Elevating Test Security in Validity Theory

Cheating and other threats to score integrity have long challenged language testing, and concerns have intensified with remote proctoring and generative artificial intelligence. Yet prevailing validity frameworks treat test security primarily as an operational matter rather than as a component of validity. We argue that high-stakes testing programs should integrate security explicitly into validity arguments. After outlining how security is treated in prominent work on validity, we propose a distinct security-related Attribution inference within argument-based validation. The Attribution inference claims that observed performance can be credibly attributed to the registered test taker’s unaided, individual effort (unless external materials or collaboration are permitted), a prerequisite for all other inferences. We articulate three warrants supporting this inference, addressing impersonation, unauthorized external assistance, and item pre-knowledge, and specify associated assumptions and potential sources of backing. We further discuss how security vulnerabilities and security measures impact the explanation, utilization, and impact inferences through construct-irrelevant variance, stakeholder trust, and fairness–privacy trade-offs. We conclude that test security evidence should be demonstrated across administration paradigms and reported routinely as part of ongoing validation, with evidentiary expectations depending on test stakes.