News Release

Carnegie Mellon experiment reveals impact of fear, anger on American perceptions of terrorism

Peer-Reviewed Publication

Carnegie Mellon University

PITTSBURGH - A national field experiment by Carnegie Mellon University scientists on American emotions and perceptions of the risk of terrorist threats following September 11 reveals a national psyche influenced in opposite ways by fear and anger.

The experiment results may have implications for better understanding of consumer behavior, the role of the media, and public support for the war on terrorism, said Jennifer Lerner, an assistant professor of Social and Decision Sciences at Carnegie Mellon and the lead author of a paper in press with the journal Psychological Science.

The experiment results also were reported to NATO officials during a recent meeting in Brussels that focused on terrorism.

The Carnegie Mellon team drew four major conclusions from the study:

  • 1) Americans who experience anger are more optimistic about the future, less likely to take precautionary actions, and more likely to favor aggressive policy responses than those who experience fear.

  • 2) Individuals see themselves as less vulnerable than the "average American," while still perceiving strikingly high personal risk in the wake of September 11.

  • 3) Men experience more anger about terrorism than women, leading them to be more optimistic than women.

  • 4) Media portrayals of the terrorist attacks strongly influence emotional responses, producing anger in some instances and fear in others.

Details of the study are included in the background fact sheet below.

The experiment involving nearly 1,000 American men and women ages 13 to 88, suggests that heightened emotions of fear and anger affect responses to the threat of terror currently facing the nation, with anger promoting greater optimism and more aggressive policies.

Feelings of fear likely fueled the sense of pessimism that contributed to the national economic downturn after September 11 and the call for tighter security; feelings of anger likely fed the sense of optimism, contributing to support for military action and the sense that threats could be controlled.

Scientifically, the experiment plows new ground. Scientists say that this is the first time that the effects of emotion have been studied in a national sample, using the random assignment to conditions of fear or anger.

Because emotions often affect economic decisions and the formation of policies, team members stressed the importance of undertaking further studies like this one.

"Citizens need to understand these processes in order to apply their hearts and minds to what might be a protracted struggle with the risks of terror," said Carnegie Mellon University Professor of Social and Decision Sciences Baruch Fischhoff, a member of the scientific team.

In addition to Lerner and Fischhoff, the team included doctoral students Roxana Gonzalez and Deborah Small, all of Carnegie Mellon's Department of Social and Decision Sciences.

Grants from the National Science Foundation and the American Psychological Association funded the study.

The meeting at NATO Headquarters, where the findings were presented, was the first scientific workshop co-sponsored by NATO and Russia to gather scientists from around the world to consider the psychological and social consequences of chemical, biological, and radiological terrorism - and to advise policy makers on prevention and mitigation measures. A preliminary summary of the meeting and this report appears on the NATO Web site,

Details of the Scientific Experiment Dealing with Emotional Response to Terrorism

In contrast to the common view that negative emotions lead to pessimism, the researchers hypothesized that the negative emotion of anger would lead to optimism, relative to the negative emotion of fear. They also hypothesized that simply asking people to reflect on fear or anger while viewing a fear- or anger-inducing media clip would elicit emotions that were strong enough to shape perceptions of twenty different risky events. Finally, they hypothesized that males would experience more anger and females would experience more fear, leading males to make relatively optimistic risk estimates. These hypotheses were tested and supported in a national field experiment with almost 1,000 American citizens, ages 13 to 88. The sample's demographics corresponded to those of the U. S. Census, so the results would generalize to the U.S. population. Using WebTVs supplied by the research corporation Knowledge Networks, the project initially asked respondents about their reactions only nine days after the attacks. Eight weeks later, using TV imagery and newspaper reports from major media organizations (e.g., CNN, the New York Times) broadcast on the Web TVs, the researchers surveyed the same people again. For this second survey, half of the sample was exposed to a fear-inducing media clip, while the other half was exposed to an anger-inducing clip.

Anger Led to Optimism; Fear Led to Pessimism

Carnegie Mellon researchers found that Americans randomly assigned to the "fear condition" perceived greater risks from terrorism, while those in the "anger condition" perceived less risk. "Brief reminders of media stories elicited emotions that shaped Americans' perceptions of their own level of risk. Stories that induced fear increased their perception that they would be hurt in a terrorist attack, while stories that induced anger reduced their perception of personal risk," Lerner explained. She added that the differential effects of fear and anger were not limited to emotions induced experimentally. Naturally-occurring fear and anger measured in the week after the attacks had the same pattern as the experimentally-induced fear and anger. "Regardless of whether we randomly exposed people to emotion-inducing media stories or if we measured naturally-occurring emotions, greater anger led to greater optimism," Lerner said.

Fear, Anger Trigger Different Precautionary Responses and Policy Preferences

Fear and anger not only produced different risk perceptions, but also different precautionary responses and different policy preferences. The Carnegie Mellon scientific team contends that these findings have important implications for the health of the U.S. economy and public support for the war against terrorism. Americans who saw a fear-inducing media story were more likely to say that they would take personal precautions, such as reducing their air travel. Americans who saw a fear-inducing media story were also more likely to support conciliatory public policies. By contrast, Americans who saw an anger-inducing story were less likely to say they would take precautions and less likely to support conciliatory policies. Overall, Americans strongly supported the public policies of "providing Americans with honest, accurate information about the situation, even if the information worries people," "investing in general capabilities, like stronger public health, more than specific solutions, like smallpox vaccinations," and "deporting foreigners in the U.S. who lack visas." There was somewhat weaker, but still positive support for "strengthening ties with countries in the Moslem world."

Men Perceive Less Risk than Women Because They Are Angrier

The Carnegie Mellon study also discovered that males (ages 13-88) were less pessimistic about risks than were females-because they were angrier. "The striking difference in risk perception between males and females is due to males experiencing greater anger and females greater fear," Lerner said.

Americans Perceived High Risk of Terrorism-But Also Say Risk Is Higher for "Other" Americans than for Themselves

The experiment found that subjects saw "average Americans" as facing much higher risks than they did. However, researchers said these results did not reflect unrealistic personal optimism. According to Lerner, many risk estimates last November reflected profound pessimism. On average, respondents saw a 21 percent chance of being injured in a terrorist attack within the next year as opposed to the 48 percent chance assigned to the average American. "This is still a very gloomy view," Lerner commented. Respondents had realistic expectations about more-everyday happenings, such as the likelihood of getting the flu. Carnegie Mellon scientists say a follow-up study will assess whether Americans' estimates of risk will have changed a year later, as well as examining the accuracy of their initial risk estimates.

Study Advances New Experimental Methods for Examining Emotion and Judgment

Many studies have looked at correlations between emotional responses and risk perceptions. However, no studies with a national sample have experimentally manipulated emotions. According to the researchers, only experimental manipulation, with random assignment to condition, allows one to conclusively examine causal relationships. Other national studies have typically involved correlational designs without experimental manipulations and random assignment, while experimental studies have typically been conducted only in the laboratory. Scientists say the Carnegie Mellon study breaks new ground by marrying the virtues of both methods: It takes experimental methodology outside the laboratory to a nationally representative sample of Americans.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.