The Effect of Non-Instrumental Information on Reward Learning

• DOI: 10.31234/osf.io/cpd8u
• Publication Date: 2023-09-28T05:02:20Z
• Authors: Jake Ryan Embrey; Amy Xiangjin Li; Shi Xian Liew; Ben R Newell
• Keywords: Representation; Value (mathematics)

Many investigations of information-seeking highlight a tendency to forgo financial reward in return for advance information that cannot be used change future outcomes. Most these experiments use tasks which contingencies are described participants. The such descriptions leaves open the question whether opportunity obtain non-instrumental influences people’s ability learn and represent underlying structure an experimental environment. In two experiments, participants completed two-armed bandit task with monetary incentives were learned via trial-by-trial experience. We find, akin description-based tasks, willing receive about delayed, unchangeable outcome. Crucially, however, there is little evidence this willingness pay driven by inaccurate representation structure: participants’ representations approximated regardless presence information. results extend previous conclusions regarding intrinsic value experience-based domain challenges probing memories experienced rewards.