What this calculator does
A point estimate is the single best guess for an unknown value, such as a population proportion or mean. This calculator gives it two ways: from successes and trials, with the maximum likelihood, Wilson, Laplace and Jeffreys estimates of a proportion, or from a confidence interval, where the point estimate is the midpoint and the margin of error is half the width.
How to use it
- Choose from successes and trials or from a confidence interval.
- For counts, enter the successes (x), the trials (n) and a confidence level for the Wilson estimate.
- For an interval, enter its lower and upper bounds.
How to calculate a point estimate
- Maximum likelihood estimate (MLE) = x ÷ n, the sample proportion
- Wilson = (x + z²/2) ÷ (n + z²), where z is 1.96 for 95%
- Laplace = (x + 1) ÷ (n + 2); Jeffreys = (x + 0.5) ÷ (n + 1)
- From an interval: point estimate = (lower + upper) ÷ 2, margin of error = (upper − lower) ÷ 2
Examples
Example 1: From a confidence interval. A 95% confidence interval runs from 62.4 to 70.8. The point estimate is the midpoint, (62.4 + 70.8) ÷ 2 = 66.6, and the margin of error is half the width: 4.2.
Example 2: A survey proportion. 180 of 400 people said yes. The point estimate is 180 ÷ 400 = 0.45. The Wilson estimate is 0.4505 and Laplace 0.4502, both nudged slightly toward 0.5.
Tips
- For large samples the estimates agree. They differ most when n is small or the proportion is near 0 or 1, where Wilson and Laplace avoid answers of exactly 0 or 1.
- Homework usually wants the MLE, x ÷ n, unless it names another method.
- Build an interval around the estimate with the confidence interval calculator or the margin of error calculator.