Darmouth Faculty Reserach and Scholarship Today
  Darmouth College

Features Dartmouth Faculty Research and Scholarship Today
Punam Keller Sending the Right Message, written by Anita Warren

Effects of depression on risk estimates


Punam Keller has a joint appointment at the Tuck School of Business, where she is the Charles Henry Jones Third Century Professor of Management, and at the Dartmouth Medical School (DMS), as adjunct professor of community and family medicine.

Can You Name Five Black Women Business Leaders?

Space Weather Starts at the Sun

Health Care by the Numbers (cover feature)

Can a Robot Predict Space Weather?

Bringing Data Retrieval into the 21st Century

Captivated by Ecological Complexities

In the Eyes of the Beholder

The Foundation for a Livable Future

Better communication and improved market segmentation by healthcare providers could mean healthier patients.

Do you see a glass as half full or half empty? Your answer may affect the way you receive healthcare in the future, according to Punam Keller's research.

Most of Keller's work over the past decade has involved applying basic marketing principles to understand how people make health-related decisions. Her work suggests ways the healthcare community can communicate more effectively by tailoring its marketing messages to different target audiences.

"You don't want to send the same communication or the same package to everyone as if they were all the same, which is what happens now," she says.

In one recently published study, Keller discovered that communicating the benefits of having mammograms may be less persuasive for some women than communicating the costs of not having them. In this study, women were asked to estimate their risk of developing breast cancer, then given more detailed risk information, and queried again. Keller found women who experienced dysphoria, a type of depression, were more receptive than nondysphoric women to the information given to them.

Dysphoric women tend to have more negative attitudes, so one might expect them to accept information that reinforces their depressed outlook on life. In Keller's study, however, participants were told their risk of developing breast cancer was lower than they had estimated. In an unexpected twist, the dysphoric women accepted this positive information and adjusted their risk estimates downward, but the nondysphoric women maintained their original high estimates.


EFFECT OF DEPRESSION ON RISK ESTIMATES. The first bar chart contains the risk estimates women (both depressive and nondepressive) gave for their chances of getting breast cancer before and after being given risk information. The second shows the results following the second experiment when women were given only standard risk information. In both cases, women who were depressive were more likely to correctly assess their risk once given accurate information. (Keller, P.A. et al. "Depressive Realism and Health Risk Accuracy: The Negative Consequences of a Positive Mood. 2002. Journal of Consumer Research, p.61.)

Keller believes the positive-oriented women resisted the information partly because they had more to lose. "Part of it was mood," she says. "The nondepressed women were in positive moods, and a lot of the literature indicates that people in a positive mood do not pay attention to information because they're afraid it will spoil their mood. In this case, it was positive information, and they still wouldn't pay attention. It was very counterintuitive. That's what makes a study fascinating."

The findings provided Keller with insight into how people think about risk-insight she has applied to another study underway, this one on adolescent decision-making. She and Ardis Olson, associate professor of pediatrics and of community and family medicine at the Dartmouth Medical School and primary investigator in the study, examined use of sunscreen, bike helmets, and seat belts, as well as eating habits and exercising, among sixth and eighth graders at five local schools. The study groups' behaviors were then compared to behavior by their parents and peers, and a model was developed to look at how risk-related decisions were made inthose five areas.

Keller and Olson found sixth graders were more influenced by their parents, and eighth graders were more influenced by their friends. Boys were more influenced by their parents, especially their fathers, and girls were more influenced by their friends. If a father uses sunscreen, for instance, the son is more likely to use it. If the mother uses it, says Keller, "it has no impact whatsoever. Girls want to see their friends use it."

The conflict adolescents experience between parental and peer influence and the uncertainty of their status between childhood and adulthood factor largely into their evaluation of risk. To a fourteen-year-old girl, the risk of getting skin cancer from sunbathing with friends is small in comparison to the risk of losing those friends by refusing to lie out in the sun with them. To a twelve-year-old boy, developing skin cancer when he's sixty is a distant risk compared to the probability of contracting another ailment, say, mononucleosis, in the more immediate future. This interplay of emotions and context in decision-making further clarifies why the one-size-fits-all approach to healthcare communication so often doesn't work.

In another study, Keller is collaborating with researchers at several other institutions to examine how orientation toward promotion or prevention affects willingness to risk change. Typically, promotion-oriented individuals focus on reaching goals, while prevention-oriented individuals focus on preventing bad things from happening. Behavioral literature suggests promotion-oriented people are always most likely to change and take higher risks than prevention-oriented people. Keller and her group want to see the conditions under which they can convince people who are prevention-oriented to switch.

The researchers ask college students enrolled in an exercise class whether they want to stay with their current program or change to a new one that includes a new exercise machine. Several completed experiments have revealed the promotion-oriented students look at things more abstractly. Because they're goal- focused, their decision is less about using an elliptical machine, cross-trainer, or treadmill and more about maximizing general fitness. By contrast, prevention-oriented people want to know what they would have to learn, what machine prevents the most injuries, and what option would minimize injuries.

"We don't know yet why they look at things differently," says Keller. "One hypothesis is that people who have a promotion focus are very much like optimists, and the reason optimists are optimistic is that they never really evaluate anything. The reason prevention-oriented people look at these lower levels - more concrete, tangible, something easily measured versus achieving 'overall fitness' - is because they're more concerned with outcomes and implementation.

There are other possible explanations. Is the task more difficult? Because if you have to do it at a concrete level, a task is much harder than if you're doing it abstractly. Promotion-oriented people are like people in a good mood; they don't want to process detailed information. That's too difficult, so they ignore it. They just look at the more general, surface level information. Prevention-oriented people are more like pessimists. They're more willing to look systematically at details. That's a different explanation from just not wanting to evaluate anything."

Ultimately, healthcare firms and professionals could use this information to reach consumers and patients better by framing information differently for different types of individuals. Knowing whether a person is oriented toward prevention or promotion, for instance, would enable pharmacies to customize the instruction sheets distributed with prescription drugs. And that, says Keller, could increase compliance.

"The key always becomes how much to customize and how much not to. There are costs associated with customizing," she notes. "But I think if you don't customize at all, you have a problem."

Keller hopes over time to accomplish two objectives. First is to combine a typology of individual differences (e.g., level of depression, gender, age) with various health contexts (e.g., safe sex, smoking, drinking) and determine what messages work best to predict future behavior. Second, she hopes her work will encourage a more systematic process of investigation and testing among policymakers who implement healthcare programs - "to help them make some choices and, hopefully, come up with more effective programs," she says. "That would make me feel very good."


Ultimately, healthcare firms and professionals could use this information to reach consumers and patients better by framing information differently for different types of individuals.

  Dartmouth Faculty Research and Scholarship Today    
Home | Features | Essays | About | Archive | Contact | Scholarship Now

Copyright © 2003 Trustees of Dartmouth College