In simple binomial validation, what should be done if outliers are found during screening?

• A. Replace outliers with more typical values
• B. Exclude outliers from the dataset
• C. Keep outliers unchanged
• D. Double their influence to test robustness

Answer: (A)

In simple binomial validation, addressing outliers by replacing them with more typical values helps prevent a single unusual observation from skewing the estimated binomial parameters, such as the probability of success. This approach preserves the dataset size and reduces the influence of potential data entry errors or measurement glitches, yielding a more stable and representative estimate of performance. Keeping outliers unchanged can distort results, and excluding them entirely reduces sample size and can bias interpretation. Doubling their influence would unrealistically amplify their impact. If the outlier is truly representative, a separate sensitivity analysis can be informative, but imputation to typical values is a reasonable screening step to maintain a reliable dataset.