Do performance-monitoring related cortical potentials mediate fluency and difficulty effects on decision confidence?

Abstract

It remains unclear to what extent the performance-monitoring (PM) processes involved in perceptual decisions are also involved in more complex reasoning decisions. To address this problem, we examined the role of PM-related cortical potentials in decision confidence. In simple perceptual decisions, the amplitude of these potentials has been shown to correlate with errors on the one hand and confidence ratings on the other. We therefore asked whether PM-related potentials mediate the effects of experimental manipulations of decision confidence on confidence ratings in an arithmetic reasoning task. In an EEG experiment, 49 participants had to quickly decide whether an equation (e.g. 9 * 7 = 65) was correct or incorrect and then report their confidence in that decision. Task difficulty and response fluency were varied to manipulate confidence. Pooling responses across correct as well as incorrect trials, we analyzed amplitudes of Error-Related Negativity (pooled ERN) and Error Positivity (pooled Pe). The results indicated that although pooled ERN and pooled Pe correlated with task difficulty, neither potential mediated the effect of task difficulty on confidence. Response fluency had an effect on confidence only for simple problems and this effect was partially mediated by pooled ERN but not by pooled Pe. These results suggest that pooled ERN and pooled Pe may reflect implicit performance monitoring, whereas pooled ERN may mediate the effect of the fluency heuristic on decision confidence.

The influence of error detection and error significance on neural and behavioral correlates of error processing in a complex choice task

“…The inversed effect, which is opposed to what Grützmann et al ( 2014 ) found, can have various reasons. First, the literature has shown that error monitoring is modulated by task difficulty (Kaczkurkin, 2013 ; Riesel et al, 2015 ; Schreiber et al, 2012 ; Somon et al, 2019 ; Voodla & Uusberg, 2021 ) and uncertainty (Bultena et al, 2020 ; Pailing & Segalowitz, 2004 ; Scheffers & Coles, 2000 ; Selimbeyoglu et al, 2012 ). While the N e amplitude is found to be attenuated, the N c amplitude increases with task difficulty and uncertainty, presumably due to a hindered error detection process (Pailing & Segalowitz, 2004 ).…”

“…These findings suggest that an effect of task difficulty or uncertainty might have counteracted an effect of error significance. However, studies that reported N e/c modulations also reported modulations of behavioral parameters by task difficulty (Pailing & Segalowitz, 2004 ; Riesel et al, 2015 ; Schreiber et al, 2012 ; Somon et al, 2019 ; Voodla & Uusberg, 2021 ). An effect of task difficulty was not reflected in our behavioral data, because the two self-evaluation conditions did neither differ significantly regarding error rates, RTs, post-response slowing, and post-response accuracy improvement nor regarding the percentage of too slow responses, which speaks against an increased task difficulty in the self-evaluation condition.…”

The Impact of Item Difficulty on Judgment of Confidence—A Cross-Level Moderated Mediation Model

“…Additionally, there are investigations that analyze these two aspects as separate factors. Voodla and Uusberg ( 2021 ) investigated the effects of item difficulty and fluency on confidence judgments. Their study found that confidence levels decrease with increasing task difficulty.…”