The adjacent category model is used when categories are ordered and somewhat equidistant and assumes that which statement is true?

• A. The predictor decreases log odds of all categories equally
• B. The predictor increases log odds of a category occurring by a fixed amount
• C. The predictor affects only a single category
• D. The model relies on linear relationship with outcome

Answer: (B)

In an adjacent-category logit model, the effect of a predictor is assumed to be the same across all neighboring category comparisons. That means a one-unit increase in the predictor shifts the log odds of being in one category versus the adjacent next category by a fixed amount, and this fixed shift applies no matter which adjacent pair you look at. In other words, the predictor changes the log odds by a constant amount across all adjacent category contrasts.

This is why the statement describing a fixed, uniform change in log odds per unit increase in the predictor best captures the model’s assumption. The other ideas don’t fit: the effect isn’t limited to a single category, nor does it imply a universal decrease across all categories, and while the model uses a linear form on the log-odds scale, saying it relies on a linear relationship with the outcome itself is not the precise point of this model.