Event related potentials

The cue-locked P300 provides an indirect measure of early attentional processes that are not accessible through behavioural measures. The P300 is a large centro-parietal deflection, peaking approximately 300ms after cue onset (Luck, 2014). Early accounts emphasized the P300 involvement in the evaluation of stimulus appearance probability: whether it follows a sequence or not (see e.g. oddball paradigm; Polich, 2007). In this framework, the component was considered to be a manifestation of the previous context update (Guo, Duan, Li, & Pallers, 2006). Subsequent research described the role of P300 in attentional resource allocation involved in task preparatory processes (e.g. Pelosi, Holly, Slade et al., 1992; Finke, Escara, & Barcelo, 2012). Given that the P300 can be used as a covert measure of resource allocation, independent of behavioural responding, it is particularly suited to investigate in a multitask paradigm that varies in task demands. Moreover, some studies extended the theory of attentional P300 by suggesting that P300 might be sensitive to emotional stimuli (Cuthbert, Schupp, Bradley, Birbaumer, & Lang, 2000; Keil, Bradley, Hauk, et al., 2002; Hajcak & Olvet, 2008). For example, it was shown that self-relevant stimuli (the ones linked to personal concerns and values) are more likely to receive attentional resources, which will be reflected by increase in P300 amplitude (Gray,  Ambady, Lowenthal, & Deldin, 2003). The study from this perspective might add to the existing understanding of negative emotions, such as maths anxiety (Suárez-Pellicioni, Núñez-Peña, & Colomé, 2016). Previous studies investigated mostly oddball task P300 and reported moderate-to-high test-retest correlation coefficients, ranging from 0.5 to 0.8 for component amplitude and from 0.4 to 0.77 for peak latency of component (Fabiani, Gratton, Karis, & Donchin, 1987; Segalowitz & Barnes, 1993; Walhovd & Fjell, 2002).

The P600 is a positive component that peaks between 500 and 1000ms after stimulus onset (Luck, 2014). P600 has mainly been studied during language processing, particularly, the component has been observed when syntactic structures are violated (e.g. Osterhout & Holcomb, 1993). More recently, research has demonstrated the involvement of the P600 in semantic violations (Swaab, LeDoux, Camblin, & Boudewyn, 2012) and mathematical rule violations (Núñez-Peña & Honrubia-Serrano, 2004). A recent review suggested that P600-related effects reflect an update of current representation of incoming information (Brouwer, Fitz, & Hoeks, 2012). It remains unclear whether P600 reflects common neurocognitive processes associated with rule violations in general, or functionally distinct mechanisms underlying processing of specific information. The paradigm used in the present study includes three types of tasks, enabling an investigation of domain generality vs. specificity of P600, comparing a component for different types of task: language and maths-related ones. While there is a wealth of research on P600 as a component, little is known about its reliability. To our knowledge, there is only one study investigating the reliability of P600 component, which showed a moderate test-retest correlation for the component amplitude (.64) and no correlation for latency (Borghans & Prince, 2013). However, there are some studies which consider P3b to be overlapping or qualitatively similar to P600 (see for example Bornkessel-Schlesewsky, Kretzschmar, Tune, et al. 2011). Studies addressing the reliability of P3b showed reliability indices as high as .88 for the area under the curve
estimates (see Cassidy et al., 2012).