Tag Archives: Fuzzy-Trace Theory

Reprinted from the Cornell Chronicle, Oct. 31, 2016

 

Researchers in the social sciences have been searching for a holy grail: an accurate way to predict who is likely to engage in problematic behavior, like using drugs. Over the years experts in economics, psychology and public health have designed hundreds of questionnaires in an attempt to understand who will binge drink or have unprotected sex – and why.

reyna_valerie_web

Valerie Reyna, Professor in Human Development

Valerie Reyna, professor of human development, and Evan Wilhelms,

Evan Wilhelms, Ph.D., Cornell Graduate School, '15

Evan Wilhelms, Ph.D., Cornell Graduate School, '15

Ph.D. ’15, of Vassar College, have just taken a sizable step toward answering those questions.

In a new study, Reyna and Wilhelms have debuted a new questionnaire that significantly outperforms 14 other gold-standard measures frequently used in economics and psychology. The measure’s 12 simple questions ask in various ways whether one agrees with the principle “sacrifice now, enjoy later.” Their study, “Gist of Delay of Gratification: Understanding and Predicting Problem Behaviors,” appeared Aug. 10 in the Journal of Behavioral Decision Making.

“People who get drunk frequently, party with drugs, borrow money needlessly or have unprotected sex disagreed more with the concept ‘sacrifice now, enjoy later’ than people who didn’t do these things,” Reyna said. “Instead, they leaned more toward ‘have fun today and don’t worry about tomorrow.’”

Having fun is generally good, she said. “But not being able to delay gratification can interfere with education, health and financial well-being, and the impact is greater for young people,” she added.

The questionnaire is based on Reyna’s fuzzy-trace theory. It says people boil down their personal values into a simple, qualitative “gist” of an idea – such as “sacrifice now, enjoy later.” When they have to make a life decision, they retrieve that gist and apply it to their situation.

In contrast, prevailing theory, with many questionnaires based on it, says that people make specific, quantitative trade-offs known as “delay discounting.” For example, those measures ask questions like, “Would you like $10 now or $11.50 in a week?”

“People do size up the trade-off, but they don’t make their decisions on that analysis,” Reyna said. “They think, ‘sacrifice now, benefit later.’ And therefore they study for the exam rather than go out to the party. It’s not about the party per se. It’s about the life principle.”

The researchers conducted four studies to get their results, comparing the measure, the Delay-of-Gratification Gist Scale, against 14 others. The Gist Scale’s questions include, “I wait to buy what I want until I have enough money,” “I think it is better to save money for the future” and “I am worried about the amount of money I owe.” Money is used as a “stand-in” or proxy for tempting rewards.

The first study asked 211 college students to take the Gist Scale and other measures that predict poor financial outcomes. The second and third studies, with 845 and 393 college students, respectively, compared the new measure against others involving delay discounting. With 47 teens and adult participants, the fourth study compared the Gist Scale against a widely used measure of impulsivity.

The Gist Scale is not only more accurate, it’s also shorter and simpler – some other measures are more than twice as long. It is also gender and age neutral, meaning it can be taken by anyone.

Reyna points out that cultures all over the world have aphorisms that encourage the ability to delay gratification. That skill can improve with practice, she said.

“Sometimes we send young people very mixed messages about struggle. I think it’s extremely important for them to know that struggle and pain are part of life and to be expected,” she said. “Staying the course, keeping your eyes on the prize – these values make a difference. And they can be taught and they can be practiced.”

By Nora Rabah, Allison M. Hermann, Thomas W. Craig, and David Garavito

David Garavito, graduate student in the Law, Psychology, and Human Development Program, under the supervision of Dr. Valerie Reyna, is working with communities in New York and around the country with support from an Engaged Cornell grant for student research. He is working with coaches and student athletes to study the effects of concussions on decision making about risks. For his dissertation research, Garavito is developing a model based upon Dr. Reyna’s Fuzzy Trace Theory (FTT) which integrates research on Alzheimer’s dementia (AD), behavioral economics and decision making, and neuroscience to study the perception of risks associated with sports-related concussions among people vulnerable to chronic traumatic encephalopathy (CTE).

