Logistic Regression Assessment Checklist

Model Setup

• [ ] Dependent variable is binary.
• [ ] Binary coding is documented.
• [ ] Predictors are correctly scaled or coded.
• [ ] Sample size is sufficient for model complexity.
• [ ] Observations are independent.
• [ ] Intercept is included unless justified otherwise.

Model Fit

• [ ] Log-likelihood reported.
• [ ] Deviance reported.
• [ ] AIC reported.
• [ ] BIC reported.
• [ ] McFadden's R-square reported.
• [ ] Cox and Snell's R-square reported.
• [ ] Nagelkerke's R-square reported.

Predictive Accuracy

• [ ] Confusion matrix exported.
• [ ] Group 0 classification accuracy reported.
• [ ] Group 1 classification accuracy reported.
• [ ] Overall classification accuracy reported.
• [ ] Threshold/cutoff documented when applicable.

Coefficients

• [ ] Logistic coefficients reported.
• [ ] Wald tests reported.
• [ ] p-values reported.
• [ ] Odds ratios reported or calculated.
• [ ] Coefficients are not interpreted as direct probability changes.

Assumptions and Limitations

• [ ] No perfect multicollinearity checked.
• [ ] Linearity with log-odds considered.
• [ ] Influential observations considered.
• [ ] Small-sample limitations stated when applicable.
• [ ] Causal claims avoided unless design supports them.