Comparing anticipation and uncertainty-penalty accounts of noninstrumental information seeking.

Abstract

Proposed psychological mechanisms generating noninstrumental information seeking in humans can be broadly categorized into two competing accounts: the maximization of anticipating rewards versus an aversion to uncertainty. We compare three separate formalizations of these theories on their ability to track the dependency of information-seeking behavior on increasing levels of cue-outcome delay as well as their sensitivity to outcome valence. Across three experiments using a variety of different stimuli, we observe a flat to monotonically increasing pattern of delay dependency and minimal evidence of sensitivity to outcome valence--patterns which are better predicted, qualitatively and quantitatively, by an uncertainty aversion information model.

A neural mechanism for conserved value computations integrating information and rewards

“…To test if the value of information also scaled with time in humans in this setting (11,67), we modified the human task to offer a choice between early vs. late access to informative cues (Fig 3E…”

“…We next asked whether the value of information scales with time. Information preferences have been shown to be influenced by the timing of task events ( 11, 17, 18, 36, 67 ), but we are only beginning to understand how these translate into subjective value and this value is computed by neural circuits ( 68 ). Notably, whereas standard temporal discounting scenarios only require evaluating a single event, the time of the extrinsic reward outcome (Tout), information seeking also requires evaluating the time when information will arrive – either the time of the cue (Tcue) for Info offers, or the time of the final reveal (Trev) for Noinfo offers (Fig 3A).…”

“…To test if the value of information also scaled with time in humans in this setting ( 11, 67 ), we modified the human task to offer a choice between early vs. late access to informative cues (Fig 3E, n=210; Fig S1). The human population was fitted with a significant mean positive subjective value of obtaining early vs. late information (Fig 3E, p < 0.001, signed-rank test), with significant positive values occurring in a substantial number of individuals as well (21%; significantly above chance, p < 0.0001, binomial test).…”

The role of risk, regret, and rejoice in non‐instrumental information seeking

“…Second, we implemented a task that interpolates a time delay between the elicitation of preference for outcome information and the receipt of that information. The implementation of an outcome delay allows the potential for participants to simulate the emotional reactions of receiving different outcomes as they wait, as is predicted by anticipation theories of information seeking (Iigaya et al, 2016; Liew et al, 2022). The task we use is a variant of the secrets task (see Figure 3) (Iigaya et al, 2016).…”

“…For instance, people acquire information simply because they are curious (Loewenstein, 1994, 2006; Noordewier & van Dijk, 2020) or find uncertainty aversive (Shani et al, 2008). Interestingly, the appetite for information generalizes to situations in which the acquired information cannot influence the outcome of decisions—a behavior often referred to as non‐instrumental information seeking (Bennett et al, 2016; Kobayashi & Hsu, 2019; Liew et al, 2022; Iigaya et al, 2016; Zhu et al, 2017). This behavior goes beyond a mere appetite for non‐instrumental information and is often manifested in people's willingness to forgo rewards for advance information (Charpentier et al, 2018; FitzGibbon et al, 2021).…”