Tamanho amostra t-teste
- Created by
- Renato Passos, Eng. de Software
- Reviewed by
- Renato Passos, Eng. de Software
Last updated: Apr 18, 2026
About this calculator
The t-test sample size calculator determines the minimum number of observations required to ensure statistical validity in an experiment. It uses the formula (z_α/2 + z_β)² · 2σ² / Δ², where z_α/2 and z_β are critical values linked to significance level and test power, σ² is population variance, and Δ is the minimum effect of interest. This prevents underpowered studies that might fail to detect meaningful results.
To use the calculator, define the significance level (typically 5%), desired power (often 80% or 90%), and estimate variance and expected effect. The output is an integer representing the minimum sample size. Note that larger Δ values reduce sample size needs, while higher variance increases them.
This tool is widely used in experimental research, social surveys, and A/B testing. However, it assumes normal distribution of data and accurate variance estimation. If these assumptions are not met, the calculated size may be unreliable.
Frequently asked questions
What is a t-test and when to use it?
A t-test compares means of small samples (typically n < 30) when population variance is unknown. Use it for independent groups or pre/post measurements in paired samples.
How to differentiate z-test from t-test?
Use the z-test when population variance is known and sample size is large (n ≥ 30). The t-test is for small samples with unknown variance.
What if population variance is unknown?
Estimate variance with a pilot sample or similar studies. Extremely low/high estimates may affect calculation accuracy.
Can I use this calculator for unequal sample sizes?
No. It assumes equal group sizes. For unequal samples, use specific formulas or power adjustments.