Test anxiety and physiological arousal : a systematic review

Abstract

Theoretical background and objectives. Test anxiety has extensively been investigated since the early 1960s. To date there are more than 2000 published studies on this construct, with many of them showing a detrimental effect on well-being, learning behavior and academic performance (e.g., Cassady & Johnson, 2002; Zeidner, 2014). In these studies test anxiety is usually conceptualized as constituting multiple components (e.g., motivational and physiological components) that are empirically distinct, albeit correlated (Zeidner, 2014). So far, test anxiety and its components have typically been assessed using self-report measures, which can be biased for example by subjective beliefs (Robinson & Clore, 2002). Heightened physiological arousal, a major component of the test anxiety response, can be assessed through physiological measures, which opens the possibility of objectively and continuously gaging individuals’ test anxiety in real-life testing situations (Harley, 2015; Shumann & Scherer, 2014). Although, theoretically one would assume high convergence between self-report and physiological measures, empirical evidence is surprisingly rare and findings are mixed (e.g., Mauss, Levenson, McCarter, Wilhelm & Gross, 2005; Mauss & Robinson, 2009). Therefore, the aim of the current review is to yield a more coherent picture of the relationship between self-report and physiological measures by systematically investigating whether higher self-reported test anxiety is associated with objectively measurable higher physiological arousal. Methods. To this end, a comprehensive literature search was conducted in PsycINFO, PubMed, and ERIC databases to identify studies addressing the relationship between self-reported test anxiety and physiological arousal applying cardiovascular measures, electrodermal measures or cortisol sampling. First, studies were preselected according to the search terms “test anxiety” AND “[PHYSIOLOGICAL MEASURE e.g., heart rate]”, which had to appear either in the title, the keywords or the abstract of the study. The next step comprised a fine grained screening of these studies, after which only published peer-reviewed journal articles in English, that used both a self-report measure of test anxiety and at least one kind of physiological measure, were included for a detailed review. References of the identified publications were additionally screened for further studies meeting the specified criteria. Results and significance. The literature search resulted in the initial identification of 189 studies with publication dates ranging from 1957 to 2016. Twenty-four studies met the specified criteria and were therefore included in the present review. Results revealed that 17 studies found significant positive correlations between self-reported test anxiety levels and at least one measure of physiological arousal (mean correlation: r = .35; range: 0.21 to 0.49). Six of the reviewed studies did not find significant correlations and in one study low anxious students were found to be more aroused than high anxious students. Taken together, these results are in line with theoretical concepts and support the assumption that higher self-reported test anxiety is associated with higher physiological arousal. Moderating variables of the relation between these two measures and methodological issues of the reviewed studies will be discussed, along with methodological and instructional implications for the assessment of physiological data in future laboratory and classroom-based research.