Use an R function such as the following to compute the logits of the cumulative proportions.

sf ← function(y)

c ( ' Y ≥ 1 ' =qlogis (mean(y ≥ 1)),

' Y ≥ 2 ' =qlogis (mean(y ≥ 2)),

' Y ≥ 3 ' =qlogis (mean(y ≥ 3)),

' Y ≥ 4 ' =qlogis (mean(y ≥ 4)))

As the Y = 3 category is rare, it may be even better to omit the Y ≥ 4 column above, as was done in Section 13.3.9 and Figure 13.1. For each predictor pick two rows of the summary table having reasonable sample sizes, and take the difference between the two rows. Comment on the validity of the proportional odds assumption by assessing how constant the row differences are across columns. Note: constant differences in log odds (logits) mean constant ratios of odds or constant relative effects of the predictor across outcome levels.