Statistics Helper

Imagine you asked ten friends how long their commute takes. If you add all ten numbers and divide by ten, that is the mean. But if one friend commutes 90 minutes while the other nine average 20, the mean jumps to 27 — higher than what nine out of ten people actually experience. The median — the middle value when sorted — would stay at 20, which is far more representative of the group.

Descriptive statistics work by reducing a list of numbers to a handful of landmarks. The mean tells you the center of gravity. The median tells you the true midpoint. The mode tells you the most common value, which matters in things like survey responses or defect counts. The range and standard deviation tell you how spread out the data is — whether values are tightly clustered or all over the place.

Standard deviation is the most useful spread measure because it is in the same units as your original data. A standard deviation of 5 kg means most values fall within 5 kg of the mean. Variance is the squared version — it is used in further calculations like ANOVA but is harder to interpret on its own. This tool gives you both, so you can use whichever your analysis requires.

Use descriptive statistics when you need to summarize a batch of numbers before presenting, comparing, or analyzing them. This tool is appropriate for student scores, survey responses, measurement series, financial figures over time, and any situation where you have a list of raw values and need to communicate what they look like as a group.

This tool is not appropriate when your values have different weights or importance. A student who took three exams worth different percentages of their grade needs a weighted mean, not a simple one. It is also not appropriate for time series where the sequence matters — average monthly revenue tells you less than the trend over those months. And it does not handle grouped frequency tables, where you know how many values fall in each bucket but not the individual values.

If you are comparing two groups and want to know whether the difference in means is statistically significant, you need a t-test or ANOVA — not just the means themselves. Descriptive statistics describe. They do not test hypotheses or establish causation.

The most common mistake is using the mean when the data is skewed. If you are summarizing salaries, house prices, or response times, the mean is inflated by high outliers. Using it as the headline figure overstates what is typical. The consequence is misrepresenting your data to an audience who will make decisions based on it.

A second mistake is confusing sample and population standard deviation. If your data is the full population — every employee at a five-person company, every part in a batch of 20 — you divide by n. If it is a sample drawn from a larger group, divide by n-1. Most calculators default to sample (n-1), including this one. Using the wrong formula understates or overstates spread, which can invalidate hypothesis tests.

A third mistake specific to this tool: entering percentage values and treating the mean as a weighted average. If three groups contributed 20%, 40%, and 60% but had different sizes, averaging those percentages gives you the wrong overall rate. You need a weighted mean, not a simple mean. The simple mean only applies when every value contributes equally.

Mean: sum all values, divide by count. For values x1 through xn, mean = (x1 + x2 + ... + xn) / n.

Median: sort values in ascending order. If count is odd, the median is the middle value at position (n+1)/2. If count is even, take the average of the two middle values at positions n/2 and n/2+1.

Mode: count how often each value appears. The value or values with the highest frequency are the mode. If all values appear equally often, there is no mode.

Sample standard deviation: calculate the mean, then for each value compute the squared difference from the mean, sum all squared differences, divide by n-1, then take the square root. Written out: s = sqrt( sum((xi - mean)^2) / (n-1) ). Variance is that same formula without the square root: s^2 = sum((xi - mean)^2) / (n-1).

Range: simply the maximum value minus the minimum value. It is the most sensitive summary statistic — a single extreme outlier changes it completely — which is why standard deviation is preferred for describing typical spread.

The sample standard deviation formula divides by n-1 because the sample mean is itself estimated from the data, not known independently — this costs one degree of freedom and makes the variance estimate slightly larger to correct for that bias. At large n, the difference between n and n-1 is negligible, but at n=5, dividing by n underestimates the true population spread by about 11%. Analysts working with small samples should be especially careful about which formula their tools use by default.