Cognition, Motivation, Coping and Learning Strategies in High, Mid and Low Test Anxious Undergraduates in a UK University: The Importance of Self-Related Cognitions

Despite increasing interest in test anxiety and its effects, few studies have examined UK students, notably not undergraduates. Although, current models of test anxiety are rich and diverse, drawing upon evidence from different cultures and education systems, it is usually from research outside the UK, which that is subsequently generalised to UK students. In addition, many studies report contrasting findings, with little agreement about which of the possible factors are most important. The current programme of research addressed these issues by investigating collectively the major factors already shown to be relevant in relation to a sample of UK undergraduates.

Initially, the important cognitive and motivational factors, together with a range of learning strategies were investigated, focussing upon whether test anxiety is more influenced by the cognitive interference caused by worry or by ineffective study skills, the two major perspectives on test anxiety. Results showed conclusively that study skills were irrelevant but there were other significant differences, notably extrinsic motivation, perfectionism and coping. Specifically, high test anxious students showed significantly higher levels of self-doubt and worry about making mistakes despite trying harder to cope with the exams, and using more approach than avoidance coping. They also showed higher aspirations but lower expectations of success than low test anxious students. These results support Cognitive models and Self Worth models of test anxiety, rather than any models involving study skills.

Contrary to many studies, a second study showed that test anxiety had no significant effect upon actual exam or coursework results in these students, providing no support for the common conception of test anxiety being responsible for performance deficits, suggesting that it is clearly unwise to generalise so widely from other studies.

With the earlier results emphasising self-doubts and yet showing no performance deficits, the effects from ego-involving instructions and test anxiety upon performance and performance estimates were examined experimentally. Whilst test anxiety led to increased state anxiety on a contrived test, performance was again unaffected, confirming that these students do not conform to existing models. However, ego-involving instructions did diminish performance for all students, leaving state anxiety unaffected. Performance deficits were from ego threat rather than test anxiety, although test anxiety was responsible for increased state anxiety at the time of the test. In addition, students’ expectations differed. Low test anxious students were overly optimistic about their test result, whereas high test anxious students were realistic. The optimism of the low test anxious students seemed to mitigate the effects of test anxiety although they may have been disappointed later with their actual results. Here, test anxiety seems to have more implications for psychological well being than for exam outcomes and shows it is important to consider how these students might best be supported. Also important is how students are motivated to study. Ego involving instructions quite often form the basis of arguments about why they should work hard, yet these results suggest they might be having an opposite effect. Since issues of self worth were clearly emerging, a final experimental study focused upon the worry component of test anxiety; the source of any cognitive interference. Results showed that cognitive inference was experienced by high test anxious students but was related not to the test itself or to ego threats coming from others. The interference originated from self-related ego threatening cognitions. There was indeed a cognitive bias, but it emanated from concerns over threats to self worth.

Overall, results demonstrate that test anxiety in UK undergraduates does indeed exist but it originates more from considerations of self worth and the extent to which this can interfere with cognitions. This offers support mainly for models of Self Worth and the Information Processing Model. There was no support for any Study Skills models or consequently the integrated models. It is clear that considerable differences in test anxiety exist among students and there seems to be a need for more specific models of test anxiety giving particular attention to differences in education systems and cultures. In addition, test anxiety did not diminish performance but results do raise concerns for the students’ psychological well being at exam time and possibly after.