Date of Award


Degree Type


Degree Name

Bachelor of Arts Honours


School of Psychology


Faculty of Computing, Health and Science

First Advisor

Dr Ken Robinson


This paper is a critical review of the research on the relative contribution of valence and arousal to the effect of emotion on performance. It is well accepted that emotion influences aspects of cognitive performance, but there are inconsistent results concerning the relative salience of valence and arousal. Some authors support the idea that valence, rather than arousal, is the primary contributor to this effect of emotion on performance. This review analysed the methodology used in studies supporting the primary role of valence, and this analysis revealed that the two dimensions may have been confounded. The literature suggested that arousal appears to be salient when the two dimensions are controlled. Furthermore, when an individual differences perspective was also integrated, this showed that both arousal and valence have particular relevance as they are differentially perceived by each individual. It was concluded that there is value in exploring an individual difference perspective, and that this may also account for the mixed results in the literature. This explorative study tested the relative contribution of valence and arousal to ratings of emotionality, and the potential for an individual differences approach to increase the explained variance in comparison to a group based approach was assessed. Thirty female participants were recruited from the psychology student population at Edith Cowan University. A ratings task was used, whereby participants had to rate a set of affective pictures in terms of overall emotionality, valence and arousal. The arousal hypothesis would predict that the highly arousing pictures, despite their valence, would have the biggest effect on performance. The Friedman and Wilcoxon non-parametric tests were used to analyse the data. From phase I, results indicated that both valence and arousal had a significant effect on emotionality rating p < .05. However, effect sizes revealed that 'arousal could explain 11.6% of the variance, in comparison to the 1.4% explained by valence. Phase II results revealed that an added 15% of the variance could be explained when individual differences were accounted for. These results are consistent with the arousal hypothesis, and highlight the power of an individual based design. There are many potential projects for further research, for instance, adapting the present design to investigate stimulus-limited effects on emotional perception which would then allow investigation of the relative contributions of valence and arousal across a time course.