�ݺ�ߣ

Fireworks from JDM
Report from the 32nd conference
on Judgment & Decisionmaking

Habit Design Meetup #4 Dec 1, 2011

Paul Whitmore Sas
paulsas@xpcxpts.com

Thursday, December 1, 2011

JDM is the premier international
conference on behavioral economics

http://www.sjdm.org/
programs/2011-program.pdf


Debt Snow Ball (David Gal)

Closing out more accounts early is a stronger
predictor of complete success than paying off bigger
balances within one account

Can Small Victories Help Win the War? A Quasi-Experiment from the Consumer Debt Settlement
Industry
Gal, David (Northwestern University); McShane, Blake (Northwestern University)
The question of how consumers should manage their debts is one of great practical significance. Using a data set
obtained from a consumer debt settlement firm, we find that closing debt accounts is predictive of eliminating
one's debts. Conversely, the dollar balance of closed accounts is not predictive of debt elimination when
controlling for number of accounts closed. These findings suggest that closing debt accounts, even when
rationally suboptimal, might motivate consumers to persist in a debt elimination goal. In addition to implications
for consumer debt management, our findings bear important theoretical and methodological implications for
research on goal pursuit.


Trafﬁc Light Ratings on Food Raise
Caloric Intake (Eric VanEpps)
Calorie labeling has zero impact
Adding “trafﬁc light” (red-yellow-green)
Makes consumption worse

Menu Labeling: Posting Calories versus Traffic Lights to Encourage Healthier Eating
VanEpps, Eric M. (Carnegie Mellon University); Downs, Julie S. (Carnegie Mellon University); Wisdom, Jessica (Carnegie
Mellon University); Loewenstein, George (Carnegie Mellon University)
We compared two approaches to labeling healthfulness of restaurant foods on menus: posting calories (popular among
American legislators) and traffic light icons (emerging in the UK and Australia). Using green, yellow and red images is
thought to be a simpler way of helping consumers to identify healthier options. Our participants (N=320) selected their meals
from manipulated menus prior to entering McDonald’s, then returned with their receipt for a survey and payment. Calorie
posting had no effect on orders, but traffic light labels led customers to order significantly more calories, strongly mediated
by ordering a larger number of traffic-light-labeled items.


“Mental Eating” & Satiation
Carey Morewedge’s Lab at CMU
Experiment published last Dec in Science showed:

Thinking about eating 30 M&Ms
vs.
Thinking of eating 3 M&Ms

Those who thought of 30 were more “habituated” and
ate less when given the option
“Thought for Food: Imagined Consumption Reduces Actual Consumption”
Carey K. Morewedge,1* Young Eun Huh,2 Joachim Vosgerau2
10 DECEMBER 2010

VOL 330

SCIENCE


Recent Satiation via Mental Imagination
from Carey Morewedge’s Lab at CMU
Feels near or Far (Yang Yang at CMU)
Feels Far or Near? How Subjective Perception of When One Last Consumed Influences Satiation
Yang Yang (Carnegie Mellon University); Galak, Jeff (Carnegie Mellon University); Redden Joseph P. (University of
Minnesota); Kyung Ellie J. (Dartmouth College)
Previous research show that satiation can be constructed based on whether people recall their past consumption, the
ease of retrieval of past consumption, and the feeling of how much they have consumed. We demonstrate in the current
research that the subjective perception of when one last consumed also has a significant impact on satiation.
Specifically, the subjective sense of having consumed more recently makes people want to eat significantly less and
feel less hungry (Experiment 1), feel more satiated and enjoy their favorite song substantially less (Experiment 2), and
purchase food with lower caloric value (Experiment 3).

Satiation is manipulation speciﬁc
Specificity in Sensitization
Huh, Young Eun (Carnegie Mellon University); Vosgerau, Joachim (Carnegie Mellon University); Morewedge, Carey K.
(Carnegie Mellon University)
Eating a food engenders habituation, a decrease in one’s responsiveness and motivation to obtain the food. General-process
theories of motivation also posit that eating a food sensitizes one to other foods, increasing one’s responsiveness and
motivation to obtain the food, but the breadth and scope of sensitization effects are unclear. We show that sensitization
occurs for foods frequently consumed in conjunction with a target food (i.e., complements), but does not occur for
unrelated foods. In three experiments, the repeated imaginary consumption of a food increased actual consumption of its
complements but did not affect consumption of unrelated foods.


