Test anxiety : theory, measurement and intervention

Test Anxiety (TA) is an emotional, physiological, behavioural and cognitive construct that has been found to play a role in student underachievement and reduced wellbeing. While existing interventions have shown some success when addressing TA and its consequences, they are often lengthy and delivered face-to-face. These qualities can present as a barrier for many university students, contributing to fewer students seeking support and greater treatment attrition. The online delivery of intervention has been found to have greater appeal amongst university students (in addition to other advantages) over face-to-face delivery. Therefore, a need for short intervention programmes delivered online exists. Consequently, this thesis set out to identify a brief, efficacious TA intervention for online delivery. Initial stages of this thesis consisted of a published review, further developed upon by an unpublished literature review. These reviews were designed to (a) develop a comprehensive theoretical framework for TA, (b) identify an effective and brief online intervention, and (c) establish a contemporary instrument with acceptable psychometric properties to measure TA. Following these aims, a theoretical model was developed implicating certain beliefs about thoughts and feelings as suitable targets for intervention. Amongst the interventions critically appraised, Cognitive Bias Modification for Interpretation (CBM-I) was selected for investigation in this thesis due to it most effectively addressing the established criteria of a brief and effective intervention that could be delivered online. Lastly, the 17-item version of the German Test Anxiety Inventory (TAI-G) demonstrated the most favourable psychometric properties out of the measures reviewed; however this had not been confirmed in an Australian university population. In response to the literature reviews, two studies were conducted (and published) using Australian university students in an attempt to assess and potentially improve upon the psychometric properties of the TAI-G. As hypothesised, the initial study (N = 224) supported the psychometric properties of the 17-item TAI-G and confirmed the use of this 17-item version over other forms. Study 2 (N = 473) expanded upon the first paper by addressing criticisms of the TAI-G by developing a revised Lack of Confidence subscale that had an item orientation consistent with the remaining subscales. Once again, and as predicted, the 17-item TAI-G proved psychometrically superior over other models tested and the revised Lack of Confidence subscale was shown to improve the interscale relationships and structural validity of the TAI-G. As a result, the 17-item TAI-G with revised Lack of Confidence subscale was chosen to measure changes in TA for the concluding ntervention study. The concluding intervention study (currently under review) finalised the aims of this project by initially examining the theoretical relationships between beliefs and TA. As expected, significant relationships were observed for participants (N = 199) between TA and the key beliefs of confidence for learning and memory, danger and uncontrollability of thoughts/ emotion, and the need to control emotion. Following this, an intervention group (n = 20) engaged in an online CBM-I intervention targeting these beliefs, with these findings compared to a control group (n = 19). In partial support of hypotheses, negative beliefs about worry showed a downward trend and beliefs for the need to control worry were significantly reduced after one week follow-up. Overall, findings support the 17-item TAI-G with revised Lack of Confidence subscale as a valid and reliable measure of TA, though the use of reverse coded items in psychometric measures was brought into question given the loss of validity and reliability observed. Further, findings support the relationship between TA and certain beliefs, while CBM-I was found to show some promise in reducing specific metacognition that could ultimately lead to a reduction in TA over multiple administrations. Recommendations for future research include developing a more engaging and aesthetically appealing CBM-I procedure to be tested in controlled laboratory conditions