Programming the statistical procedures from SAS

Interrater agreement between two raters: How to handle data missing/unknown in cells

Reply
Contributor
Posts: 58

Interrater agreement between two raters: How to handle data missing/unknown in cells

Hi SAS Community,

I am looking to create an analysis plan for a study that aims to compare interrater agreement between two assessments. In theory a 2 x 2 table with Kappa stats should give me what I am looking for, but here is the catch:

Data from each test (assessment) have up to 33% missing information i.e., there is no way to correctly assign 33% cases to the appropriate cells in the 2 x 2 table

Given this limitation in data collection, is there any way to think of an analysis plan and test statistic that could handle this? My feeling is that we can only quantify % observed agreement in this case and will not be able to quantify the %chance agreement. Thoughts

Ask a Question
Discussion stats
  • 0 replies
  • 124 views
  • 0 likes
  • 1 in conversation