‘‘Less Can Be More’’ in Risk Graphics
Brian Zikmund-Fisher

<

Viewers of survival-only pictographs had better accuracy

Omitting redundant mortality outcome statistics can be an effective method of risk communication

A Demonstration of ‘‘Less Can Be More’’ in Risk Graphics
Brian J. Zikmund-Fisher, PhD, Angela Fagerlin, PhD, Peter A. Ubel, MD MEDICAL DECISION MAKING/NOV–DEC 2010


Rapid Reset Rewards for Speed Addicts

Method to pay people for pee’ing clean
Sequential clean pees increases payment
Most motivating: Rapid Reset, so 3 clean returns

Frequent rewards improve outcomes for methamphetamine addiction: A behavioral economic analysis of
a contingency management treatment program
Ling, Kimberly (UCLA); Krishnamurti, Tamar (Carnegie Mellon University); Shoptaw, Steven (UCLA)
Contingency management (CM) is an effective methamphetamine substance abuse treatment where escalating
payments are made in exchange for abstinent behavior. In a randomized, controlled trial of methamphetamine-
dependent gay men, 78 participants were randomly assigned to receive CM payments as part of a larger trial of
behavioral treatments. This study examines features of purchases (e.g. frequency, magnitude, hedonic vs.
utilitarian type) made with CM payments to elucidate the ability of CM incentives to yield positive treatment
outcomes for methamphetamine dependence. Regression of purchase features on treatment outcomes assessed
preferences for redemption as reflecting an economic hedonic or consumable substitute.


Overwork (Chris Hsee)

5 mins of music (work = play noise burst)
5 mins of eating chocolates earned

Players worked TOO hard, earned more
chocolate than they wanted to eat
Medium Maximization
Overworking
Hsee, Christopher K. (University of Chicago); Zhang, Shirley (Applicant to PhD programs)
We introduce an experimental simulation paradigm to study whether and when people overwork - work and earn more than
what one needs or can consume. We found that people are more likely to overwork when production rate is high than low, that
overworking is due to mindless accumulation rather than uncertainty protection, and that encouraging people to focus on the
consequence or not allowing people to earn more when they have enough to consume can reduce their tendency to overwork
and at the same time increase their subjective happiness.


Optimal quitting point (Cade Massey)

30 balls in urn, either 18 red or only 12

Win by drawing 15 - pay 10 cents per ball

People often overpay to persist
Pressing On or Giving Up: Goal Pursuit Under Uncertainty
Massey, Cade (Yale University); Koehler, Derek J. (University of Waterloo);
Goal pursuit depends on motivation and belief revision (e.g., assessing the likelihood that the goal is
achievable). The burgeoning literature on goals and decision-making has focused on the former
(e.g., Heath, Larrick & Wu, 1999; Fishbach & Dhar, 2005). We consider the latter, investigating how
participants decide whether to abandon a goal. We find that participants over-persist in the pursuit
of immediate goals, foregoing the benefits of alternative pursuits. This bias is strongest in the face of
early, negative feedback. We suggest the need to abandon potentially profitable goals because of high
opportunity costs is important, common and, apparently, difficult.


Implementation intentions - commit to
vaccine (Katherine Milkman)
Scheduling is a classic “channel factor”
Lift is signiﬁcant, but not huge

Using Implementation Intentions Prompts to Enhance Influenza Vaccination Rates
Milkman, Katherine (The Wharton School, The University of Pennsylvania); Beshears, John (Graduate School of
Business, Stanford University); Choi, James J. (School of Management, Yale University); Laibson, David (Department
of Economics, Harvard University); Madrian, Brigitte C. (Harvard Kennedy School, Harvard University)
We experimentally evaluate whether planning prompts increase influenza vaccination rates. Over 3,000 employees at a
firm received reminder mailings listing the times and locations of vaccination clinics at their office sites. Treatment
condition mailings also included a prompt to write down either the date or the date and time the employee planned to be
vaccinated. Vaccination rates increased meaningfully with the specificity of the planning prompt received: by 1.5
percentage points over the control in the date group and 4.2 percentage points in the date and time group. The effect size
doubled for employees at offices with only one-day clinics.


