A comparison of the logistic RAMP for the Discard- and Unobserved-mortality studies (Fig. 3) with Score as the only predictor indicated that the probabilities of mortality were lower for all Scores for the Discard-mortality study. This is further highlighted by the fact that Scores two through six did not have overlapping 95% Confidence Intervals, and that the Score at which a crab has a 50% or greater probability of mortality (Score50) is Score-three for the Discard-mortality study, and Score-two (1.9) for the Unobserved-mortality study. Moreover, logistic model selection on the combined datasets determined that the most parsimonious model included Score, binned-width, Study, and an interaction between Study and Score as predictors (Table 2). The significant interaction indicates that Study has a measurable influence on the relationship between Score and probability of mortality, after accounting for binned-width. Results from the logistic regression analyses indicate significant differences between RAMPs for the Discard- and Unobserved-mortality studies.