Abstract
Traditionally, research in language testing is conducted using quantitative methods, such as psychometrics and statistics..Since the 1980s, however, an increasing number of qualitative research methods, such as introspection and verbal protocols, have been implemented. This is based on the concern that quantitative methods are unable to discover and explain cognitive processes and factors which influence testing and assessment satisfactorily. By and !arge, two areas have since been investigated using introspective methods. Firstly, the field of test-taking strategies: Are the tests we develop able to elicit those language behaviours which we aim to measure? Secondly, the field of rating lest products such as oral and written texts: What kind of strategies and cognitive processes can be observed in raters? On which aspects of the text do raters base their assessment and why do they do so? Thus, introspective methods in the context of language test quality have two benefits: We learn about how tests fnnction and whether we can measure that which we sought to. Additionally, the findings might be used for our language behavior as learners and our assessment behaviour as teachers and raters. Both aspects are crucial for test validity. This article aims to givean overview of the state of the art in language testing with regard to introspective methods and discusses its potential and limitations.