EZ Statistics

Effect Size

Calculator

Parameters

Results

Notes:

  • Effect size measures the magnitude of an effect independent of sample size
  • Different types of tests use different effect size measures
  • Effect sizes help in planning sample sizes for future studies
  • Guidelines for effect sizes are field-dependent
  • Consider practical significance alongside statistical significance

Learn More

Understanding Effect Size Measures

What is Effect Size?

Effect size quantifies the magnitude of differences or relationships between groups. It helps researchers understand the practical significance of their findings beyond statistical significance (p-values).

Effect Size Formulas and Examples

Formula:

d=Xˉ1Xˉ2spooledd = \frac{\bar{X}_1 - \bar{X}_2}{s_{pooled}}
spooled=(n11)s12+(n21)s22n1+n22s_{pooled} = \sqrt{\frac{(n_1-1)s_1^2 + (n_2-1)s_2^2}{n_1 + n_2 - 2}}

Example Calculation:

Given:

  • Mean1=75.3,Mean2=70.1\text{Mean}_1 = 75.3, \text{Mean}_2 = 70.1
  • s1=12.4,s2=11.8s_1 = 12.4, s_2 = 11.8
  • n1=30,n2=32n_1 = 30, n_2 = 32
spooled=29(12.42)+31(11.82)6012.1s_{pooled} = \sqrt{\frac{29(12.4^2) + 31(11.8^2)}{60}} \approx 12.1
d=75.370.112.10.43d = \frac{75.3 - 70.1}{12.1} \approx 0.43

Result: Small-to-medium effect size

Available Effect Size Measures

Mean Difference (Cohen's d)

  • Small effect: 0.2
  • Medium effect: 0.5
  • Large effect: 0.8
  • Used for comparing two independent groups

Paired Difference (Cohen's d)

  • Uses same scale as regular Cohen's d
  • Accounts for correlation between measures
  • Typically larger than independent d
  • Used for before-after or matched pairs

Proportions Difference

  • Small effect: 0.1
  • Medium effect: 0.3
  • Large effect: 0.5
  • Used for comparing success rates between groups

ANOVA (Eta-squared)

  • Small effect: 0.01
  • Medium effect: 0.06
  • Large effect: 0.14
  • Used for comparing multiple groups

Calculation Examples

Independent Groups Example

Treatment vs Control Group:

  • • Treatment: Mean = 75, SD = 10, n = 30
  • • Control: Mean = 70, SD = 10, n = 30
  • • Cohen's d = 0.5 (Medium effect)
  • • Interpretation: Moderate treatment effectiveness

Paired Design Example

Before vs After Intervention:

  • • Before: Mean = 20, SD = 5
  • • After: Mean = 25, SD = 5
  • • Correlation = 0.7, n = 25
  • • Cohen's d = 1.2 (Large effect)

Proportions Example

Success Rates Comparison:

  • • Group 1: 40 successes out of 100
  • • Group 2: 60 successes out of 100
  • • Effect size = 0.4 (Medium-large)
  • • Substantial difference in success rates

ANOVA Example

Comparing Multiple Groups:

  • • F-value = 10.5
  • • df between = 2, df total = 100
  • • η² = 0.17 (Large effect)
  • • Strong evidence of group differences

Interpretation Guidelines

Important Considerations:

  • Effect sizes should be interpreted within your field's context
  • Small effects can be meaningful in some research areas
  • Consider practical implications alongside numerical values
  • Use effect sizes to plan sample sizes for future studies
  • Report effect sizes alongside confidence intervals when possible

Related Calculators

Sample Size Calculator

Power Analysis Calculator

Two Sample Paired T-Test Calculator

One-Way ANOVA Calculator

Help us improve

Found an error or have a suggestion? Let us know!