Developing effective risk
communication:
How to find out what the audience needs
Wändi Bruine de Bruin
Centre for Decision Research, Leeds University Business School (UK)
Department of Engineering & Public Policy, Carnegie Mellon University (US)
Risk communication
• The goals of risk communication include
– Improving public understanding
– Promoting behavior change to reduce risks
and to increase public disaster preparedness
• To evaluate whether these goals are
achieved, we can conduct
– Randomized controlled trials
– Meta-analyses
Features of effective communications
1. Content reflects scientific knowledge of
diverse experts, ensuring
– Accuracy
– Balance
2. Content is based on research (interviews
and survey) with members of the
intended audience, increasing likelihood
of
– Using wording that recipients understand
– Covering what recipients need to know
Research with intended audience
is often omitted
• Experts often design communications
without finding out what their audience
needs
• Potential reasons include
– Domain experts are overconfident about how
well they know their non-expert audiences
– Existing communications are often not
evaluated, so experts never find out whether
(or not) their communications were effective
– Domain experts may lack the training to
conduct interviews and surveys
Failing communications
• The few communications that are
evaluated often turn out to be ineffective
• When designing communications without
input from the intended audience, experts
– Use difficult wording
– Omit information their audience needs to
make and implement informed decisions
Example: difficult wording
• When describing flood risk and protection,
experts talk about a “100-year flood”
• Non-experts expect “100 year flood” to
happen at regular 100-year intervals
• It may be better to describe flood as “1%
chance per year”
• However, note that
people worry more
about flood levels
than about
flood frequency
Example: difficult wording
• When describing flood risk and protection,
experts talk about a “100-year flood”
• Non-experts expect “100 year flood” to
happen at regular 100-year intervals
• It may be better to describe flood as “1%
chance per year”
• However, note that
people worry more
about flood levels
than about
flood frequency
Example: difficult wording
• When describing flood risk and protection,
experts talk about a “100-year flood”
• Non-experts expect “100 year flood” to
happen at regular 100-year intervals
• It may be better to describe flood as “1%
chance per year”
• However, note that
people worry more
about flood levels
than about
flood frequency
Example: Omitting information
Example: Omitting information
Encouraging people to wash their hands will
be ineffective if they don’t know how to
wash their hands
Mental Models Approach
1. Expert model: What should people know?
– Conduct interdisciplinary literature review
– Convene expert panel
– Conduct risk analysis
2. Lay model: What do people already know?
– Conduct interviews to identify beliefs, barriers to
behavior change, relevant wording
– Conduct follow-up surveys to examine prevalence of
interviewees’ beliefs
3. Communication: What do people still need to know?
– Compare expert model and lay model
– Address misunderstandings and other barriers to
behavior change in interviewees’ preferred wording
4. Evaluation: Does the communication work?
– Conduct validation study
Interview and survey procedure
• Interviews
– Are designed to identify people’s beliefs and barriers
to behavior change in their preferred wording
– Start with “Tell me what you know about..” and follow
up with “can you tell me more about..”
– Are repeated until no more new ideas emerge (n≈20)
• Surveys
– Are less labor-intensive than interviews, and better for
use with larger samples
– Can be conducted by mail or online
– Ask participants to rate agreement with interviewees’
beliefs and self-report behaviors
(Bruine de Bruin & Bostrom, PNAS, in preparation)
Sexually Transmitted Infections
(STIs)
• Goal: To reduce STIs in young American women
• Background: Most American sex education is ineffective and
just repeats the basic facts
• Interviews and surveys: Young women already know about
STIs and how to prevent them -- but lack skills to
communicate with partners
• Communication: An interactive video that taught negotiation
skills (and not just basic facts) reduced STIs compared to
controls
• Take-home message: To implement behavior change,
people need to know more than just the basic facts
(Bruine de Bruin et al., HIV/AIDS Prevention in Children and Youth, 2007;
Downs et al., Social Science & Medicine, 2004) 13
Carbon Capture and
Sequestration (CCS)
• Goal: To inform public debate about CCS
• Background: CCS aims to reduce CO2 emissions, but public
resistance may hinder widespread deployment in the US
• Interviews and surveys: Most people have not heard of
CCS, but they become concerned about the risks when they
learn about the risks and benefits of CCS, and want to talk
about wind and solar instead
• Communication: Information about the risks, costs and
benefits of 10 low-carbon technologies increased
acceptance of some CCS
• Take home message: To make informed decisions, people
need to understand all options and their risks and benefits
(Fleishman, Bruine de Bruin & Morgan, Risk Analysis, 2010;
Palmgren et al., Environmental Science & Technology, 2004) 14
Examples of other applications
• Pandemic influenza
(Bruine de Bruin et al., Global Public Health, 2006;
Downs, Bruine de Bruin, & Fischhoff, Vaccine, 2008)
• Xenotransplantation
(Bruine de Bruin et al., Risk Analysis, 2006)
• Smart meters
(Krishnamurti et al., Energy Policy, 2012)
• Climate change
(Bostrom et al., Risk Analysis, 1994)
• Hurricane modification
(Klima, Bruine de Bruin, & Morgan, Risk Analysis, 2012)
Expertise needed for
developing effective communication
• Diverse domain experts are needed to
ensure accuracy and balance of content
• Social scientists are needed to conduct
interviews and surveys with members of
the intended audience, to increase
likelihood of
– Using wording that is understood by recipients
– Covering content the audience needs
Recommendations
• Conduct interview and survey research in
China to inform Chinese communications
• Develop and test communications in China
• Build database of effective communications
• Conduct meta-analyses to identify features
of effective communications
• Publish guidelines on effective
communications
Relevant references
Bruine de Bruin, W., & Bostrom, A. (2012). How to find out
what to address in science communications: An
introduction for scientific experts. Invited paper for the
Proceedings of the National Academy of Sciences
Fischhoff, B., Brewer, N.T., & Downs, J.S. (2011).
Communicating risks and benefits: An evidence-based
user’s guide. Washington DC: Food and Drug
Administration.
(http://www.fda.gov/oc/advisory/OCRCACACpg.html)
Morgan, M.G., Fischhoff, B., Bostrom, A., & Atman, C.
(2002). Risk communication: The mental models
approach. New York, NY: Cambridge Univesity Press.
