Free online tool
Free statistics in your browser — t-tests, ANOVA, chi-square and correlation
Paste a column of numbers and this free tool runs real statistics in your browser: descriptive stats, one-sample, two-sample (Welch) and paired t-tests, a chi-square test of independence, Pearson and Spearman correlation, and one-way ANOVA — with the test statistic, degrees of freedom, an exact p-value, and a plain-English verdict. It is a free SPSS alternative for the tests most students and analysts actually run: no install, no account, no upload, and nothing is stored. p-values are computed from the regularized incomplete beta and gamma functions — the same mathematics statistics packages use — and shown to 4 decimal places, with anything below 0.0001 shown as “< 0.0001” instead of a misleading 0.0000.
Quick answer
Quick answer
Paste a column of numbers and this free tool runs real statistics in your browser — descriptive stats, one-sample, Welch two-sample and paired t-tests, chi-square, Pearson and Spearman correlation, and one-way ANOVA — with the test statistic, degrees of freedom, an exact p-value, and a plain-English verdict. No install, no account, no upload: a free SPSS alternative for the tests most people actually run.
Open the interactive Browser Statistics Studio
The interactive tool loads in the page. Enable JavaScript to enter data, process content, and download results.
Worked example — 10 exam scores vs a pass mark of 80
The built-in sample: 10 exam scores — 72, 85, 90, 78, 88, 95, 67, 82, 91, 76 — tested against 80. All values below were produced by the tool's own engine and cross-checked against SciPy.
| Mean | 82.4 — median 83.5, sample SD 9.0333 |
|---|---|
| Standard error | 9.0333 / √10 = 2.8566 |
| One-sample t-test | t = (82.4 − 80) / 2.8566 = 0.8402, df = 9, p = 0.4226 |
| Verdict | Not significant at the 0.05 level — the gap could easily be chance |
| Chi-square sample | 2×2 table 30, 10 / 20, 40: chi-square = 16.6667, df = 1, p < 0.0001 — significant |
| ANOVA sample | Three groups, means 5, 9, 10: F = 9.2647, df 2 and 15, p = 0.0024 — significant |
Which test do I need?
| Question | Test | What you paste |
|---|---|---|
| What are my numbers like? | Descriptive statistics | One number per line |
| Does one mean differ from a known value? | One-sample t-test | One number per line + the value to test against |
| Do two independent groups differ? | Two-sample t-test (Welch) | Two columns, one pair per line |
| Did the same subjects change? | Paired t-test | Two columns: before, after |
| Are two categories associated? | Chi-square test of independence | One row of counts per line |
| Do two columns move together? | Pearson or Spearman correlation | Two columns, one pair per line |
| Do three or more groups differ? | One-way ANOVA | One group per line |
Reference p-values — checked against the engine
| Statistic | p-value | Verdict at 0.05 |
|---|---|---|
| t = 2.0, df = 10 | 0.0734 | Not significant |
| t = 2.228, df = 10 | 0.0500 | Not significant — just above 0.05 |
| Chi-square = 3.84, df = 1 | 0.0500 | Not significant — just above 0.05 |
| Chi-square = 16.67, df = 1 | < 0.0001 | Significant |
| F = 9.26, df 2 and 15 | 0.0024 | Significant |
| r = 0.9719, n = 10 | < 0.0001 | Significant |
About this tool
About the Browser Statistics Studio
Paste a column of numbers and this free tool runs real statistics in your browser: descriptive stats, one-sample, two-sample (Welch) and paired t-tests, a chi-square test of independence, Pearson and Spearman correlation, and one-way ANOVA — with the test statistic, degrees of freedom, an exact p-value, and a plain-English verdict. It is a free SPSS alternative for the tests most students and analysts actually run: no install, no account, no upload, and nothing is stored. p-values are computed from the regularized incomplete beta and gamma functions — the same mathematics statistics packages use — and shown to 4 decimal places, with anything below 0.0001 shown as “< 0.0001” instead of a misleading 0.0000.
Detailed guide
How to get useful results from Browser Statistics Studio
A free SPSS alternative that runs in your browser
SPSS, Stata, and similar packages charge hundreds of dollars a year and need an install; most of their everyday use is a handful of tests — a t-test here, a chi-square there, a correlation for a coursework table. Browser Statistics Studio covers exactly those tests, free, in a tab: paste your data, pick the test, and read the statistic, the degrees of freedom, the p-value, and a verdict in plain English.
Nothing is uploaded and nothing is stored — the calculations run in JavaScript in your browser, so your data never leaves your device. There is no account, no trial clock, and no export locked behind a paywall: the results download as a CSV. It is not a replacement for a full statistics package — it does not do regression, multiple-comparison corrections, or one-tailed tests — but for the standard tests below, it gives the same numbers.
