It’s fairly common for developers of language proficiency tests to say that test results are only valid for a certain period of time from the test date. But why is that the case?
The question is more complicated than it might seem at first! Understanding the validity period of language test results involves (at least) four factors: stakes of test use, change in true proficiency, score comparability, and stewardship of test results.
Allow us to explain…
Test stakes are high
Saying that a test is “high-stakes” means that decisions based on its results will seriously impact the test taker—and will be difficult to reverse. When the stakes are high, there's less room for error in decision-making.
The DET is used for high-stakes purposes, primarily admitting students to colleges and universities around the world. At these institutions, decision-makers rely on test results to accurately judge whether someone meets the applicable language proficiency criteria. So, they must be confident that a test taker's results don’t overestimate their current proficiency.
The more time that has passed since the test was taken, the greater the possibility that someone’s language proficiency has changed (for better or for worse), which brings us to the next factor…
How to lose a language in 10 days (give or take a few)
Anyone who has studied a language in a class but never used that language post-graduation (adios, Spanish!) knows that it is certainly possible to forget a lot of language over time. Experts call this process of losing proficiency language attrition.
But whether it's been five years or fifteen years since you studied a language in school, you likely still remember simple words for things like “hello” and “goodbye.” This indicates that while much can be lost, some knowledge is effectively permanent.
Research has shown that even if people haven’t used a language in decades, they can still recognize and understand a surprising amount of it. This seems especially true the more proficient someone has become in the language they were learning, suggesting people may forget language at different rates once they stop using it.
However, while receptive comprehension (the ability to read and listen to a language) seems to persist longer, productive ability (writing and speaking) declines more quickly. This can lead to the sensation of feeling “rusty” in a language: conversation is slow and halting, because you can’t quite recall the words or grammar patterns you want to use. If you’ve lost the ability to recall words needed in conversations or writing, then your proficiency has arguably declined, whether or not you’d still understand those same words if you heard or read them.
All language learners have their own mix of strengths and weaknesses, but ultimately, test scores are used to understand someone’s communicative competence—that is one’s ability to do things with the language. So from a theoretical perspective, an expiration date on language test results is warranted to prevent overestimation of a test taker’s current/near-future language proficiency.
Of course, it is also quite possible that someone’s proficiency has increased since taking a test, which is usually not a concern for test results users (what’s not to like about more proficiency?!). Often, universities don’t require recent test results from international applicants who have already earned a degree from an English-medium university, as this accomplishment provides strong evidence someone has maintained —if not improved—their proficiency.
Which score means more?
All languages change over time; no one was talking about texting and smartphones 40 years ago, for example. So a good language test must also change at least once in a while to keep pace with the language itself and the contexts in which it is used.
Technological advancements also present new opportunities to improve the measurement of language proficiency, for example, by enabling innovative assessment tasks and different administration formats, such as paper- and computer-based testing.
When a test changes, it is important to conduct comparability research. This includes research to demonstrate score comparability, meaning that scores from two test versions have the same meaning.
The DET is committed to annual test updates that incorporate new developments in assessment science, artificial intelligence, and technology. When these updates are made, such as adding new item types, the DET’s assessment scientists use operational test data and psychometric methods to ensure that the distribution of test scores stays the same, and that individual test takers would receive the same score on both versions (within the confidence interval for the test scores).
Why test score stewardship matters
The final factor in language test validity periods is stewardship of scores. When a testing organization says that test results are valid, that means they are endorsing the results’ legitimacy. In the case of the DET, this endorsement includes answering any questions an institution might have about a particular test taker’s results, which can involve reviewing the recorded test session and performing additional analyses.
Providing this level of support for institutions that use our test results requires considerable time and resources; the costs of offering it in perpetuity for each test result would be astronomical, especially considering the security measures necessary to maintain data protection standards.
Keeping the price of the DET as low as possible is central to the DET’s mission to lower barriers to education, and since the vast majority of people who take the DET plan to use their scores within the next year or so, the added cost and complexity of stewarding scores forever is not warranted.
Nothing lasts forever!
While language attrition can vary widely among individuals due to many factors, setting an expiration date on test results is practical. This approach ensures that all test takers are evaluated under a uniform policy.
Continued research is essential, and it’s likely we’ll see further evolution in the language testing industry’s policies regarding the validity period of test results, just as we’ll continue to see evolution in the language itself!
To learn more about the research that goes into developing the Duolingo English Test, check out our publications!