Maximum likelihood estimation and likelihood ratio test for square tables with missing data.

W. Shih
DOI: https://doi.org/10.1002/SIM.4780060111
Statistics in Medicine
Abstract:Missing data complicate the analysis of paired categorical data. This paper considers with the use of the EM algorithm, the maximum likelihood estimation and likelihood ratio test for incomplete square tables with missing data. An example involving the susceptibility of micro-organisms to antimicrobial drugs illustrates the procedure.
What problem does this paper attempt to address?