Über die Charakteristika der Leistung in Konzentrationstests: Prozesskomponenten, Mechanismen und Übungseffekte

Sustained attention tests have been in use for more than a hundred years and they are applied in various psychological disciplines from traffic to business to neuropsychology (Schmidt-Atzert & Amelang, 2012). Typically, these tests consist of many homogenous and simultaneously presented stimuli and the test-taker is required to continuously apply a simple rule to them (like finding targets or assessing whether two single-digit numbers add up to a third) until the test is over (Westhoff & Hagemeister, 2005). A large body of correlational validity evidence supports sustained attention tests’ convergent (there is evidence for a common factor; e.g. Schmidt-Atzert et al., 2006) and discriminant validity (correlations with higher cognitive abilities are low to moderate; e.g. Krumm et al., 2009). However, some research questions are difficult to address using only correlational approaches to validity. For example, until today, there is no generic process model of performance in sustained attention tests and the large practice effects in these tests are well confirmed, but poorly understood. Therefore, a more in-depth analysis of the processes involved in sustained attention tests requires an experimental approach. Along these lines, the aim of the present thesis was to investigate the processes involved in sustained attention tests and the locus of the practice effect in these tests. In the first contribution of the present thesis, based on theoretical considerations and earlier research, a process model of sustained attention tests was proposed. According to the model, performance in these tests should depend on: How fast the test-taker 1) perceives an item, 2) performs a simple mental operation to solve an item, 3) responds to indicate the answer to an item and 4) deliberately shifts towards the next item. In two studies (N Study 1 = 103, N Study2 = 100), three sub-components jointly predicted 55 to 74 % of the variance in three sustained attention tests. Perceptual and item solving processes showed to be strong predictors of performance, while there was a consistent trend towards a small influence of motor speed on test performance. Regarding item shifting, it was confirmed that the requirement to constantly stay on task and deliberately shift between items (the so-called self-paced mode) affected performance compared to a force-paced mode (sometimes also called computer-paced mode) which involved short intervals between successive stimuli. However, interindividual differences in item shifting were low in reliability and unrelated to performance in sustained attention tests. Thus, it remains to be elucidated how the requirement to continuously work on and deliberately shift between items impacts information processing in sustained attention tests. In the second contribution of the present thesis, it was examined whether the process model had to be extended when considering the characteristic presentation mode of sustained attention tests. Typically, in these tests, many stimuli are presented simultaneously and should therefore require the test-takers to focus on the currently relevant stimulus but could also enable them to preprocess upcoming stimuli. In order to assess the role of focusing and preprocessing in sustained attention tests, a modified version of the d2 test of sustained attention was created and its stimulus arrangement was manipulated. There was no effect of focusing in the present data, which might be due to the use of relatively simple distractors. However, it was shown that there was a large preview benefit (with regard to speed and error rates) in the conditions that allowed a preprocessing of upcoming stimuli. Moreover, interindividual differences in the extent to which preprocessing took place were reliable and substantially correlated with performance in three different sustained attention tests. Thus, a new component of performance in sustained attention tests was revealed: The preprocessing of upcoming stimuli. This component of test performance could very likely be linked to everyday sustained attention tasks that also benefit from a proactive scanning of potentially relevant information in the environment, like driving a car or proofreading a text. However, this potential link is yet to be investigated before we can draw conclusions about whether preprocessing constitutes an inherent part of sustained attention or rather an ancillary effect of the way sustained attention tests are designed. The subject of the third contribution of the present thesis were the large practice effects in sustained attention tests that have frequently been reported and range between one-third and one standard deviation (z. B. Bühner et al., 2006; Scharfen, Peters, et al., 2018; Westhoff & Dewald, 1990). Based on the previously proposed process model, it was investigated which of the sub-components of sustained attention tests benefitted from practice and to which extent. Therefore, in this study (N Study = 100), several cognitive tasks were administered twice to examine the effects of practice on the sub-components. It was shown that practice affected several sub-components, among them perceptual and motor processes but especially item solving processes and potentially also the coordination of the sub-components. However, neither the deliberate shifting between items nor the preprocessing of upcoming items became more efficient through practice. Altogether, the effects of practice on the sub-components of sustained attention tests seem to be rather complex and therefore, taking measures in order to reduce the practice effect in these tests or to recognize pre-exposure to these tests appears challenging. To sum up, using experimental test validation approaches, the present thesis allows a more in-depth insight into the processes involved as participants work through sustained attention tests and therefore, allows a better understanding of what these tests measure. It has again been shown that performance in these seemingly simple tests relies on several cognitive processes. However, the results of the current studies also raise new questions: For example, it is yet to be investigated how the continuous self-paced processing of stimuli impacts information processing in these tests, whether preprocessing corresponds to everyday sustained attention tasks and finally, how practice affects the validity of sustained attention tests. Altogether, the field of psychological assessment can strongly benefit from experimental approaches to validity. In future studies, it might be promising to also apply mathematical models of cognitive processes to address some of the open questions.