Proportional odds (ordinal) regression for likert scales in SPSS

Introduction to regression with ordinal response variable (eg. DV is a likert scale)
0:44 Example: proportional odds model
1st pass of the output:
4:00 Model fit stats, parallel lines test;
6:32 Interpretation of the "slope" parameters (SPSS, Stata, R v SAS)
2nd pass:
12:05 choice of link function affects the interpretation of the parameters; models with covariates, interactions;
14:23 Goodness of fit test: what if there is a high proportion of empty cells?
16:20 What if parallel assumption is not satisfied?
16:48 Where I am talking about binary logistic for B and C, I meant to say run a binary logistic for each of B and C choosing one rating value to divide the ratings into 2 groups (low, and high), then you can get the odds ratio by taking the ratio of the odds from each of B and C.