Equal-share average
Best when each observation should count equally. Add everything, divide by the count.
Find the ordinary mean, weighted average, combined group average, target average, or average rate of change — with the working shown and common averaging mistakes called out.
Paste a spreadsheet column or separate values with commas, spaces, semicolons, tabs, or new lines. Invalid tokens are rejected instead of silently ignored.
One pair per line. Weights may be counts, credits, quantities, or percentages such as 92:40%. They do not need to add to 100.
| Value | Weight | Share | Value × weight |
|---|---|---|---|
| Calculate to see contributions. | |||
This fixes the classic “average of averages” mistake. Group size must be a positive whole number.
Average rate of change is the secant slope: Δy ÷ Δx. It describes change over an interval, not an instantaneous derivative.
Best when each observation should count equally. Add everything, divide by the count.
Use for grades with credit hours, quantities at different prices, ratings with vote counts, and combined group means.
Useful when extreme values can drag the mean away from a typical observation, such as income or house-price data.
Geometric mean suits positive multiplicative factors; harmonic mean is useful for positive rates under specific equal-distance or reciprocal settings.
Deterministic practice — everyone gets the same question for the date.
What is the arithmetic mean?
Simple mean uses Σx ÷ n. Weighted mean uses Σ(wx) ÷ Σw. Combined group averages are the same weighted-mean problem with group size as the weight.
List sums use compensated summation. Geometric mean uses logarithms to reduce overflow risk. Root-mean-square is scaled before squaring. Results still use IEEE-754 JavaScript numbers, so extremely large or tiny values can lose precision.
The selected percentage is removed from each tail using a whole-number floor count. On short lists, a 10% trim may remove zero values; the result then equals the ordinary mean.
This is an educational calculator. It does not decide whether a mean is statistically appropriate, repair biased data, infer missing weights, or replace a documented statistical workflow for regulated or scientific analysis.