On tests of quasi-independence in psychological research
Abstract
Presents a method for analyzing rate agreement matrices and other contingency tables common in psychological data. The method, unlike standard chi-square analyses, allows for different treatment of errors and correct responses.