Which base-rate condition is associated with the best incremental validity?

Which base-rate condition is associated with the best incremental validity?

• A. Base rate is moderate
• B. Base rate is high
• C. Base rate is low
• D. Base rate is irrelevant

Answer: (A)

Incremental validity measures how much a new predictor adds to our ability to predict a criterion beyond what existing predictors already provide. Base rate, the proportion of people in the population who would meet the criterion (for example, who would be successful on the job), shapes how much extra predictive value a new measure can offer. When the base rate is moderate, there are enough successful and unsuccessful individuals to meaningfully distinguish between them. A new predictor can meaningfully improve selections, leading to a noticeable increase in correct decisions and utility. If the base rate is very high, most people would meet the criterion regardless of the new predictor, so the added predictive power yields limited gains. If the base rate is very low, true successes are rare, so even a strong predictor has less opportunity to improve the classification and the relative usefulness of the added information diminishes. Therefore, the condition associated with the best incremental validity is a moderate base rate.

Incremental validity measures how much a new predictor adds to our ability to predict a criterion beyond what existing predictors already provide. Base rate, the proportion of people in the population who would meet the criterion (for example, who would be successful on the job), shapes how much extra predictive value a new measure can offer.

When the base rate is moderate, there are enough successful and unsuccessful individuals to meaningfully distinguish between them. A new predictor can meaningfully improve selections, leading to a noticeable increase in correct decisions and utility. If the base rate is very high, most people would meet the criterion regardless of the new predictor, so the added predictive power yields limited gains. If the base rate is very low, true successes are rare, so even a strong predictor has less opportunity to improve the classification and the relative usefulness of the added information diminishes.

Therefore, the condition associated with the best incremental validity is a moderate base rate.