Standard Deviation Calculator

Sample, almost always. Use the population figure only when your numbers are genuinely everything you are describing, such as the marks of every student in one class. A survey, a batch of measurements or a month of readings is a sample of something larger, even when it does not feel like one.

Because a sample's own mean sits closer to that sample than the true mean of the wider group does, so the squared deviations come out too small. Dividing by n minus 1 instead of n compensates. It is called Bessel's correction, and it matters most on small datasets: the gap between the two figures shrinks as the count grows and is negligible past a few hundred values.

By linear interpolation, which is what Excel's QUARTILE function and NumPy's percentile default both use. There are nine recognised definitions of a quartile and they disagree on small datasets, so matching the two tools people are most likely to check against is the useful choice.

Then there is no mode and the page says so, rather than reporting the first value or the lowest. If several values tie for the most frequent, all of them are shown.

However it comes. Commas, spaces, semicolons and line breaks all work, so a column copied out of a spreadsheet pastes straight in. Anything that is not a number is skipped rather than counted as zero, which is the behaviour that quietly ruins an average.