15:00 - 16:30
Wed-B16-Talk VII-
Wed-Talk VII-
Room: B16
Chair/s:
Linus Hof
Core capacities of the mind like reasoning and decision making are exercised as responses to specific information-processing tasks. It is often assumed that these responses are strategic, taking into account resource limitations and trade-offs between the costs and quality of information-processing mechanisms. Yet, when the input information is missing, search must become part of the mind’s strategic response. This symposium features two tasks, inductive inferences and decisions under uncertainty, to highlight the strategic nature of information search (sampling). Marlene Hecht shows that if people consult their social network to make uncertain inferences, their search through the network is best described as sequential, limited, and less impactful for online contacts. Kevin Tiede presents work indicating that people increase their sampling effort to alleviate informational imbalances between described and experienced choice options. Linus Hof and Mikhail Spektor expand the symposium’s view on decisions from experience, demonstrating, for example, how sampling and integration strategies can interact to produce distinct choice patterns and psychoeconomic profiles. Doron Cohen concludes by presenting a simplified drift diffusion model. He uses the model to reconsider basic assumptions of sequential sampling approaches, which treat
information search as an evidence accumulation process. As a whole, the collection of talks suggests that our explanations of cognitive capacities and the phenomena they produce can be improved by postulating how these capacities implement a strategic information search.
Back
No Description–Experience Gap in Choices Between a Described and an Experienced Option
Wed-B16-Talk VII-02
Presented by: Kevin Tiede
Kevin Tiede 1, 2, Wolfgang Gaissmaier 2, Thorsten Pachur 3, 1
1 Max Planck Institute for Human Development, 2 University of Konstanz, 3 Technical University of Munich
People choosing between risky options seem to evaluate the options differently depending on whether they learn about them from a summary description (decisions from description) or from drawing sequential samples from the payoff distribution (decisions from experience). Does this impact of learning mode on the evaluation of risky options—referred to as description–experience gap—depend on whether the choice is between options presented in the same vs. different learning modes? And do people draw more samples from an experienced option when the alternative option is described (vs. experienced) in order to align the certainty about the payoff distributions across options? We examined these questions by comparing people's choice and search behavior in a mixed-mode condition, where they chose between a described and an experienced option, with behavior in a purely description- or experience-based condition. Using cumulative prospect theory to model choices and measure people’s subjective representations of outcome and probability information, we found clear differences between the purely description-based and the purely experience-based conditions. In the mixed-mode condition, however, the discrepancies in the subjective representations of the described and the experienced options disappeared. As expected, per-option search effort was higher in the mixed-mode condition than in the purely experience-based condition. Our findings underscore the importance of studying the many facets of the choice context—that also includes the learning mode of context options—in order to fully understand both information search and the mechanisms of preference construction in risky choice.
Keywords: risky choice, decisions from experience, sampling, description–experience gap, cumulative prospect theory, hierarchical Bayesian modeling
Back