Garavito and undergraduates in the Laboratory for Rational Decision Making are working with a growing number of students, coaches, and administrators from high schools and colleges in New York (including Watkins Glen and Moravia high schools, Cornell University, and Ithaca College), Colorado, and Minnesota. Engagement with sports communities has provided the team with the opportunity to educate--and listen to-- the public about current research on concussions and how values or principles can affect perceptions and decisions about concussion risk. Garavito has found that coaches are very supportive of research projects that aim to help keep athletes safe and further knowledge about concussions. Many athletes have enthusiastically agreed to volunteer in Garavito’s studies. The student team also has been working with the Ithaca Youth Bureau and the experiences of coaches and educators at the center have been essential in the development of interactive activities to teach youth about the brain and concussions. The ultimate goal of the concussion intervention is to strengthen healthy values and educate people about risks, and the importance of reporting symptoms of concussions.

Chronic traumatic encephalopathy is a form of dementia, like Alzheimer’s, which results in an accumulation of tau proteins in the brain, however, the onset and progression of CTE is related to a history of concussions. Athletes in contact sports are particularly vulnerable to CTE because many athletes fail to report concussions and their symptoms – a very risky decision that could result in brain damage and cognitive impairment. Using the FTT framework, Garavito and his team of undergraduates are studying this underreporting phenomenon.

Although current research on the underreporting of concussions has brought about the creation of laws mandating concussion education nationwide, research based on FTT has shown that not all education programs are equal. How people process information can have a profound effect on how they make decisions (Mills, Reyna, Estrada, 2008; Widmer, Wolfe, Reyna et al., 2015). Dr. Reyna’s research has shown that adolescents are more likely to rely on verbatim, surface-level details, whereas adults tend to rely on qualitative reasoning and the bottom-line gist of information (Reyna & Farley, 2006; Reyna, Estrada, DeMarinis et al., 2011). For example, if told an athlete has a 2/3 chance of having another concussion if they go back out on the field after having had a concussion on the same day, adolescents make their decisions about risk by considering the numbers and take the chance in order to play more sports because the benefits, to them, outweigh the risks. Adults, on the other hand, get the point that the mere possibility of a catastrophic injury, no matter how small, of getting another concussion (which could result in permanent brain damage or death), is not worth the risk for more playing time. This difference in information processing has not been studied in the underreporting or the perception of risks in sports-related injuries.

Many educational programs emphasize the acquisition of verbatim fact-based knowledge in the hope that these details will help the public understand and make better decisions. Unfortunately, this can lead to the opposite effect – giving people more detailed information causes them to engage in greater precise deliberation and leads them to take unnecessary risks. Football players, for example, may know verbatim facts about the symptoms of concussions, but still “gamble” by not reporting their symptoms, instead of choosing the “sure thing” of being safe and reporting them. This risky decision-making among athletes, in turn, is exacerbated by impairment from prior concussions.

Currently, Garavito, Dr. Reyna, and their team of undergraduates, are using scales based on FTT, to test several important hypotheses. These scales are sensitive measures that can detect if a person is relying more on categorical than fact-based thinking. Fuzzy Trace Theory predicts that categorical or gist-based thinking is more developmentally advanced and can deter people from taking dangerous risks. Garavito hypothesizes that adolescents affected by cumulative concussions may rely less on categorical thinking than non-concussed adolescents. This could lead concussed adolescents to engage in greater risk-taking, in general. Garavito and Reyna are studying whether FTT measures can cue developmentally advanced categorical thinking. Cuing adolescents to engage in categorical thinking will lead them to approach dangerous risks like adults, and is consistent with Dr. Reyna’s research on other types of risk-taking behavior (Reyna, Wilhelms, McCormick et al., 2015).

Nora Rabah is a Biology and Society major in the College of Agriculture and Life Sciences.

Allison M. Hermann is the Research and Outreach Manager for the Laboratory for Rational Decision Making.

Thomas W. Craig is the Law, Psychology and Human Development Program Assistant.

