Choice Symposium 2023: Cognitive Costs Session Schedule
DAY 1: Wednesday, August 9
3:00pm - 7:00pm: Registration
7:00pm - 10:00pm: Dinner
DAY 2: Thursday, August 10
9:10am - 9:25am: Dan Bartels
Settling in and Opening Remarks
9:25am-9:45am: Alex Moore
Heuristics and Strategies in Consumer Search
Consumers often find good products even when given a bewildering number of options, but little research has looked at how consumers navigate through potential choices as they search for good products. In my research, I look at different strategies and heuristics that consumers use to navigate as they look for good products, compare these strategies to optimal models of search, and discuss how these choices may be adaptive when people are faced with limited cognitive resources.
9:47am-10:07am: Mark Ho
Algorithmic, Representational, and Information Theoretic Costs in Decision-Making
Accounts of cognitive costs in decision-making should be both psychologically plausible and theoretically general, but these goals are sometimes at odds. I will discuss three different formalizations of cognitive costs and projects in which they have been applied.
10:09am-10:29am: Joan Ongchoco:
Direct consequences of perception and imagination on decision-making
Perception (and imagery) and decision-making often take up two sides of theoretical and sociological divides — but here, I will present two studies that try to bridge these two together. First, I will discuss several experiments in which mental imagery of different options can lead to benefits for impending decisions — and then present a computational framework through which we can explore the nature and scope of this process. Second, I will then show how perception — especially how visual processing segments experience into discrete events — can have surprisingly direct consequences for decision-making, restricting the impact of a well-known cognitive bias such as the anchoring effect.
10:29am-11:15am: COFFEE BREAK/DISCUSSION
11:15am-11:35am: Christin Shulze
Beating the odds: Cognitive effort in repeated decisions from experience
In repeated decisions from experience, people often deviate from maximization in striking ways—they underweight rare events and probability match instead of maximizing probability or returns. I will discuss findings suggesting that these prominent deviations from maximization are associated with effortful pattern search—that is, people try to beat the odds in probabilistic choice, but inevitably fail—rather than arising from cognitive constraints that favor simple strategies. I will consider how maximization, in turn, can arise from low-effort, satisficing strategies and how these behaviors develop across childhood.
11:37am-11:57pm: Stefan Bucher:
Cognition-Optimized Information Filters
In this talk, I propose mitigating information overload with information filters that account for cognitive costs. I will introduce a multi-attribute choice model of a rationally inattentive consumer, and discuss the information design problem of a Sender selecting the information available to the consumer.
11:57am-12:30pm: Summary Discussion
Discuss Group Interests
12:30pm-2:00pm: LUNCH
2:00pm-2:20pm: Ada Aka
Inferring Consideration Sets: A Computational Model of Naturalistic Memory-Based Decision Making
We combine established theories of consideration set formation and memory search with techniques from natural language processing to propose a new computational framework. This framework explores how people retrieve and choose among hundreds of common choice items stored in memory. Consequently, it provides data-driven insights into the core cognitive mechanisms at play in memory-based decision making.
2:22pm-2:42pm: Giovanni Compiani
A Method to Estimate Discrete Choice Models that is Robust to Consumer Search
We propose a method to recover consumer preferences for product attributes that is valid regardless of whether consumers have full information about those attributes or need to engage in search to uncover some of them. Our method only requires data on attributes and final choices and can be used to predict behavior and quantify welfare in the scenario where consumers have full information
2:47pm-3:07pm: Summary Discussion
Discuss Group Interests
3:07am-3:45am: COFFEE BREAK/DISCUSSION
3:45pm-4:05pm: Gabriele Oettingen
Processes of Disengagement from Desired Futures, Missed-Out Pasts, and Inevitable Endings
Mental contrasting has been shown to be a self-regulation strategy that is mentally effortful, but at the same time highly effective in helping people to set priorities: To actively engage in attaining desirable and feasible futures and to actively disengage from attaining undesirable or unfeasible futures. I will be talking about how mental contrasting can be used to actively decide against pursuing futile endeavours, whether they lie in the future, pertain to counterfactual pasts, or refer to deflecting inevitable endings.
4:07pm-4:27pm: Ian Krajbich
Learned Impatience
We study a reinforcement learning task where reward information is broken down into two components, one immediate and one delayed, and so at any given time, subjects receive feedback from the option they just chose as well as their previous choice. We find that our subjects put significantly less weight on the delayed feedback, resulting in a systematic bias to choose options with low total reward but high immediate reward, and that this effect is partly explained by an attentional (gaze) bias towards immediate rewards.
4:27pm-5:30pm: Summary Discussion: Emerging Themes
DAY 3: Friday, August 11: Joint Session
9:00am-9:20am: Todd Gureckis
Can cognitive discovery be incentivized with money?
We examine the degree to which discovery of patterns can be incentivized by financial rewards. Across five experiments, we find no effect of incentive on rule discovery performance, but in a sixth experiment requiring category recognition but not learning, we find a large effect of incentives on response time and small effect on task performance.
9:20am-9:40am: Ryan Oprea
What makes rules complex?
If we think of complexity (as computer scientists do) as the cost of implementing an algorithm, in principle we can measure "complexity" in humans directly by eliciting people's willingness to pay to avoid being asked to implement algorithms on data. When we do this for a range of algorithms, we find significant structure that allows us to map what formal features of algorithms make them complex and therefore aversive to decision makers.
9:40am-10:00am: Eric Johnson
TBD
10:00am-11:00am: COFFEE BREAK/DISCUSSION
11:00am-11:20am: Steph Smith
TBD
11:20am-11:40pm: Daria Dzyabura
Product Images & Consumer Search
In a large online apparel retail dataset, we find that an item's click rate and its purchase rate are uncorrelated. However, when we compare the frequently clicked products with frequently purchased products based on "standard" attributes, such as price, discount, category, they appear similar. Only when we decompose the image features do we observe large differences. We explore why these large discrepancies in propensity to click and purchase may occur, and how online retailers should optimize for different scenarios.
11:40pm-12:00pm: Gerald Haubl
Choice Architecture and the Dynamics of Cognitive Costs
I plan to share a few key findings from an ongoing research program on the dynamic effects of choice-architecture tools (particularly defaults) on preferences and behavior. I will attempt to do so through the lens of cognitive costs
12:00pm-2:00pm: LUNCH
2:00pm-2:25pm: Nick Reinholtz
Considering Variability
Optimal behavior often requires understanding the variability of outcomes. I will discuss some of my research on people's understanding of variability in different decision-making domains (investing, consumer search, etc.)
2:27pm-2:52pm: Ben Newell
Cognitive Misers or Active Thinkers? Effort, Boredom and Cognitive Demand
I will discuss work examining how the type of cognitive demand affects people’s willingness to seek or avoid effortful tasks. I will also show how the context in which a task is performed radically alters the balance between effort and boredom. I will then try to relate these findings to discussions about achieving effortful behaviour change via effortless channels.