The project was divided into four sub-projects (SP). In SP 1 we compared the effects of positive (win) and negative (loss) reward signals in the context of different cognitive control and response modes. In several behavioral paradigms, we found that reward cues improve performance globally, while reward-related target stimuli promoted valence-action biases (win=approach; loss=avoid). These biases are maladaptive when they are not in line with the task goal - e.g. when an avoid response is required to gain a reward. Neural data (fMRI) revealed that reward cues increase activity in attentional control regions globally, but there was also evidence for interactions where valence-incompatible responses were associated with enhanced neural activity compared to compatible ones, likely reflecting additional resources to overrule the inherent bias.
In SP 2, we investigated the effects of reward signals on cognitive control by embedding win and loss prospect in a conflict task. The main observation was that win and loss stimuli similarly improve performance, which challenges the common view that expected losses have a higher value than equivalent wins. Neurophysiological data from this paradigm (EEG and pupil) supports the notion that win outcomes have a stronger motivational value. A second set of studies focused on the effects of incidental action-outcome contingencies on cognitive control. We found evidence that these can have a selective impact on a certain trial type (i.e. conflict trials). Using a similar design, we tested in how far reward and affective (emotional) outcomes lead to similar performance adjustments (see SP3).
SP 3 further explored the overlap between reward and affective valence both in terms of behavioral and neural modulations. First, we observed similar valence-action biases triggered by the anticipation of affective stimuli, which are comparable to the biases for reward stimuli (SP 1). Second, we found overlapping neural activity modulations (fMRI) for reward-related and affective stimuli when considering their absolute valence (positive vs. negative). Third, we explored potential compatibility effects between reward and affective valence and found that positive affect stimuli facilitate performance in reward trials (compatible valence), while negative affect stimuli impair performance (incompatible valence). These behavioral effects were accompanied by neural modulations indexing effortless versus effortful processing, respectively.
SP 4 focused on the overlap between reward and affective valence in the context of long-term memory by employing an encoding-retrieval procedure (with fMRI during encoding). Behaviorally, we observed that both reward and positive affect improved subsequent retrieval independently. The neural data analysis is still in progress due to a major delay during the covid pandemic in 2020. The analysis will focus on the contribution of two neurotransmitter systems, noradrenalin and dopamine, which both have been implicated in memory processing, as well as processing rewarding and affective stimuli. Finally, in a pilot study we found that white matter integrity in several tracts was inversely related to reward-triggered response speeding, which may reflect higher susceptibility to reward stimuli.
Twelve articles directly emerging from the SPs have been published in peer-reviewed journals. Related studies and collaborations led to 12 additional publications (including 5 in the clinical domain), 3 book chapters, 2 articles in revision, and 1 under review. We have presented the above results at different (inter)national conferences and research facilities, and have organized a symposium.
