Document Type : Original Article
Seminar für Sprachlehrforschung (Department of Foreign Language Research), Ruhr-Universität Bochum, Bochum, Germany.
This article comments on theoretical misconceptions and misuses of statistics in Khodadady & Hashemi’s (2011) paper “Validity and C-Tests: The Role of Text Authenticity”. Firstly, it is pointed out that the pertinent C-Test literature is not adequately dealt with. Then, it is argued that the authors misconstrue the notion of the C-Test when they apply the term to a single (longer) C-Test text such as their AC-Test (Authentic C-Test). Subsequently, it is shown that there are serious flaws in the data analysis and interpretation. Here, the main focus is on local item dependence, which is not taken into account, and misconceptions with regard to Cronbach’s Alpha, an issue of relevance to a wider audience.