In the context of regression, the average leverage value equals (k+1)/n. What do k and n stand for?

• A. k is the number of predictors; n is the number of participants
• B. k is the number of observations; n is the number of predictors
• C. k is the number of predictors; n is the number of observations
• D. k is the number of samples; n is the number of groups

Answer: (A)

Leverage in regression reflects how much influence a given observation has on the fitted model, and the average leverage across all observations is tied to how many parameters you estimate. The formula (k+1)/n comes from the hat matrix, where the sum of its diagonal elements equals the number of estimated parameters, p. If your model has k predictors plus an intercept, then p = k+1, so the average leverage is (k+1)/n. Here, k is the number of predictors in the model, and n is the number of observations (participants) in the dataset. The other options mix up what counts as predictors or observations, so they don’t align with this standard interpretation.