Clinical research · Study planning

Sample size calculator

Estimate the minimum enrollment needed to detect a planned treatment effect with your selected statistical power.

Parallel groups · crossover · repeated measures

Study assumptions

Choose your design and outcome, then enter the expected values from prior evidence or a pilot study.

Parallel design · number of samples

Two samples · independent groupsOne sample vs. known population

Primary endpoint

Dichotomous (yes / no)Continuous (means)

Anticipated incidence

Group 1 proportion (%)

Group 2 proportion (%)

Enrollment ratio Group 2 : Group 1Use 1 for equal group sizes, or 2 for twice as many participants in group 2.

Type I / II error rates

Alpha (α) significance

Power (1 − β)

Two-sided significance test · commonly α = 0.05 and power = 0.80–0.90.

Your result will appear hereEnter study assumptions and select Calculate sample size.

Minimum required enrollment

Sample size

—

Calculation:

Definitions & design guide

Parallel design · two samples

A parallel study assigns different participants to separate groups. Choose two independent samples to compare two groups, or one sample to compare a cohort with a known population value.

Key statistical terms

Sample size: The number of participants with usable outcomes needed for the planned analysis. Add an allowance for expected dropouts or missing outcomes.
Alpha (α): The planned false-positive rate for the primary test.
Power (1 − β): The chance of detecting the planned effect if it is present.
Pooled SD: A combined estimate of variability across treatment groups. For the multi-treatment crossover calculation, it is the root mean of the squared treatment SDs.
Within-person correlation: How similarly the same person’s outcomes move across treatments or time points. It changes the variance of paired differences.
Carryover: An effect of an earlier treatment that remains during a later treatment period. A crossover plan should account for washout and carryover in its design and analysis.

These calculators are planning estimates. They do not adjust for dropout, clustering, interim looks, or all analysis-specific corrections. Confirm assumptions and analysis methods with a statistician before finalizing a protocol.

Sample size calculator

Study assumptions

Parallel design · number of samples

Primary endpoint

Anticipated incidence

Anticipated incidence

Anticipated means

Anticipated mean

Cross-over structure

Anticipated treatment means and SDs

Expected discordant treatment outcomes

Repeated-measures structure

Anticipated means and SDs at each repeat

Expected baseline-to-follow-up transitions

Equivalence test assumptions

Type I / II error rates