Ordinal alpha is an alternative to Cronbach's Alpha that is more accurate with Likert-scale responses. Stata does not included baked in calculations of Ordinal Alpha. Does anyone have a workflow for calculating this? I can get the polychoric correlations (as opposed to Pearson's)  through stata but I'm not seeing a good method forward from there. 

The attached paper has a method in R, but not stat a method. 

Any help is much appreciated! 

More Simon Greenland-Smith's questions See All
Similar questions and discussions