Statistics and regression from a list of numbers
Paste your data, get n, mean, median, mode, quartiles and both standard deviations; add a second column and the same page fits a line, a parabola or an exponential.
Applies to CalcMode statistics and regression calculator, /calculator/statistics-calculator · Updated 2026-11-09
The page holds two cards under the Statistics tab: one-variable statistics and two-variable regression. The tab row also offers Distribution, in case you need probabilities rather than summaries. Press Calculate to run a card, so you can paste a column and edit it before looking at any number - a link that already carries its data (?data=) is the exception, because then the answers are on screen the moment the page opens.
Entering the data
- Separate values with commas, spaces or semicolons - the parser treats all three the same, and line breaks are fine.
- There is no frequency column. To count a value twice, write it twice: 2, 4, 4, 4, 5, 5, 7, 9 is eight values.
- The default set is that same textbook example - 2, 4, 4, 4, 5, 5, 7, 9 - so the page is never blank.
- A token that is not a number stops the run with A value is not a number; an empty box says No data.
What comes back
| Row | For 2, 4, 4, 4, 5, 5, 7, 9 |
|---|---|
| Number of values n | 8 |
| Sum Σx | 40 |
| Mean x̄ | 5 |
| Median | 4.5 |
| Mode | 4 |
| Min / Max | 2 / 9 |
| Range | 7 |
| Quartiles Q1 / Q3 | 4 / 5.5 |
| Population standard deviation σx (divided by n) | 2 |
| Sample standard deviation sx (divided by n−1) | 2.13809 |
The median is the average of the two middle values when n is even, which is why this set reads 4.5 and not one of the data points. If every value appears once, Mode reads None - every value appears equally often.
Two-variable regression
- Write one point per line as x,y - or separate the points with semicolons: 1,2.1 2,3.9 3,6.2 4,7.8 5,10.1.
- Pick the model: Linear A + B·x, Quadratic A + B·x + C·x², or Exponential A·e^(B·x).
- Read the equation card, then the rows: intercept a, slope b, the quadratic term c when the model has one, correlation r, coefficient of determination r², number of points n.
- Type a value in Predict at x and the fitted y is added to the list.
| Model | Equation for the default five points | r² | Predicted y at x = 6 |
|---|---|---|---|
| Linear | y = 1.99x + 0.05 | 0.9973053289 | 11.99 |
| Quadratic | y = 0.02142857143x² + 1.861428571x + 0.2 | 0.997467225 | 12.14 |
| Exponential | y = 1.662830828·e^(0.3834343339x) | 0.9348456157 | 16.59543191 |
The quadratic barely improves on the line here, and the exponential is visibly worse - that is what the r² column is for. Two points always fit a line exactly, so a straight-line fit on two points tells you nothing about the model.
Refusals you should recognise
- Enter at least two points - a single point cannot define a line.
- A quadratic fit needs at least three points.
- All x values are the same - a vertical scatter has no slope.
- An exponential fit needs y > 0, because it is fitted through a logarithm.
- Each line needs x and y, for example 1,2.5 - the message appears when a line was typed as one number.
Open the tool
These links go straight to the web tool. The ones that carry an example open with it already filled in, so the result appears without typing anything; the rest open the tool ready for your own input.