Analyzing Max-Diff Using the Rank-Ordered Logit Models With Ties

From Survey Analysis
Jump to: navigation, search

The Rank-Ordered Logit Model With Ties is a model specifically designed for analyzing rankings that contain ties (i.e., partial rankings). Max-Diff experiments can be viewed as partial rankings. Consider an experiment where a respondent is shown five options: A, B, C, D and E. If the respondent chooses C as Best and E as worst this can be viewed as a partial ranking with C given the highest rank, A, B and D in equal second place and E ranked last (i.e., C > A = B = D > E).

Worked examples using the rank-ordered logit model with ties are presented in:

See also

A more up-to-date version of this content is on

Personal tools