Logistic Regression Sample Size Calculator
Below is a table summarizing key aspects related to sample size for logistic regression.
| Aspect | Description |
|---|---|
| Effect Size | A measure of the strength of the relationship between the independent and dependent variables. |
| Significance Level (α) | The probability of rejecting the null hypothesis when it is true, commonly set at 0.05. |
| Power (1 – β) | The probability of correctly rejecting the null hypothesis; commonly set at 0.80 or 0.90. |
| Number of Predictors | The number of independent variables included in the model, which influences sample size. |
| Event Rate | The proportion of the outcome occurring in the sample; can affect the required sample size. |
| Sample Size Formula | Sample size (N) can be estimated using formulas such as: |
| N = (Zα/2 + Zβ)² × (p(1-p)) / d², where: | |
| Zα/2 = critical value for significance level | |
| Zβ = critical value for power | |
| p = estimated event rate | |
| d = margin of error (difference in proportions of the event rate) | |
| Sample Size Calculators | Online tools and software (e.g., G*Power, R packages) can help estimate sample sizes. |
| Guidelines | As a rule of thumb, at least 10-15 events per predictor variable is often recommended. |
| Final Considerations | Consider potential loss to follow-up or incomplete data when determining the final sample size. |
Example Calculation
- For a logistic regression model with:
- Two predictors
- An expected event rate of 20%
- A significance level of 0.05
- A power of 0.80
Using a sample size calculator or software, you would input these parameters to find the necessary sample size for your study.
Resources
- Online calculators and software like G*Power or R’s
pwrpackage can assist in calculating sample size for logistic regression based on the parameters you specify.
This table provides a concise overview of important factors to consider when determining the sample size for logistic regression analyses.