Valuation and estimation from experience

Abstract

The processing of sequentially presented numerical information is a prerequisite for decisions from experience, where people learn about potential outcomes and their associated probabilities and then make choices between gambles. Little is known, however, about how people's preference for choosing a gamble is affected by how they perceive and process numerical information. To address this, we conducted a series of experiments wherein participants repeatedly sampled numbers from continuous outcome distributions. They were incentivized either to estimate the means of the numbers or to state their minimum selling prices to forgo a consequential draw from the distributions (i.e., the certainty equivalents or valuations). We found that participants valued distributions below their means, valued high-variance sequences lower than low-variance sequences, and valued left-skewed sequences lower than right-skewed sequences. Though less pronounced, similar patterns occurred in the mean estimation task where preferences should not play a role. These results are not consistent with prior findings in decision from experience such as the overweighting of high numbers and the underweighting of rare events. Rather, the qualitative effects, as well as the similarity of effects in valuation and estimation, are consistent with the assumption that people process numbers on a compressed mental number line in valuations from experience.

Biased confabulation in risky choice

“…In addition, both memory and choice exhibited an overweighting of extreme outcomes, even when those outcomes were drawn from a continuous distribution. Experiment 2 highlights how this memory bias emerges even in the absence of preferential choice, suggesting that the choice biases are driven by these memory biases and are not the cause of the memory biases (see also Vanunu et al (2020) and Olschewski, Newell, Oberholzer, and Scheibehenne (2021)). Notably, the memory recalls were not veridical, nor evenly distributed.…”

“…Interestingly in our experiments the relationship between memory and estimation was stronger in the non-preferential compared to the preferential task, although this may be due to the simpli ed nature of the non-preferential task and consequent better memory recall overall. Olschewski et al (2021) found that people consistently under-weighted the average of continuous outcomes in an estimation task. In our task, we demonstrate that people show under or overweighting depending on whether a risky option contains the high or low value extremes and the e ects of under-weighting are stronger than over-weighting.…”

The role of episodic sampling in evaluation

“…One possibility is that our requirement to also remember the items presented an additional processing burden that pushed our participants into using an intuitive strategy. Considering our findings relating to task order, a compelling question for future research might be whether memory-based evaluation is yet another alternative to analytic and intuitive evaluation (Brezis et al, 2015), the extent to which people adaptively switch between these based on the task environment, and the extent to which estimation is the basis for evaluation (Olschewski, Newell, Oberholzer, & Scheibehenne, 2021).…”

Learning the lie of the land: How people construct mental representations of distributions.

“…For example, explanations of people's risk behaviour in uncertain environments often invoke the strong assumption that event-frequencies are represented as probability distributions. Imagine that, in a modified version of our task and after having asked participants to learn the sequence of numbers, we asked them to place a bid on how much they would be willing to pay to receive a random draw from the sequence (cf., Olschewski et al, 2021). A random number is the drawn from the distribution, and participants receive their bid minus the value of the number drawn.…”