Hassan Ijaz

Ai, Web & Design
← Back to all topics
Statistical InferenceTopic 13 of 58

Sample size determination

Power calculator with visual effect size demonstrations showing required samples for statistical significance

Concept Overview

Sample size determination is crucial for ensuring statistical power while maintaining efficiency. Too small samples may miss important effects, while unnecessarily large samples waste resources.

Power Analysis Components

Effect Size

Magnitude of difference we want to detect

Larger effects easier to detect

Significance Level (α)

Probability of Type I error

Typically 0.05

Power (1-β)

Probability of detecting true effect

Typically 0.80 or 0.90

Sample Size (n)

Number of observations needed

What we're trying to determine

Sample Size Formulas

One Sample Mean

n = (z_α/2 + z_β)² × σ² / δ²

δ = effect size, σ = standard deviation

Two Sample Means

n = 2(z_α/2 + z_β)² × σ² / (μ₁ - μ₂)²

For equal group sizes

Proportions

n = (z_α/2 + z_β)² × [p₁(1-p₁) + p₂(1-p₂)] / (p₁ - p₂)²

For comparing two proportions

Practical Considerations

  • Pilot Studies: Use preliminary data to estimate parameters
  • Multiple Outcomes: Adjust for multiple comparisons
  • Dropout Rates: Inflate sample size to account for attrition
  • Cost Constraints: Balance statistical power with resources

Key Insight: Sample size requirements grow with the square of precision needed. To halve the margin of error, you need 4 times the sample size!

The power calculator below shows visual demonstrations of effect sizes and required sample sizes. Adjust the parameters to see how they influence statistical significance.

Interactive Visualization

Loading interactive visualization...

Power calculator with visual effect size demonstrations showing required samples for statistical significance