Effects of stress on cognitive task performance: the role of working memory capacity

Major formulations of stress are presented (Cannon, 1932; Selye, 1951-1956) both to clarify the nature of what has proved to be a familiar but vague construct, and also to provide background to the theoretical context (see Koolhaas et al., 2013, for a review). In Chapter 1, the interplay between stress and Working Memory (WM) is introduced where WM: (a) represents a ‘domain-free’ or ‘domain-general’ ability to control attention, (b) is separable from short-term memory (STM), and (c) is an important component of the cognitive architecture most affected by stress (Schoofs, Preuss, & Wolf, 2008; and Wolf & Smeets,2009). Current immediate-memory theories (e.g., cognitive interference theories; see Eysenck, Derakshan, Santos, & Calvo, 2007) are reviewed in the following chapters, chapters 2 and 3, which represent a robust approach to stress and cognition. The main focus is on anxiety––a form of stress (Mauricio, 2009)––within general populations rather than within anxious ones, and there is an emphasis on individual difference characteristics in anxiety as a disposition or trait, typically assessed by self-report scales of anxiety such as Spielberger’s State-Trait Anxiety Inventory. Two hypotheses are developed and tested using the results obtained from two separate studies, Study 1 (Chapter 4) using traditional paper and pencil measures of anxiety, and Study2 (Chapter 5) using more physiological measures, including galvanic skin conductance and heart rate variance. The first hypothesis examined the impact both of domain-general and domain-specific individual difference characteristics on performance in a complex aviation task environment (cf. Sohn & Doane, 2003). Results indicated that increase in WM capacity (WMc, i.e., domain-general attention) reduces the role of domain-specific skill variables (or incoming ability e.g., experience and training). The second hypothesis tested an interaction hypothesis to determine whether a combination of high WMc and high anxiety would predict variance in a complex aviation task environment. A criterion measure of task performance (‘flight error’) was related to the predictor measures, where flight error was reduced in high anxious individuals with high WMc. A second Study sought to further assess the interplay between the study variables, assessing whether the results of Study 1 could be replicated in a more stressful context. A stimulation shock box, together with the threat of shock, was used to ensure that “lasting affect was actually elicited” (Shackman et al., 2006, p. 42). The same hypotheses were tested again and generally confirmed the findings of Study 1, except this time in what was evidently a stressful context. Physiological measures of galvanic skin conductance and heart rate variance were also captured and confirmed the presence of the target emotion. Overall, the results suggest that although threat-based cognition (i.e., anxiety) is often associated with some underlying restriction in the functional capacity of WM, it may also, given particular conditions, be associated with recruitment of attentional control, depending on the processing resources available. More generally, the present thesis accords with theories of attention for complex cognitive tasks and is consistent with a recent review by Eysenck et al. (2007, but see also Eysenck & Derakshan, 2011).