“…more frequent, lower-stakes quizzes.” We have a idea.
You would think that in your attempts to harness the “testing effect” you could quite reasonably fall victim to the impairment of “test anxiety,” effects. The answer is relatively simple. More frequent, lower stakes
quizzes retrieval. That – and control the spacing. Otherwise know as Successive Relearning.
Silah et al., (2021) assess “how the quantity and relative weight of assessments contribute to the effects of test anxiety on performance and metacognitive accuracy.” And it is the metacognitive accuracy that most interests me. In essence, the research explores the use of high and low stakes tests and whether these tests increase or reduce anxiety and performance to recall.
Study 1: Students took six low-stakes quizzes each worth 10% of their final grade.
Study 2: Students took two high-stakes exams each worth 40% of their final grade.
All students provided their state anxiety and predicted their scores before and after each assessment (a judgement of retrospective learning.) Students in both classes also provided their trait (overall) anxiety after the final assessment.
In both studies, students’ higher post-state anxiety appeared to be associated with worse assessment performance however pre and post-state anxiety; however decreased across the quarter in study 1 but remained constant in study 2. Furthermore, students’ metacognitive accuracy appears to be influenced by trait anxiety when taking low-stakes quizzes, while performance is related to trait anxiety when taking high-stakes exams.
The researchers reported that score calibration (a measure of metacognitive accuracy) increased throughout the academic quarter. Higher post-exam state anxiety was associated with worse assessment performance.
It is well reported that the testing has many direct and indirect benefits and researchers were right to acknowledged that “…low-stakes quizzes may prompt students to adjust their study practices.” Rather, I would suggest it was a combination of both the “stakes” and the “frequency” that was important – effectively creating a Successive Relearning sequence.
The research concludes, “less-frequent, higher-stakes assessments may trigger students’ baseline levels of anxiety,” thus diminishing the both direct and indirect benefits of the testing effect.
High stakes testing overwhelms students with anxiety and as a result, interacts with their metacognitive accuracy.Silah et al., (2021)
That students feel more pressure to perform better on these exams compared to testing environments with more-frequent, lower-stakes quizzes, as each individual assessment would not necessarily have as substantial of an impact on a student’s overall course grade comparatively.
At RememberMore we have a hunch that there may be “response measures” in additional to quantitative outcome, is it correct or not, to help inform teaching and learners self regulation and metaconigtion.
High stakes testing overwhelms students with anxiety and as a result, interacts with their metacognitive accuracy.
Let’s not stop there. Make more of the indirect benefit of Successive Relearning with these five easy ideas.
- Explain to learners why retrieval practice or the testing effect works.
- Tell learners ahead of time to expect regular testing.
- Before the test, ask learners for forecast their performance. Reward for forecast accuracy.
- Build learner agency, have learners mark their (interim) test.
- Always provide the corrective feedback.
Matthew Wemyss made a good point over on Linkedin. Having read the post:
This is a very interesting read. I have always interpreted ‘High stakes’ tests as any test that did count towards a grade, as there is a pressure there. For me ‘Low stakes’ have been just been the teacher and student/class, with the results been used for (re) teaching, not being added to a system.
I think Matthew raises a poignant point. What teachers consider “low/high” stakes is crucial to how you interpret these findings.