Free · No signup · Runs in your browser

CSV A/B TestWelch t-test

Compare two groups in a CSV column and get a t statistic, two-sided p-value, and a clear significant / not-significant verdict at p < 0.05 - computed entirely in your browser.

01 · How it works

Three steps, then done.

Pick the column that labels your two groups and the numeric column you want to compare. The tool splits rows by group, computes each group's mean and standard deviation, and runs a Welch t-test that does not assume equal variances.

1

Load your CSV

Drop in a file with one column identifying the group (for example a control / variant label) and one numeric column to measure, such as conversion, revenue, or time.

2

Pick group and value columns

Choose the group column - it must contain exactly two distinct values - and the numeric value column. Blank cells and non-numeric values are skipped automatically.

3

Read the verdict

You get a per-group summary (n, mean, std), the mean difference, the Welch t statistic, degrees of freedom, a two-sided p-value, and whether the result is significant at p < 0.05.

02 · Why ours

Why use this CSV A/B test toolhonest stats, zero upload

A statistically sound two-group comparison that never sends your experiment data anywhere.

  • 01

    Welch by default

    Welch's t-test does not assume the two groups share a variance, so it stays valid when your control and variant have different spread or unequal sample sizes.

  • 02

    Real p-values

    The two-sided p-value comes from the t-distribution via a regularized incomplete beta function - not a rough lookup table - so small samples and large t statistics are handled correctly.

  • 03

    Fully private

    Everything runs client-side in your browser. Your raw experiment rows never leave the page, so there is nothing to upload and nothing stored on a server.

  • 04

    Clear verdict

    Beyond the raw numbers you get a plain significant: yes / no flag at the conventional 0.05 threshold, so you can act without re-deriving the test by hand.

"Welch's t-test compares two groups without assuming equal variance - the right default for messy real-world A/B data."
csvtodashboard.com
03 · FAQ

ab test questions.

What test does this run?
An unpaired, two-sided Welch t-test (also called the unequal-variances t-test). It compares the means of two independent groups without assuming their variances are equal, which is the safer default for most A/B experiments.
An A/B test compares precisely two groups. If your chosen group column has one group, or three or more, the tool returns a short note instead of a result. Filter your data or pick a column with exactly two distinct values.
From the t statistic and Welch-Satterthwaite degrees of freedom, the tool evaluates the Student t-distribution tail using a regularized incomplete beta function, giving an accurate two-sided p-value rather than an approximation from a table.
Rows with a blank group label are ignored, and any value cell that is empty or non-numeric is skipped. Only rows with both a valid group and a parseable number contribute to each group's statistics.
No. The entire calculation happens in your browser using client-side JavaScript. Your CSV is never transmitted to or stored on any server.
Related tools
CSV Group By
Aggregate rows by a key column with sums, averages, and counts to summarize each segment before testing.
Open →
CSV Profiler
Profile every column's type, distribution, and missing values to sanity-check your data before running a test.
Open →
CSV Sort & Filter
Narrow your CSV to the exact two groups you want to compare before running the A/B test.
Open →
CSV to Pivot
Cross-tabulate group versus metric to eyeball differences before a formal significance test.
Open →
CSV Validate
Catch malformed rows and bad numeric cells so your test runs on clean, trustworthy data.
Open →