First, it is possible to interpret only the sign and the significance of the coefficients. This is a particularly useful way to compare ordered choice regression outcomes to linear regression outcomes. The drawback is that the sign conveys little information. Second, one can compute marginal probabilities (MP), which are defined as the derivative of the probability function.Afterward, the average over all individual marginal probabilities is computed — called the average partial effect. Third, it is possible to compute the marginal probabilities evaluated at a specific value. Commonly, the sample-average is used. Hence it is called the partial effect at the average. The second and third interpretation are similar and share a common drawback.For ordered choice models, the marginal probabilities must be computed for every possible value that the dependent variable can take on. Thus, the interpretation can become very cumbersome.