When the outcome is different than expected: Subjective expectancy shapes reward prediction error at the FRN level

Abstract

Converging evidence in human electrophysiology suggests that evaluative feedback provided during performance monitoring (PM) elicits two distinctive and successive ERP components: the feedback‐related negativity (FRN) and the P3b. Whereas the FRN has previously been linked to reward prediction error (RPE), the P3b has been conceived as reflecting motivational or attentional processes following the early processing of the RPE, including action value updating. However, it remains unclear whether these two consecutive neurophysiological effects depend on the direction of the unexpectedness (better‐ or worse‐than‐expected outcomes; signed RPE) or instead only on the degree of unexpectedness irrespective of direction (i.e., unsigned RPE). To address this question, we devised an experiment in which we manipulated the objective reward probability and the subjective reward expectancy (via instructions) in a factorial within‐subject design and explored amplitude changes of the FRN and the P3b. A 64‐channel EEG was recorded while 32 participants performed a speeded go/no‐go task in which evaluative feedback based on the reward probability either violated expectancy (thereby creating a RPE) or did not. This violation corresponded either to better‐ or worse‐than‐expected events. Results showed that the FRN was larger when RPE occurred than when it did not, but irrespective of the direction of this violation. Interestingly, in these two conditions, action value was updated for the positive feedback selectively, as shown by the P3b amplitude. These results obey a two‐stage model of PM assuming that unsigned RPE is first rapidly detected (FRN level) before the positive feedback’s value is updated selectively (P3b effect).

Dissociating the effect of reward uncertainty and timing uncertainty on neural indices of reward prediction errors: A reward positivity (RewP) event-related potential (ERP) study

“…In conclusion, the results of this study add to the current body of literature in discussing the underlying generators of the RewP by directly comparing effects of reward and timing uncertainty on RewP amplitude. The current results confirm that uncertain trials elicit greater RewP amplitude, when compared to certain trials (Hajcak et al, 2007;Pfabigan et al, 2011;Walentowska et al, 2019;Xu et al, 2011), which supports theoretical arguments that the RewP indexes a reward prediction error. These results also add, however, that this effect is driven by uncertainty specifically surrounding the reward and is unchanged by uncertainty surrounding the timing of feedback presentation.…”

“…Additionally, the current study provides evidence for increased motivational salience of rewards compared to the timing of reward or non-reward presentation (cf., Pegg & Kujawa, 2020). As expected, we found that larger RewP amplitudes were present on uncertain reward trials compared to certain reward trials, confirming previous studies that have shown the same or similar effect (Hajcak et al, 2007;Pfabigan et al, 2011;Walentowska et al, 2019). However, contrary to our hypothesis, RewP amplitude did not significantly differ between certain and uncertain trials on the timing cued task where participants were cued to the timing of reward presentation but not the presence/absence of reward.…”

“…RewP amplitude is more positive for gain trials as opposed to loss trials (in the case of the FRN, more negative for loss trials; Lohse et al, 2020;Miltner et al, 1997;Proudfit, 2015), and is greater for rewards that are larger in magnitude (Meadows et al, 2016). Uncertain outcomes also elicit greater RewP amplitude compared to certain outcomes (Hajcak et al, 2007;Pfabigan et al, 2011;Walentowska et al, 2019), but RewP amplitude may be equally effected whether the outcome is better than expected or worse than expected (Ferdinand et al, 2012;however, see Mulligan & Hajcak, 2018). Notably, higher RewP amplitude during feedback processing is associated with behavioral adjustments on subsequent trials following erroneous responses (Cohen & Ranganath, 2007;Lohse et al, 2020;Schultz, 2016), suggesting that the RewP indexes the signal between neural networks that adjusts future behavior.…”

Monetary payoffs modulate reciprocity expectations in outcome evaluations: An event‐related potential study

“…The findings in the current study do not provide supportive evidence for the negative outcome bias reflected by FRN (e.g., loss outcomes always elicit a more negative FRN). Instead, it supports evidence showing that the FRN reflects the degree of unexpectedness irrespective of valence (i.e., better‐ or worse‐than‐expected outcomes; Ferdinand et al., 2012; Walentowska et al., 2019). Future studies will have to explore the reason why FRN only respond to expectation violation rather than outcome valence or the interaction of outcome valence and expectation violation in the Chicken game task involving monetary reward.…”

“…Proudfit (2015) proposed that the fERN, MFN, and FRN are reward positivity (RewP), which provides an appropriate description of the scalp‐recorded difference between the ERP elicited by the gain (correct) and loss (incorrect) feedback. Besides an index of differentiating the valence of feedback, the feedback‐locked ERP component, especially FRN, also can be regarded as an “expectation violation” signal that codes for expectation independent of valence (Ferdinand et al., 2012; Walentowska et al., 2019). The violation of expectations leads to a more negative‐going FRN deflection (Gu et al., 2020; Pfabigan & Alexopoulos, 2011; Van der Molen et al., 2013).…”