Track 5: Judgment & Decision Making I
Dan Schley (Erasmus University)

5B. Context Effects

Friday, March 4
11:15am – 12:45pm EST
Discussant: Dan Schley (Erasmus University)
MC: Ioannis Evangelidis (ESADE)
Calendar Invite: Add to calendar
Student Coordinator: Donald Gaffney (Vanderbilt University) (donald.r.gaffney@Vanderbilt.Edu)

Competitive Papers

Distribution Builder vs. Slider Scales: What Is the Best Way To Elicit Belief Distributions?
Authors: Beidi Hu (University of Pennsylvania), Joseph Simmons (University of Pennsylvania)
Presenting Author: Beidi Hu (University of Pennsylvania)
Researchers have recently embraced the practice of eliciting consumers’ entire subjective belief distributions. Our research compares two popular methods: Distribution Builder vs. SPIES (using slider scales). In 7 pre-registered studies (N = 10,815), we investigate whether Distribution Builders vs. SPIES yield different response patterns and, specifically, whether one method leads to more accurate distributions. We find that Distribution Builder usually elicits more accurate belief distributions, at least partially because SPIES users tend to complete the task “in order,” causing them to exhibit predictable response biases.
How Do Fresh Start Cues Shape Consumer Preferences for Formal (versus Casual) Stimuli?
Authors: Zhenyu Jin (University of South Carolina)
Presenting Author: Zhenyu Jin (University of South Carolina)
Consumers often encounter tradeoffs between formal and casual styles in product offerings (e.g., dress shoes vs. sneakers; formal vs. casual lifestyles; formal vs. casual activities). Surprisingly, little research has examined situational factors that may influence consumer preferences for formal versus casual stimuli. Across four studies, we propose and demonstrate that fresh start cues (i.e., cues that remind people about “making a fresh start regardless of the past or present circumstances”) induce consumers’ meaning-seeking motives (i.e., seeking for something meaningful, important, significant, and serious), which in turn increase their preferences for formal over casual product offerings.
Skewed Stimulus Sampling has Distorted Consumer Research
Authors: Hannah Perfecto (Washington University in St. Louis)
Presenting Author: Hannah Perfecto (Washington University in St. Louis)
Life has its ups and downs, but the world participants are asked to inhabit is often only up. Researchers overwhelmingly frame decisions as choices (vs. rejections) with pleasant, positive outcomes in pleasant, positive contexts (e.g., buying products, free bets). Past work has shown these positive decisions reduce deliberative thinking, which, in many cases, increases reliance on heuristics and succumbing to biases. This stacks the deck against study participants and may over-represent how often our real-life decision-making is flawed. We demonstrate the severity of this issue by attenuating four well-known findings (N=4,543) simply by changing the valence of the stimuli.If a goal of marketing research is to understand how people think about the world around them, this exclusion of negativity has resulted in an incomplete picture.
When it is Best to be Last: How Constructed Distributions Influence Sequential Judgments
Authors: Siyuan Yin (The Wharton School, University of Pennsylvania), Maurice Schweitzer (Wharton School, University of Pennsylvania)
Presenting Author: Siyuan Yin (The Wharton School, University of Pennsylvania)
When evaluating alternatives, individuals often view options in an ordered sequence. Across five pre-registered experiments and a field dataset, we find that people evaluate attractive options more favorably when they are presented last than when they are presented first. We also find that individuals are more enthusiastic about both their selected option and the entire choice set when they evaluate alternatives in ascending quality order than descending quality order. Both experience and expertise moderate these results. We also introduce Constructed Distribution Theory to show that individuals construct reference distributions as they evaluate alternatives.