David Garavito is a graduate student in the Law, Psychology, and Human Development Program at the College of Human Ecology.

HD-Today e-Newsletter, Summer 2016 Issue

By Allison M. Hermann, Ph.D.

LRDM lab members and 4-H Career Explorations students

LRDM lab members and 4-H Career Explorations students

The Laboratory for Rational Decision Making (LRDM), led by Dr. Valerie Reyna in Human Development, welcomed 24 high school students from 18 different counties throughout New York State as part of a 3-day course in decision making research called, “Getting the Gist.” The high school students journeyed to Cornell University as part of the 4-H Career Explorations Conference that offers secondary school students the opportunity to attend courses and workshops and learn about STEM research.

get-the-gist-add

James Jones-Rounds, Lab Manager of the HEP Lab

The high school students became guest LRDM lab members and learned how to turn their questions about risky decision making into experiments. They created an experiment, collected and analyzed the data, and discussed the results. The student career explorers also toured the Center for Magnetic Resonance Imaging Facility and the EEG and Psychophysics Laboratory and saw how decision research uses brain imaging technologies to examine what brain areas are activated when making risky decisions.

Dr. Reyna’s graduate students' David Garavito, Alisha Meschkow and Rebecca Helm, and research staff member, Bertrand Reyna-Brainerd, presented lectures on Dr. Reyna’s fuzzy trace theory and research design and led interactive discussions with the visiting students about the paths that led the graduate students to the LRDM at Cornell. In addition, three undergraduate members of the lab, Tristan Ponzo (’18), Elana Molotsky (’17) and Joe DeTello (’19) delivered poster presentations of current lab research projects. Feedback from one of the career explorers expressed the gist of the program, “Yes, I definitely feel like I have a better understanding of why I make the decisions I do.”

For juries awarding plaintiffs for pain and suffering, the task is more challenging – and the results more inconsistent – than awarding for economic damages, which is formulaic. Now, Cornell social scientists show how to reduce wide variability for monetary judgments in those cases: Serve up the gist.

As an example of gist, juries take into account the severity of injury and time-scope. In the case of a broken ankle, that injury is a temporary setback that can be healed. In an accident where someone’s face is disfigured, the scope of time lasts infinitely and affects life quality. In short, “meaningful anchors” – where monetary awards ideally complement the context of the injury – translate into more consistent dollar amounts.

Valerie Reyna

Valerie Reyna

“Inherently, assigning exact dollar amounts is difficult for juries,” said Valerie Reyna, professor of human development. “Making awards is not chaos for juries. Instead of facing verbatim thoughts, juries rely on gist – as it is much more enduring. And when we realize that gist is more enduring, our models suggest that jury awards are fundamentally consistent.”

The foundation for understanding jury awards lies in the “fuzzy trace” theory, developed by Reyna and Charles Brainerd, professor of human development. The theory explains how in-parallel thought processes are represented in your mind. While verbatim representations – such as facts, figures, dates and other indisputable data – are literal, gist representations encompass a broad, general, imprecise meaning.

VHans Chronicle

Valerie Hans

“Experiments have confirmed the basic tenets of fuzzy trace theory,” said Valerie Hans, psychologist and Cornell professor of law, who studies the behavior of juries. “People engage in both verbatim- and gist-thinking, but when they make decisions, gist tends to be more important in determining the outcome; gist seems to drive decision-making.”

In addition to authors Reyna and Hans for the study, “The Gist of Juries: Testing a Model of Award and Decision Making,” the other co-authors include Jonathan Corbin Ph.D. ’15; Ryan Yeh ’13, now at Yale Law School; Kelvin Lin ’14, now at Columbia Law School; and Caisa Royer, a doctoral student in the field of human development and a student at Cornell Law School.

The research was funded by grants from the National Institutes of Health, Cornell’s Institute for the Social Sciences, the Cornell Law School and Cornell’s College of Human Ecology.

