PRINCIPLES OF LANGUAGE ASSESSMENT: RELIABILITY AND VALIDITY

This article examines two foundational principles of language assessment—reliability and validity—and their implications for test design, administration, and interpretation. After situating reliability and validity within contemporary theoretical frameworks, the paper reviews major types and sources of reliability (e.g., internal consistency, inter-rater, test–retest) and validity (content, criterion-related, construct, consequential and argument-based approaches). The article emphasizes the interactive relationship between reliability and validity: reliability is necessary but not sufficient for validity. Practical strategies for enhancing both properties in language tests are discussed, including task design, rater training, sampling procedures, standardization, statistical analysis, and validation evidence collection. The article concludes with recommendations for test developers, teachers, and researchers to adopt a principled, evidence-based approach to language assessment that foregrounds consequences, fairness, and continual validation.