Individual differences in task-switching paradigms

Previous attempts to completely eliminate task-switch costs by improving advance preparation (e.g., Schneider, 2016, 2017) and by extensive practicing task rules (e.g., Zhao, Wang & Maes, 2018) had limited success; most researchers reported significant “residual” switch costs in typical task-switching paradigms, suggesting incomplete preparation of tasks. In most task-switching studies it was ignored that participants may have different switching abilities and/or strategies. It has been shown, however, that some participants perform better than others, showing only small switch costs for specific target-response mappings (Lindsen & De Jong, 2010), or reduced switch costs after extended practice of task rules (Stoet & Snyder, 2007). Other researchers have claimed that a few select participants showed superior performance in multi-tasking as well as other cognitive and perceptual tasks (Haaf & Rouder, 2017, 2018; Ramon et al., 2016; Robertson, Noyes, Dowsett, Jenkins & Burton, 2016; Strayer & Watson, 2012). It is therefore possible that in task-switching a few participants may also perform considerably better than others, showing minimal or no errors and no switch costs across different conditions. The present thesis aimed to study individual differences in task switching by monitoring group-average as well as individual performance under different conditions. Chapter 1 provides an overview of different task-switching paradigms and accounts to explain task-switch costs and residual switch costs. In particular, I discuss studies that considered individual differences in various tasks and experimental paradigms. In order to detect superior performance in task switching and to study individual differences, I conducted an experiment and two follow-up studies that are documented in Chapter 2. I employed Generalised Linear Mixed-effects Models (GLMMs; Bolker et al., 2009) on single-trial RTs in order to investigate whether individual participants exhibit comparable switching effects and whether best performers vary across different conditions, paradigms, and experiments. Seven psychological scales and a Raven intelligence test were also employed in order to better understand possible factors that may be related to differences in performance in task switching. The results of Chapter 2 suggest that there are considerable individual differences in task switching and that smaller individual switch costs may be due to more efficient preparation during cue-stimulus intervals. Furthermore, I suggest that efficient task preparation may be linked to better executive control, general intelligence, higher motivation, and lower levels of impulsivity. Since efficient preparation after task-cueing plays an important role in reducing or even eliminating switch costs, Chapter 3 sought to facilitate faster cue-based preparation, thereby reducing switch costs in typical participants. Previous research has suggested that task-switch costs are smaller for “transparent” compared to “non-transparent” or standard cues (e.g., Logan & Schneider, 2006; Schneider, 2016). In three experiments I compared “non-priming” cues with carefully designed “priming” cues that indicate the upcoming target feature and response mappings. I found that participants who used priming cues showed smaller and non-significant switch costs in their response times, independent of the interval between cue and target stimulus. These participants also showed more homogenous task-switching performance. In a related EEG study that investigated the temporal dynamics of preparation I provide evidence that priming cues elicited significantly larger cue-locked positivity in switch trials compared to repeat trials at electrode Pz, in different cue-stimulus intervals. Similar to previous results on transparent cues, this suggests that preparation can be facilitated (e.g., Karayanidis & Jamadar, 2014). Gender-related individual differences in task switching are investigated in Chapter 4. Previous reports on gender differences have been controversial (e.g., Polunina, Bryun, Sydniaeva & Golukhova, 2018; Stoet, O’Connor, Conner & Laws, 2013). In Chapter 4 I studied whether gender differences are present in the preparation phase of task-switching by manipulating the cue-stimulus interval. The results of the experiment in Chapter 4 suggest that females may have a slight advantage in task switching for longer preparation intervals. While individual switch costs varied considerably for female and male participants, females showed smaller congruency effects in trials with task repetition. I suggest that female participants were faster than males in task preparation. In addition, some females seemed to employ a different strategy in repeat trials. For example, they may have performed different tasks by recalling cue-stimulus-response associations without applying the task rules. In Chapter 5, I summarise possible reasons that may lead to superior switching performance and discuss other factors that may account for different switching performance across individuals. Limitations and future directions of research are also outlined. I suggest that individual differences should be considered and reported in order to improve model and hypothesis testing and to make results more replicable. Finally, I discuss whether superior switching abilities can be related to other tasks and situations