Descriptive statistics first
Every analysis should start with a description of the data, and the tool reports it all: count, sum, mean, median, mode, sample and population standard deviation and variance, first and third quartiles, interquartile range, minimum, maximum, and range. The built-in sample — 10 exam scores from 67 to 95 — has a mean of 82.4, a median of 83.5, and a sample standard deviation of 9.0333.
Two numbers are easy to mix up. The sample standard deviation divides by n − 1 and is the one inferential tests use; the population standard deviation divides by n and describes only the values in front of you. Quartiles use linear interpolation, the same method as R and Excel PERCENTILE.INC. If the mean and median disagree by a lot, the data are skewed and the median is the safer summary.
t-tests: one sample, two samples, paired
The one-sample t-test asks whether a sample mean differs from a known value — for example, whether a class average of 82.4 differs from a pass mark of 80. The built-in sample gives t = 0.8402 with 9 degrees of freedom and p = 0.4226, which is not significant: a gap of 2.4 points could easily be chance with 10 students.
The two-sample test compares two independent groups, and this tool uses Welch's version, which does not assume the groups have equal variances — the safer default. The paired test is for the same subjects measured twice, such as before-and-after scores, and works on the differences: the sample's mean improvement of 3.0 points gives t = 11.2250 with 7 degrees of freedom and p < 0.0001, a clear significant result. Each test shows the group means and standard deviations alongside the statistic, so the size of the difference is visible next to its p-value.
Chi-square test of independence
Chi-square answers a different question: are two categories associated? Paste a table of counts — one row per line, columns separated by commas — and the tool compares the observed counts with the counts independence would predict, cell by cell. The built-in 2×2 table (30, 10 / 20, 40) gives chi-square = 16.6667 with 1 degree of freedom and p < 0.0001, a significant association.
The test needs counts, not percentages, and the chi-square approximation is only reliable when every expected count is at least 5 — the tool shows the smallest expected count and warns you when it falls below 5. A significant result says the categories are associated; it does not say which cells drive the association, so the observed-vs-expected table is shown to make that visible.
Correlation: Pearson vs Spearman
Correlation measures how two columns move together. Pearson's r measures a straight-line relationship and needs roughly normal, continuous data; Spearman's rho ranks the data first, so it measures any steady (monotonic) relationship and works for ordered data too. The built-in sample — 10 pairs of study hours and scores — gives Pearson r = 0.9719 and Spearman rho = 0.9695, both with p < 0.0001: a strong positive relationship.
Strength is a matter of size, not of the p-value: below 0.3 is usually called weak, 0.3–0.7 moderate, above 0.7 strong. And a significant correlation is not causation — it says the columns move together, not that one causes the other. Both tests report a t statistic and degrees of freedom (n − 2) alongside the coefficient, so the result can be checked against a textbook.
One-way ANOVA
ANOVA compares three or more groups at once: instead of running several t-tests (which inflates the false-positive rate), it asks whether any group mean differs from the rest. Paste one group per line. The built-in sample — three groups of six with means 5, 9, and 10 — gives F = 9.2647 with 2 and 15 degrees of freedom and p = 0.0024, significant.
The full ANOVA table is shown: sums of squares, degrees of freedom, and mean squares between and within groups, with the F ratio. A significant F means at least one group differs — it does not say which one, so a follow-up with pairwise comparisons is the next step. ANOVA assumes independent observations from roughly normal distributions with similar variances; with very unequal group sizes or heavy skew, treat the p-value with caution.
Reading p-values honestly
A p-value is the probability of seeing a result at least this extreme if nothing were really going on. Below 0.05, the result is called significant — unusual enough under “nothing going on” to be worth reporting. The tool marks that verdict plainly, but significance is not importance: with enough data, a trivial difference becomes significant, and with little data, a large difference can stay non-significant.
Three numbers in the tables below were checked against the engine itself: a t of 2.0 with 10 degrees of freedom gives p = 0.0734 (not significant), a chi-square of 3.84 with 1 degree of freedom gives p = 0.0500 (just above the line, not significant), and an F of 9.26 with 2 and 15 degrees of freedom gives p = 0.0024 (significant). Running many tests on the same data inflates the chance of a false positive, so a single p-value just under 0.05 from a fishing expedition is weak evidence. For a thesis, a paper, or a decision that affects people, have a statistician check the analysis.
Quick guide
How to use this tool
- 01Pick the test — descriptive statistics, a t-test, chi-square, a correlation, or one-way ANOVA.
- 02Paste your data in the format shown under the box: one number per line for descriptives, two columns for t-tests and correlations, one row of counts per line for chi-square, one group per line for ANOVA.
- 03Press Run test and read the statistic, degrees of freedom, and p-value first, then the plain-English verdict.
- 04Open the supporting tables — descriptives, observed vs expected, or the ANOVA table — to see what drives the result.
