DEVELOPING AND VALIDATING A METALINGUISTIC KNOWLEDGE TEST OF WRITTEN ACADEMIC DISCOURSE FOR UNIVERSITY UNDERGRADUATES IN PAKISTAN

Abstract

Metalinguistic knowledge plays an important role in language learning. Its evaluation involves learners’ ability to analyze their explicit knowledge of certain aspects of language. Various tests have been developed by researchers keeping in view the specific domain and the population. The present study aimed to develop and validate a Metalinguistic Knowledge Test (MKT) by focusing on a particular domain of written academic discourse. Before the test development, vocabulary knowledge of the undergraduate participants was noted using Vocabulary Size Test by Nation and Beglar (2007) as it helped in knowing the reading comprehension level of participants which was later used in the selection of the appropriate passages for the test. A test consisting of 132 MCQs was developed keeping in view the framework of Bialystok and Ryan (1985). Before test administration, initial piloting was done on 39 participants to notice any issues in the test. Then, actual data for the newly developed test was collected using convenience sampling from 385 undergraduate participants studying in 36 different universities in Pakistan. It was an untimed test, and this decision was made based on the administration of previously developed tests. After the test administration and data collection, it was analyzed to check validity and its measures, item difficulty, item discrimination, distractor effectiveness, reliability, and bias in terms of gender and subject of the participants. After performing calculations based on the participants’ responses, it was revealed that the remaining 63-item test was valid and highly reliable, resulting in a value of .93. In addition, factors like gender and subject did not independently bias the test. This test was significant as no test was developed in the past on metalinguistic knowledge of written academic discourse.