Update: On Sept. 1, 2015, the National Science Foundation awarded a grant for $389,996 to Cornell for support of the project “Quantitative Judgments in Law: Studies of Damage Award Decision Making,” under the direction of Valerie P. Hans and Valerie F. Reyna.

Flip a coin, roll the dice, pick a number, throw a dart -- there is no shortage of metaphors for the perceived randomness of juries. That belief that "anything can happen," is applied to some extent to the whole jury model, but nowhere more so than on the topic of civil damages. When jurors are asked to supply a number, particularly on the categories that are intangible (like pain and suffering) or somewhat speculative (like lost earnings), the eventual number is sometimes treated as a kind of crapshoot. What's more, the process that the jury goes through to get to it is seen as a kind of black box: not well understood or necessarily subject to clear influence.

What litigators might not know is that the subject of civil damages is a great example of social science research beginning to close the gap. Based on research within the last decade, we are coming closer to opening that black box in order to see a jury process -- albeit not fully predictable, but at least more knowable. The way jurors arrive at a number is increasingly capable of being described through the literature, and these descriptions have implications for the ad damnum requests made by plaintiffs and for the alternative amounts recommended by defendants. A study just out this year (Reyna et al., 2015), for example, provides support for a multistage model describing how jurors move from a story, to a general sense or 'gist' of damages, and then to a specific number. Their work shows that, while the use of a numerical suggestion -- an 'anchor' -- has a strong effect on the ultimate award amount, that effect is strongest when the anchor is meaningful. In other words, just about any number will have an effect on a jury, but a meaningful number -- one that provides a reference point that matters in the context of the case -- carries a stronger and more predictable effect. This post will share some of the results from the new study framed around some conclusions that civil litigators should take to heart.

No, Jurors Aren't Random

The research team, led by Cornell psychologists Valerie Reyna and Valerie Hans, begins with the observation that "jurors are often at sea about the amounts that should be awarded, with widely differing awards for cases that seem comparable." They cover a number of familiar reasons for that: limited guidance from the instructions; categories that are vague or subjective; judgments that depend on broad estimations, if not outright speculation; and a scale that begins at zero but, theoretically at least, ends in infinity. Add to that the problem that jurors often have limited numerical competence (low "numeracy") or an aversion to detailed thinking (low "need for cognition"). All of that means that there isn't, and will never be, a precise and predictable logic to a jury's damages award. But it doesn't mean that they're picking a number out of a hat either. Research is increasingly pointing toward the route jurors take in moving to the numbers.

There is a Path Jurors Follow

As detailed in the article, jury awards will vary on the same essential case facts, but a number of studies have found a strong "ordinal" relationship between injury severity and award amounts: More extreme injuries lead to higher awards, and vice versa. Increasing knowledge of the process has led researchers to describe a jury's decision in steps, both to better understand it, and also to underscore that some steps are more predictable than others. The following model, for example, is drawn from Hans and Reyna's 2011 work (though I've taken the liberty to make some of the language less academic):

Slide1

That breakdown of steps is backed up by the research reviewed in the article, but it isn't just useful for scholars. Watch mock trial deliberations and you are likely to see jurors moving through that general sequence. Knowing that jurors are going to implicitly or explicitly settle on a 'gist' (large, small, or in between) before translating that into a specific number is also helpful to litigators in providing the reminder 'speak to the gist' as well as to the ultimate number.

So Use an Anchor, but Make it Meaningful

The model also points out the advantages of giving jurors guidance on a number. Research supports the view that it is a good idea to try to anchor jurors' awards by providing a number. Suggesting a higher number generally leads to a higher award, and vice versa. Reyna and colleagues found that even the mention of an arbitrary dollar amounts (the cost of courthouse repairs) influences the size of awards. But the central finding of the study is that it helps even more if the dollar amount isn't arbitrary. "Providing meaningful reference points for award amounts, as opposed to only providing arbitrary anchors," the team concludes, "had a larger and more consistent effect on judgments." Not only are the ultimate awards closer to the meaningful anchor, but they are also more predictable, being more tightly clustered around the anchor when that anchor is meaningful and not arbitrary.