- 05Download the results CSV to keep the numbers, and check the assumptions note before quoting a p-value.
Capabilities
Features and supported formats
- Descriptive statistics: mean, median, mode, sample and population SD and variance, quartiles, IQR, min, max, and range.
- One-sample, Welch two-sample, and paired t-tests with statistic, degrees of freedom, exact p-value, and a plain-English verdict.
- Chi-square test of independence for any table size, with an observed-vs-expected table and a warning when an expected count falls below 5.
- Pearson and Spearman correlation with r, t statistic, degrees of freedom, and p-value — Spearman uses average ranks for ties.
- One-way ANOVA with the full table: sums of squares, degrees of freedom, mean squares, F, and p-value.
- One-click sample data for every test, clear, and CSV export of the results — all in the browser, nothing uploaded.
Supported formats
Input: numbers pasted one per line or comma-separated (descriptives, one-sample t-test); two comma-separated columns, one pair per line (Welch, paired, Pearson, Spearman); one row of counts per line (chi-square); one group per line, values comma-separated (ANOVA) — up to 20,000 numbers. Output: test statistic, degrees of freedom, p-value to 4 decimals (below 0.0001 shown as “< 0.0001”), a significant / not significant verdict at 0.05, supporting tables, and CSV export.
Answers
Frequently asked questions
Is Browser Statistics Studio really a free SPSS alternative?
For the tests most people run in SPSS — descriptive statistics, t-tests, chi-square, correlation, and one-way ANOVA — yes: it gives the same statistics and p-values, free, in your browser, with no install and no account. It is not a full SPSS replacement: it does not do regression, factor analysis, multiple-comparison corrections, or one-tailed tests.
How do I run an online t-test with this calculator?
Pick the t-test you need — one-sample to compare a mean against a known value, two-sample (Welch) to compare two independent groups, or paired for before-and-after measurements on the same subjects. Paste your numbers (two columns for the two-sample and paired tests), press Run test, and read the t statistic, degrees of freedom, p-value, and verdict. The built-in sample loads in one click.
How does the chi square test calculator online work?
Paste a table of counts — one row per line, columns separated by commas, for example “30, 10” on one line and “20, 40” on the next. The tool computes the counts independence would predict for each cell, the chi-square statistic, degrees of freedom, and p-value, and shows the observed-vs-expected table. It warns you if any expected count falls below 5, where the approximation becomes unreliable.
How do I use the ANOVA calculator online?
Choose one-way ANOVA and paste one group per line, values separated by commas — at least two groups, at least two values each. The tool returns the full ANOVA table (sums of squares, degrees of freedom, mean squares, F, and p-value) plus the group means and a verdict. A significant F means at least one group differs; it does not say which one.
What is a statistics calculator for 2027 — is this up to date?
The mathematics has not changed: t, chi-square, F, and correlation tests are the same in 2027 as they were decades ago, and this tool computes their p-values from the regularized incomplete beta and gamma functions, the same approach statistics packages use. What is current is the delivery — no install, no account, runs in any modern browser on phone or desktop.
What does the p-value mean?
A p-value is the probability of seeing a result at least this extreme if nothing were really going on. Below 0.05 the result is called significant at the 0.05 level — unusual enough to report. It is not the probability that your hypothesis is true, and it says nothing about how large or important the effect is.
Does this tool upload my data?
No. Numbers are pasted into a text box and calculated only in your browser tab with JavaScript. Nothing is sent to a server, nothing is stored, and no value ever leaves your device.
How many numbers do I need?
Descriptives and the one-sample t-test need at least 2 numbers; two-column tests need at least 3 pairs; chi-square needs a table of at least 2 rows and 2 columns; ANOVA needs at least 2 groups of at least 2 values. Small samples make p-values noisy — a non-significant result from 8 observations is weak evidence, not proof of no effect.
Welch or Student t-test — which one is used here?
The two-sample test here is Welch's, which does not assume the two groups have equal variances. It is the safer default and stays valid when the groups' spreads differ; when the variances are equal, its answer is nearly identical to Student's.
Pearson or Spearman correlation — which should I pick?
Pick Pearson when both columns are continuous and the relationship looks like a straight line. Pick Spearman when the data are ranks or ordered categories, when the relationship is steady but curved, or when outliers and skew make Pearson unreliable. Spearman ranks the data first (ties share the average rank) and then applies Pearson's formula to the ranks.
Can I use it on my phone?
Yes. The paste box works in mobile browsers and the whole calculation runs in the tab. Paste a column of numbers from your notes or spreadsheet app and the results appear in seconds.
Does it do one-tailed tests or confidence intervals?
No. Every test here reports the standard two-tailed p-value at the 0.05 level, and no effect-size confidence intervals are shown. If your hypothesis is directional or you need an interval, use the reported statistic and degrees of freedom in a textbook table, or a full statistics package.