Extensive evidence linking exercise with stress reduction has prompted many organisations to implement exercise stress management interventions in order to combat employee stress. These interventions however, have generally yielded low levels of effectiveness. Determining factors that can enhance the effectiveness of exercise stress management interventions is important for organisations implementing such interventions. The main purpose of the present study was to investigate the role that outcome expectancy has in the relationship between exercise and stress. Stress perceptions and exercise behaviours were assessed in 54 university students over the 4 weeks leading up to final examinations. Outcome expectancy was assessed once during this period. Heart rate was also assessed in 20 students on two occasions, including the exam. In contrast to the majority of research linking exercise with lower stress, individuals who exercised more than their own average during the study period had higher levels of exam stress over time, whilst variance in exercise levels from the group average was not significantly related to exam stress over time. Conversely, there was a significant difference in physiological exam stress (heart rate) between high and low exercise groups overtime for the overall heart rate average, but not the sleep or exam period heart rate averages. More specifically, it was found that the low exercise group had a stronger negative physiological reaction to the exam overall. No significant moderation effects of outcome expectancy on the relationship between exercise and stress were found. The results indicate that exercise is related to both self-reported and physiological indicators of stress, and that exercise at different levels (within-person and between-person) have differing effects on exam stress.

Test anxiety is an individual personality trait, which results in elevated state anxiety in situations of performance evaluation. For school-age children, high-stakes examinations occurring at the culmination of programmes of study are where they frequently experience such evaluation. Alongside its impact on an individual's wellbeing, heightened test anxiety has been reliably linked to deficits in performance on examinations and assessments. Attentional bias has been shown to be an aspect of many forms of anxiety and is considered to have role in the maintenance of state anxiety, though the mechanisms underlying this are not fully clear. However, Attentional Control Theory (Eysenck, Derakshan, Santos, & Calvo, 2007) implicates preferential allocation of attention to threat in its explanation of performance deficits associated with test anxiety. The presence of attentional bias in test anxiety appears theoretically plausible with some empirical support (e.g. Putwain, Langdale, Woods and Nicholson, 2011); however, its reliability is under question. This study aims to investigate the presence of attentional bias in test anxiety, with a view to further understanding its underlying mechanisms and informing the development of interventions to ameliorate its effects. To ensure ecological validity, this study was conducted in schools and colleges, with a sample of 16-18-year olds following high-stakes programmes of study. Full investigation of test anxiety requires individuals to experience heightened state anxiety through performance evaluation threat; hence, the Trier Social Stress Test (TSST) was modified to make it applicable to this context and population. This study was conducted in two experimental phases, both of which adopted a mixed methodological approach to provide quantitative and qualitative data. The preliminary phase evaluated the materials and anxiety manipulation protocols. The main phase employed the modified-TSST in collaboration with a dot-probe task to investigate participants' attentional bias when under high performance evaluation threat. No patterns of attentional bias were uncovered to indicate a consistent relationship to either trait test anxiety or attentional control. However, there was a level of congruence between how some individuals describe themselves in evaluative situations and the attentional bias they displayed. Further investigation employing mixed methodological approaches such as Single Case Experimental Design is recommended to identify and address attentional bias in test anxiety.