The Concise Encyclopedia of Applied Linguistics

VValidation of Language Assessments

JONATHAN SCHMIDGALL AND XIAOMING XI

Validation has been traditionally defined as the process of determining the extent to which an assessment measures what it was designed to measure. The hypothesized language knowledge, skills, or abilities that an assessment is designed to measure is referred to as the construct. The construct of a language assessment may be more narrowly defined, such as “the vocabulary knowledge pilots need to communicate with air traffic controllers.” Often the construct is more broadly defined, such as “the language knowledge and skills needed to participate in English‐medium postsecondary education.” The construct definition provides the basis for making meaningful interpretations about language knowledge or abilities based on test scores. Thus, the process of validation involves conducting research and collecting evidence to determine whether an assessment is primarily measuring the defined construct versus some other knowledge or skills, or whether other factors that may be irrelevant and unintended—like familiarity with test taking—may be involved. In other words, validation evaluates whether test scores support the intended interpretation about language knowledge or abilities.

For most testing researchers today, the concept of validation has expanded beyond this core definition to include the extent to which an assessment is used as intended and facilitates intended outcomes. In this view, validation is also conducted ...