Bootstrap Logistic Regression Calculator

Estimate event probability, odds ratios, and classification risk. Inspect coefficient impact with resampling-based interval summaries. Make binary outcome analysis clearer for practical, data-driven decisions.

Calculator Inputs

Use the responsive grid below. Large screens show three columns, medium screens show two, and small screens show one.

Single-page white layout
Predictor Label Coefficient Observed Value Bootstrap SE

Formula Used

Linear predictor: z = β₀ + Σ(βᵢ × xᵢ)

Predicted probability: p = 1 / (1 + e-z)

Predicted odds: odds = ez

Odds ratio for each predictor: ORᵢ = eβᵢ

Coefficient interval: βᵢ ± zα/2 × SEboot

OR interval: exponentiate the coefficient interval limits.

Bootstrap probability interval: simulate repeated coefficient draws, convert each draw into a probability, then read percentile limits.

Classification rule: predict class 1 when p ≥ threshold; otherwise predict class 0.

How to Use This Calculator

  1. Enter the model intercept and its bootstrap standard error from your fitted logistic model.
  2. Fill in each predictor label, its coefficient, the observed predictor value, and the bootstrap standard error.
  3. Set the sample size, expected base event rate, desired confidence level, and decision threshold.
  4. Optionally provide the actual binary outcome to unlock scoring metrics such as log-likelihood, pseudo R², AIC, BIC, and Brier score.
  5. Press the button to generate point estimates, interval summaries, stability statistics, and exportable result tables.

Example Data Table

Outcome Age BMI Smoker Exercise Hours Family History Cholesterol Ratio
15231.01314.8
04426.20503.4
15933.81115.2
03824.90603.1
16130.51214.9
04727.10403.6

Frequently Asked Questions

1. What does this calculator estimate?

It estimates the linear predictor, event probability, odds, predictor odds ratios, interval limits, and resampling-based stability for a binary logistic regression scenario.

2. Why are bootstrap standard errors useful?

Bootstrap standard errors reflect uncertainty from repeated resampling. They help you judge whether each coefficient and predicted probability remain stable across changing samples.

3. What does the predicted class mean?

The predicted class is the outcome assigned after comparing the estimated probability with your chosen threshold. Above the threshold predicts 1; below predicts 0.

4. How should I interpret odds ratios?

An odds ratio above one raises event odds as the predictor increases by one unit. An odds ratio below one lowers event odds.

5. What is classification stability?

Classification stability shows how often the model keeps the same class during repeated resampling draws. Higher values suggest more reliable threshold-based decisions.

6. Why do I need a base event rate?

The base event rate provides a null reference for comparing your fitted prediction against a simpler benchmark. It helps compute pseudo R² and related diagnostics.

7. Can I use fewer than six predictors?

Yes. Set unused predictor coefficients, values, and bootstrap standard errors to zero, or replace their labels with placeholders to keep the layout consistent.

8. Does this replace full model fitting software?

No. It is best for quick scenario analysis, interpretation, and reporting. Use dedicated statistical software when you need full estimation, diagnostics, and validation.

Related Calculators

permutation logistic regressionpermutation confidence interval

Important Note: All the Calculators listed in this site are for educational purpose only and we do not guarentee the accuracy of results. Please do consult with other sources as well.