And Here is what "Meaningful" Might Mean

Of course, the notion of what is "meaningful" might carry just as much vagueness as the damages category itself, and the research team is not fully explicit on what makes a number meaningful in a trial context. That might be a fitting question for the next study, but in the meantime, the team gives at least some guidance. To Reyna et al., recommended amounts are "meaningful in the sense that their magnitude is understood as appropriate in the context of that case." In the study, the "meaningful" anchor was that one that expressed a pain and suffering amount as being either higher or lower than one year's income. That's not a perfect parallel -- one is payment for work, the other is recompense for suffering -- but it does reference something that the jury is used to thinking about and using as a rough way of valuing time. By nature, meaningful anchors will vary from case to case, but there are a few general numbers that could be used as a reference point, like median annual salary; daily, weekly, or monthly profits; or remediation costs.

Of course, those are well-known devices that attorneys will regularly apply. Still, it is good to know that, at a preliminary level at least, they have the social science stamp of approval. Those kinds of anchors and reference points work because they give jurors a way to get a gist of the claimed damages, and a way to bring abstract numbers into the jurors' own mental universe.

Copyright Holland & Hart LLP 1995-2015.

By H. Roger Segelken
Reprinted from Cornell Chronicle, December 16, 2014

When the doctor says, “I could prescribe antibiotics for your sniffles, but it’s probably a virus – not bacterial,” do you decline? Many patients expect antibiotics, although overprescription is a major factor driving one of the biggest public health concerns today: antibiotic resistance.

Now researchers at Cornell, George Washington and Johns Hopkins universities have figured out why: “Patients choose antibiotics because there’s a chance [prescription medications] will make them better, and they perceive the risks of taking antibiotics as negligible,” says Cornell psychologist Valerie Reyna.

With her co-authors, the professor of human development has published new research with important implications for communicating about antibiotics: “Germs Are Germs, and Why Not Take a Risk? Patients’ Expectations for Prescribing Antibiotics in an Inner-City Emergency Department,” in the journal Medical Decision Making.

That’s encouraging news for health educators, Reyna says, noting: “Patients might expect doctors to prescribe antibiotics because patients confuse viruses and bacteria – and think antibiotics will be effective for either. Most educational campaigns attempt to educate patients about this misconception. However, we found fewer than half of patients in an urban ER agreeing with the message, ‘germs are germs.’”

Patients who understand the difference between viruses and bacteria – and take antibiotics anyway – are making a strategic risk assessment, Reyna says: “Our research suggests that antibiotic use boils down essentially to a choice between a negative status quo – sick for sure – versus taking antibiotics and maybe getting better. This risk strategy promotes antibiotic use, particularly when taking antibiotics is perceived as basically harmless.”

Fuzzy-trace theory

The Broniatowski-Klein-Reyna study is the first to apply “fuzzy-trace” theory to how people think about antibiotics. The theory predicts that patients make decisions based on the gist (or simple bottom line) of information.

As Reyna explains: “The goal is to make better decisions, getting antibiotics to patients who need them but not overusing them so the rest of the public is safe. Understanding how patients think is crucial because their expectations influence doctors’ decisions.”

Adds David Broniatowski, assistant professor of engineering management and systems engineering at GWU, and the report’s first author: “We need to fight fire with fire. If patients think that antibiotics can’t hurt, we can’t just focus on telling them that they probably have a virus. We need to let them know that antibiotics can have some pretty bad side effects, and that they will definitely not help cure a viral infection.”

The third author is Dr. Eili Klein, assistant professor in the Department of Emergency Medicine at the Johns Hopkins University and a fellow at the Center for Disease Dynamics, Economics and Policy.

Reyna is the director of the Human Neuroscience Institute, co-director of the Cornell University Magnetic Resonance Imaging Facility, and a co-director of the Center for Behavioral Economics and Decision Research, all in the College of Human Ecology. She is a developer of “fuzzy-trace theory,” a model of the relation between mental representations and decision making that has been widely applied in law, medicine and public health.

The study was supported, in part, by funds from the National Institutes of Health and the U.S. Department of Homeland Security.