by a group with the Split by field. Descriptives are never a formality — they are the health check that decides which inferential test your data will tolerate. This guide walks the workflow and the options that matter. New to jamovi? Start with the first-analysis guide.
The descriptives workflow in five moves
- Open the dialog: Exploration → Descriptives. Move numeric variables into Variables; categorical ones can go there too — jamovi adapts the output to each variable’s measure.
- Tick the right statistics: under Statistics, choose mean, median, mode, SD, variance, skewness and kurtosis. Skip anything you cannot explain to your reader.
- Split by a group: the Split by box repeats the whole table per level — female/male, control/treated — which is exactly the per-group table most reports require.
- Add plots: histograms, box plots and bar plots render in the same output; tick Normality plot when a t-test is coming next.
- Check missingness: n per variable exposes gaps early — surprises here are cheap, surprises at the test stage are not.
Reading the output like an analyst
| Pattern in the output | What it tells you |
|---|---|
| Mean far from median | Skew — report the median, consider rank-based tests |
| SD larger than the mean | Huge relative spread — inspect for outliers or a mixed population |
| Skewness beyond ±1 or kurtosis < −1 | Non-normal shape — verify with the normality plot before t-tests |
| Different n across variables | Missing data — decide listwise vs pairwise handling before testing |
The same logic lives in every package — R Commander’s parallel workflow is in our R Commander descriptives guide — but jamovi’s advantage is that the whole thing lands formatted and plot-annotated in one step. Once shape and spread are understood, the natural next stop is the t-test walkthrough; the variable types driving all these choices are covered in the data types guide.
For one-to-one coaching on reading your own data honestly before testing it, Ampersand Academy teaches statistics with jamovi one-to-one.
Frequently asked questions
How do I get descriptive statistics in jamovi?
Exploration menu, Descriptives. Move variables into the Variables box, tick the statistics and plots you need under Statistics, and optionally use Split by to repeat the output per group. Everything renders in one formatted table.
Which descriptive statistics should I always report?
Sample size, mean and standard deviation for symmetric variables, median with interquartile range for skewed ones. Add minimum and maximum where ranges matter to the reader.
What does Split by do in the Descriptives dialog?
It repeats the entire summary for each level of a grouping variable – the per-group table that precedes any group comparison, without building separate analyses.
Why do descriptives matter before hypothesis tests?
They reveal skew, outliers and missing data that decide which test is valid. A t-test on badly skewed small samples fails assumptions that a ten-second histogram would have caught.
Can jamovi show frequency tables for categorical variables?
Yes – categorical variables in the Descriptives dialog produce frequency counts and percentages, and bar plots render alongside when the plot options are ticked.