Adding 2nd tier info reduces
attractiveness (Stephen Spiller)
Garcia showed that
A hotel with 5-star restaurant
beats
A hotel with 5-star restaurant + 3 star pool

Representing Money as a Set: The Influence of Irrelevant Alternatives on the Perceived Value of Money
Spiller, Stephen A. (UCLA); Ariely, Dan (Duke University)
How do people represent money? Because money is fungible and people can allocate it in so many different ways, it is
difficult to conceptualize. We propose people represent money as a set of products. As a result, they value it according to
the set of purchases it allows, not just the marginal purchase it allows, implying that irrelevant alternatives may influence
its perceived value. We test this hypothesis by adding less attractive uses to a set (or merely making them more accessible)
and showing that these less attractive options influence the perceived value of the medium.


Phantom Firsts (LaBoeuf) & Imminent
Ends (O'Brien)
Sequence Matters
Placement can be manipulated, so that the middle is
treated as a ﬁrst or last item
Forceful Phantom Firsts: Framing Experiences as Firsts Amplifies their Influence on Judgment
LeBoeuf, Robyn A. (University of Florida); Williams, Elanor F. (University of Florida); Brenner, Lyle (University of Florida)
First experiences are highly influential. We show that experiences merely framed as firsts can also disproportionately influence judgment. In
five experiments, one piece of a series of information was framed to appear to have ‘first’ status: for example, the final review in a sequence
of hotel reviews happened to be 2010’s first review. Such information had greater influence on subsequent judgments (e.g, of the hotel’s
quality) than it did when it was not framed as a first. Results further suggested that this effect arises largely because ‘phantom first’
information receives greater weighting (and not merely more attention) than other information.

Imminent ends make eminent friends: Last bites of food, last beats of music, and last job applicants are judged better than others
O'Brien, Edward H. (University of Michigan); Ellsworth, Phoebe C. (University of Michigan)
People make more positive judgments during 'last' experiences of significant life events (e.g., graduation). Five experiments extended this to
everyday experiences. Chocolates, cake, and songs were judged better when they were described as the 'final' of a series. Moreover, 'final'
individuals in groups of neutral targets and job applicants were judged better than others. These findings extend the power of endings to
everyday lifethe same foods, music, and people are judged best when they happen last. Findings are discussed in terms of maximizing the
efficacy of marketing campaigns, consumption experiences, and social impressions, but also more troubling implications.


Signing ﬁrst increases honesty (Lisa Shu)

Signature at the top of a form reduces the amount
of claimed expenses
Factor could be something other than dishonesty
(e.g., changing the framing shifts attention, rather
than explicitly reducing tendency to deceive)

When to Sign on the Dotted Line? Signing First Makes Ethics Salient and Decreases Dishonest Self-Reports
Shu, Lisa L. (Harvard University); Mazar, Nina (University of Toronto); Gino, Francesca (Harvard University); Ariely, Dan
(Duke University); Bazerman, Max H. (Harvard University)
We examine the effectiveness of a simple measure to reduce dishonesty: signing on the dotted line. This intervention has
relevance for transactions that rely on truthful self-reporting, such as taxes, business expenses, and billable hours. Departures
from honest reporting can lead to significant economic losses. Using both lab and field experiments, we find that signing before,
rather than after, a self-report task raises the saliency of ethics, and subsequently decreases dishonesty. Signing on the dotted
line shifts the moral gaze inward, raising the saliency of ethical standards and promoting more ethical actions going forward.


Other fun findings

Budget Backfire - switches attention to quality, and
drives overspending (Jeff Larson, BYU)
Reported compliance in midwife practice (100%)
differed from observed (4%) (U of Gloucestershire)
Ubiquity of Mechanical Turk (~1/3 of studies)
Incidental conflict more motivating - FOMOx2 (Liad
Weiss, Columbia)
False Positive Psychology (Joe Simmons & Leif


�ݺ�ߣ

Habit Design #4 12012011

More Related Content

Habit Design #4 12012011