Test Anxiety in Children: What the Research Tells Us and What Assessment Design Can Do

Test anxiety is not a peripheral concern in assessment design. Research consistently finds that it affects a substantial proportion of school-age children, that its effects on performance are measurable, and that they are not simply reflections of lower underlying ability. Students with high test anxiety perform worse on assessments than their actual knowledge and skills would predict, meaning that assessments conducted without attention to anxiety produce systematically inaccurate information about academic capability.

The Cassady and Johnson research

In 2002, Jerrell Cassady and Ronald Johnson published a study in Contemporary Educational Psychology examining the relationship between cognitive test anxiety and academic performance in a sample of undergraduate students. They found that cognitive test anxiety, which they defined as the tendency to engage in worry and self-critical thinking during evaluations, was a significant independent predictor of academic performance even after controlling for measured ability.

Importantly, their research distinguished between the emotional component of test anxiety (feeling anxious, physiological arousal) and the cognitive component (worrying, intrusive thoughts, negative self-evaluation). The cognitive component was the stronger predictor of performance impairment. Students who worried about their performance during tests, rather than simply feeling nervous, showed greater interference with their ability to retrieve and apply knowledge.

This finding points toward a design principle for assessment contexts that aim to elicit accurate rather than anxiety-suppressed performance: the goal should be to reduce worry and negative self-evaluation, not simply to create a calm physical environment.

What the broader research shows

A meta-analysis by Hembree (1988), published in the Journal of Research and Development in Education, synthesised findings from over 500 studies on test anxiety in school-age children. He found that between 10 and 25 per cent of students experience test anxiety at levels that meaningfully affect their performance, with higher rates among younger children and among students from backgrounds where assessment failure carries significant social consequences.

The research also identified factors that exacerbate test anxiety: time pressure that feels insufficient, unfamiliar formats, assessment framing that emphasises high stakes, and the absence of any preparatory support for what to expect. Conversely, assessments that provide clear instructions in advance, include practice opportunities, and are framed as informational rather than definitively judgmental tend to produce lower anxiety and more accurate measurement of actual capability.

Design responses to the anxiety research

Several aspects of Evalent's assessment design reflect these research findings. Each section of the assessment opens with a short video guide that explains what the student is about to do, shows them an example, and normalises the experience of finding the tasks challenging. This addresses the unfamiliarity factor that the research identifies as a significant anxiety driver.

The assessment is designed to be completed at the student's own pace, with generous time allowances for each section rather than tight time pressure. The progression from section to section is clearly marked. Students know how many sections there are and approximately how long they will take.

The mindset section, positioned at the end of the assessment, serves partly as a decompression from the academic tasks. By asking students to reflect on how they approach learning rather than to demonstrate academic performance, it closes the assessment on a different register and may reduce the sense of having "succeeded or failed" at the end.

None of these design choices eliminate test anxiety. But the research suggests they can meaningfully reduce it, which in turn produces assessment data that more accurately reflects what students actually know and can do, rather than how well they perform under conditions of elevated anxiety.

Sources: Cassady, J. C., and Johnson, R. E. (2002). Cognitive test anxiety and academic performance. Contemporary Educational Psychology, 27(2), 270-295. Hembree, R. (1988). Correlates, causes, effects, and treatment of test anxiety. Review of Educational Research, 58(1), 47-77.

See how Evalent approaches student wellbeing throughout the assessment experience.