Flash Talks

Logarithmic Axis Graphs Distort Lay Judgment
Authors: William Ryan (UC Berkeley), Ellen Evers (UC Berkeley Haas)
Presenting Author: William Ryan (UC Berkeley)
COVID-19 data is often presented using graphs with either a linear or logarithmic scale. Given the importance of this information, understanding how choice of scale changes interpretations is critical. To test this, we presented laypeople with the same data plotted using differing scales. We found that graphs with a logarithmic, as opposed to linear, scale resulted in laypeople making less accurate predictions of growth, viewing COVID-19 as less dangerous, and expressing both less support for policy interventions and less intention to take personal actions to combat COVID-19. Education reduces, but does not eliminate these effects. These results suggest that public communications should use logarithmic graphs only when necessary, and such graphs should be presented alongside education and linear graphs of the same data whenever possible.
Asymmetric Extrapolation of Improving and Declining Trends
Authors: Sokiente Dagogo-Jack (University of Georgia), Joshua Beck (University of Oregon), Justin Angle (University of Montana)
Presenting Author: Sokiente Dagogo-Jack (University of Georgia)
When predicting how outcomes will change in the future, consumers often use an extrapolation heuristic, expecting improving trends to persist and also expecting declining trends to persist. We find that the extent of this extrapolation depends on the direction of the trend. Specifically, consumers extrapolate declining trends more readily than commensurate improving trends. Two experiments demonstrate asymmetric extrapolation of various trends, from energy consumption to obesity rates. Corroborating these findings, an analysis of secondary data from a nationally representative survey demonstrates asymmetric extrapolation when predicting financial outcomes.


How the Voice Pitch Influences Processing Styles? The Effectiveness of Matching Auditory Pitch and Verbal Message in Advertising
Authors: Kaijun Zhang (Xiamen University), Jun Ye (Xiamen University)
Presenting Author: Kaijun Zhang (Xiamen University)
This research examined the effect of voice pitch on the consumers’ processing styles. With evidences from four main experiments, we proved that low voice pitch induces more abstract processing via the perceived distance. As such, this study raises a novel source of construal. In addition, this study shows that when both auditory (e.g., voice pitch) and verbal (e.g., temporal benefit) advertising elements induce the same level of construal, advertising effectiveness increases. This effect is due to the increasing processing fluency induced by the construal fit.
Reason Based Defaults
Authors: Shweta Desiraju (University of Chicago), Berkeley Dietvorst (University of Chicago)
Presenting Author: Shweta Desiraju (University of Chicago)
Defaults often involve increasing uptake of one option that serves only the majority of consumers. We introduce and test a new default, a “reason default”, which describes the reasons for choosing the default and alternative options. In Studies 1 & 2, we find that reason defaults help consumers who would be better served by an alternative option opt out of the default and increase satisfaction with the choice architecture. In Study 3, participants reported opinions about reason and standard defaults and felt that choices with reason defaults were less effortful and more transparent among other things.
How Pictogram Arrangements Influence Consumer Judgments
Authors: Gaurav Jain (Rensselaer Polytechnic Institute), Sunaina Shrivastava (Manhattan College), Zeynep Tolun (Rensselaer Polytechnic Institute)
Presenting Author: Zeynep Tolun (Rensselaer Polytechnic Institute)
Frequency pictograms are one of the most common graphical representations of quantitative information used in communications. The paper shows that individuals react differently to objectively equivalent information when represented in a sorted pictogram versus an unsorted pictogram. We show that individuals form a more optimistic judgment when presented with numerical information in the form of a sorted pictogram versus an unsorted pictogram. Using the backdrop of attribute framing across four studies, we demonstrate the above-posited phenomenon and find evidence for an optimism bias-based underlying mechanism.
Pastel Power: The Effects of Pastel Colors on Consumers' Attitudes toward Negative Information
Authors: Sasawan Heingraj (University of Texas Rio Grande Valley), Suwakitti Amornpan (University of Texas Rio Grande Valley), Michael Minor (University of Texas Rio Grande Valley), Areerat Pansuppawatt (Mahasarakham University), Yada Samart (Mahasarakham University)
Presenting Author: Sasawan Heingraj (University of Texas Rio Grande Valley)
Recently, online harassment on social media platforms becomes one of the dominant reasons causing anxiety disorders—the most common mental illness in the US. The characteristics of social media complicate this problem and prevent social media companies from controlling the content available on such platforms. Since this platform allows users to share their thoughts and customize the texts and background using various styles and colors, this study examines the effects of colors (i.e., pastel colors) on consumers’ attitudes toward negative information. Based on our preliminary IAT results, we found that consumers exhibited an automatic association between pastel colors and positive concepts. Additionally, when consumers posted a negative review on social media using pastel colors, other consumers tended to perceive them as having less tendency to take any negative action. Social media companies and public policymakers may consider utilizing pastel colors to weaken the negative effects on the social media community.
Return to schedule