Quick answer: R Commander gives you three descriptive routes: Statistics → Summaries → Active data set (everything at once), Numerical summaries (mean/SD/quantiles per variable, grouped if you like), and Graphs (histograms, boxplots, bar plots). Descriptives are never optional: they check data sanity, reveal outliers, and tell you which inferential test the data will tolerate. Import logistics are in our data import guide; here we focus on the summaries themselves.
The three routes, and when to use each
| Route | Menu path | Use when |
|---|---|---|
| Whole-dataset summary | Statistics → Summaries → Active data set | First look: n, missing, mean, quartiles for every column |
| Numerical summaries | Statistics → Summaries → Numerical summaries | Chosen variables, optional grouping factor, chosen stats |
| Frequency distributions | Statistics → Summaries → Frequency distributions | Categorical variables: counts and percentages per level |
Reading the numerical output like a pro: compare mean vs median — a mean notably higher than the median means right-skew (income, length-of-stay, growth curves) and flags the outliers a histogram would show. SD vs IQR tells the robustness story; and n per variable exposes missingness patterns before they surprise a t-test. What those distributions mean statistically — and why skew matters for test choice — is in our distributions guide.
The graphs that earn their place
- Histogram (Graphs → Histogram): the shape check before any t-test or ANOVA — normal-ish, skewed, bimodal?
- Boxplot (Graphs → Boxplot): group comparisons at a glance and the fastest outlier detector; tick Identify outliers with mouse to label points interactively.
- Bar/line + quantile-comparison plot (Graphs → QQ plot): the normality eyeball that pairs with Shapiro-Wilk when reviewers ask.
One commandment: report descriptives in a table (per group: n, mean, SD, median, min–max) before any p-value appears — the APA-style habit that examiners and reviewers reward. The full logic of why descriptives gate inference is in the introduction to biostatistics, and the descriptive-first workflow with menus-only software is paralleled in our jamovi first-analysis guide.
Next stop: the tests
When descriptives are on paper, the inferential catalog — t-tests, ANOVA, chi-square, correlations, non-parametrics — is mapped menu-by-menu in statistical tests in R Commander; and if you haven’t installed R Commander yet, the two-minute setup is in the installation guide. For one-to-one practice describing real datasets, Ampersand Academy teaches R and statistics one-to-one.
Frequently asked questions
How do I get summary statistics in R Commander?
Statistics menu, Summaries. Active data set summarizes every column at once; Numerical summaries lets you pick variables, statistics and an optional grouping factor; Frequency distributions handles categorical variables.
What does it mean when the mean is much larger than the median?
The variable is right-skewed – a few large values pull the mean up. Report the median and interquartile range instead, and consider rank-based tests rather than t-tests for inference.
How do I make a histogram or boxplot in R Commander?
Graphs menu: Histogram for distribution shape, Boxplot for group comparison and outliers. Both ask for the variable, and the boxplot can label outliers with a mouse click.
How do I group summaries by a factor?
In the Numerical summaries dialog, move the grouping variable into the Summarize by groups box. Every statistic is then reported per level – the descriptives table most reports require.
Which descriptive statistics belong in a thesis table?
Per group: sample size, mean, standard deviation, and for skewed variables the median with interquartile range. Add minimum and maximum where ranges inform readers. This one table precedes every inferential test.