Compare Two Data Sets: Mean, Variation & Histograms

Compare two measurement data sets using descriptive statistics and shared-scale histograms. Copy a table for Excel or save a PNG, entirely in your browser.

No signup · Calculated in your browserInput data and results are not stored or sent to servers, analytics or external APIs. Inputs are lost when you leave, reload or switch language.

The file contains your input data. It is not sent to a server. Results are not saved. Maximum file size: 2 MiB.

1. Enter data to compare

Enter one number per line, with 2–10,000 values per condition. Blank lines are ignored. Omit headings, units and thousands separators.

Use the same measurement and unit for both conditions. One-sided limits are supported. Values on a limit count as within specification.

Paste measurements and export the comparison

Enter 2–10,000 numbers per condition, one per line. Use a decimal point and omit headings, units and thousands separators. Blank lines are ignored. Use the same measurement and unit for both conditions; names and specification limits are optional.

Compare the data, then copy the table for Excel or save the shared-scale chart as a PNG. If clipboard access is unavailable, select and copy the fallback text. Changing an input clears the previous results until you compare again.

Read the statistics and histograms

The mean describes the arithmetic center and the median the middle of the sorted values. Sample standard deviation is the square root of the sum of squared deviations from the mean divided by n−1.

Both histograms use the same horizontal range, bin widths and percentage scale. Each percentage is relative to its own condition's sample size. The common range includes specification limits and is split into 5–20 bins, so distant limits can make the distributions look narrow.

Values on a specification limit count as within specification. One-sided limits are supported. The within-specification percentage describes only the entered sample; it is not a guarantee of population yield or future quality.

What this comparison cannot establish

This tool provides descriptive statistics, not hypothesis tests, confidence intervals or paired-data analysis. Similar means do not prove equivalence, and a before/after difference does not establish causality. Check measurement methods, lots, collection periods and sample sizes before drawing conclusions.

Sources and edition dates

English edition updated 2026-09-21. Translation date: 2026-09-21. Japanese source content updated: 2026-09-16.

Reviewed by RYO on 2026-09-21. Published 2026-09-21.