Overview
A point estimate is a single best guess for a population value. For a proportion, it's the sample proportion p̂, shown here with its margin of error and confidence interval.
Enter the number of successes, the sample size and a confidence code: 1 for 90%, 2 for 95% or 3 for 99%.
Worked examples
| Inputs | Result |
|---|---|
| Successes (x)58Sample Size (n)120Confidence level95% | p-hat: 0.48 (48.33%)SE: 0.05Margin of error: 8.94%CI: [39.39%, 57.27%] |
Show 2 more examples
| Successes (x)30Sample Size (n)100Confidence level95% | p-hat: 0.3 (30%)SE: 0.05Margin of error: 8.98%CI: [21.02%, 38.98%] |
| Successes (x)450Sample Size (n)1,000Confidence level99% | p-hat: 0.45 (45%)SE: 0.02Margin of error: 4.05%CI: [40.95%, 49.05%] |
p-hat by sample size
With the calculator's starting values, the p-hat is 0.48. The table shows how it changes with the sample size.
| Sample Size (n) | p-hat |
|---|---|
| 60 | 0.97 |
| 80 | 0.73 |
| 100 | 0.58 |
| 120starting value | 0.48 |
| 140 | 0.41 |
| 160 | 0.36 |
| 180 | 0.32 |
Other inputs: Successes: 58 x. Change them in the calculator above for your own numbers.
How it works
The sample proportion, p̂, is successes ÷ sample size. Its standard error is √(p̂(1 − p̂) ÷ n). The margin of error is z × SE, and the confidence interval is p̂ ± that margin. This Wald interval is less accurate for small samples or proportions near 0 or 1, where the Wilson interval is better.
The formula
Practical tips
- 58 of 120 is a p̂ of 0.483, with a 95% interval of about 39.4% to 57.3%.
- For a rough rule, check that n × p̂ and n × (1 − p̂) are both at least 10.
- The margin of error is largest when p̂ is 0.5.
Frequently asked questions
What is p-hat?
The proportion of successes in your sample, used to estimate the population proportion.
How big a sample do I need?
For ±5% at 95% confidence near 50%, about 385.
Is it the same as a percentage?
Yes, p̂ × 100 is the percentage.