Frequency Table Maker

Choose exact values or class intervals before reading the table

Ungrouped mode keeps each distinct value or label as its own row. Grouped mode is for numeric measurements where nearby values should be summarized into class intervals. The two views answer different questions, so the studio keeps the grouping decision visible instead of silently binning every dataset.

Compare class rules instead of treating one heuristic as truth

Sturges, square-root and Rice rules are practical starting points for the number of classes, not universal statistical laws. You can override them with an explicit class count, a fixed class width and a custom first lower boundary. Every grouped row shows its midpoint, frequency, relative frequency, cumulative count and cumulative percentage.

Check the raw data separately from the grouped approximation

Mean, median, quartiles and sample standard deviation are calculated directly from the valid raw numeric values. They are not reconstructed from class midpoints. Invalid numeric tokens are listed in the audit instead of disappearing silently, and the final class explicitly includes the observed maximum so a boundary value is not dropped.

Use each chart for the evidence it actually shows

The histogram/bar view emphasizes frequency by row or class. The frequency polygon connects class midpoints or exact rows to show shape. The ogive plots cumulative frequency and is useful for locating cumulative thresholds. None of these charts turns a poor class rule into a better statistical model.

Keep datasets local unless you explicitly export them

Calculations run in the browser. Project JSON and local saves include the dataset only when you choose those actions. The settings-only share link excludes values, which makes it useful for reproducing the chosen class rule without putting the raw observations in the URL.