Teaching statistics to meet
the needs of policy makers
Marion Campbell
Health Services Research Unit
University of Aberdeen
HSRU is funded by the Chief Scientist Office of the Scottish Government Health and Social
Care Directorates. The author accepts full responsibility for this talk.
Background
•
•
•
•
Teaching statistics is a means to an end
The end aim …
• to interpret data accurately and
meaningfully to make inferences
about the greater population
Often the end user is a decision-maker – who
has to make decisions on behalf of
themselves/wider population
As such … in addition to teaching the skills of
doing statistics, we need to teach the skills of
communicating statistics
Health Services Research Unit
Who are the ‘decision makers’?
Policy makers
Public &
patients
Health service
managers
Clinicians & health
professionals
Health Services Research Unit
Policy makers
• Macro-level questions
• What types of services should
be provided in the NHS?
• Should the NHS fund particular
types of drugs/treatments?
• … what would be the impact of
those decisions on the health
& wealth of the nation?
• Societal perspective
Health Services Research Unit
Health service managers
Issues of …
• Local priorities and
needs
• Cost-effectiveness of
specific healthcare
treatments
• How should services be
delivered locally?
• How can the
organisation be most
efficient?
Health Services Research Unit
Health professionals
• Questions about clinical
effectiveness …
• Which treatments have
the best outcomes for
my patients?
• Eg Which procedure is
better for knee
replacement?
Health Services Research Unit
Patients
• Questions about the
patient journey &
treatment choices …
• How long should I expect
my recovery to take?
• What are the risks to
undertaking different
treatments?
Health Services Research Unit
Communicating with policy makers
•
•
•
•
•
•
Policy makers and scientists (incl
statisticians) are different
Policy makers have a different
agenda
They have many conflicting
sources of advice
They are non-specialists
They are busy
They have different timescales
Health Services Research Unit
Differences: academia and policy
/
Source: http://blogs.lse.ac.uk/impactofsocialsciences/2014/02/11/policy-world-versus-academia
Health Services Research Unit
The concept of evidence
● Differs between academics and
policymakers
● Academics:
− ‘evidence’ = academic research
findings, RCTs, novel methods
● Policy-makers: often use and value
other types of evidence:
− public opinion,
− political feasibility
− knowledge of local contexts
Health Services Research Unit
Engaging policy makers
Research evidence is most likely to be used by policy-makers when
it meets the following needs:
● Relevant - addresses questions of interest to policy-makers
● Accessible - can be easily found and understood by policymakers
● Immediate - evidence is provided in a timely manner for current
problems
● Useful - information provides solutions to problems
● Quality - information is credible and scientifically rigorous
● Collaborative - early and sustained engagement with policymakers will increase their understanding of the research and
their confidence in using it
● Targeted - identifies a specific audience and key messages
Health Services Research Unit
The John Lavis model
Full text
Paragraph
Headline:
• Short & catchy
• Retain essence of overall
message
Sentence
Headline
Sentence (have 2 versions):
1. Highlight the research
evidence
2. Highlight the implications for
policy
Paragraph:
• Why is the issue
important?
• What does the research tell
us about the issue?
• To what extent does
current decision-making
differ from optimal
decision-making?
• Who should act and what
should be done?
Headline
Full report:
• Expand on paragraph
points
• Full references
The 1:3:25 model
1
page
One pager:
• Main messages
• The “so what” not just
the “what”
3
page
Three pager:
• Condensed version of
findings
• More like newspaper
rather than academic
abstract
25 page
Headline
Headline
25 pager:
• Plain language
• Anecdotes to convey
messages
• Cover 7 categories
Seven sections for 25 pager
•
•
•
•
•
•
•
Context
Implications
Approach
Results
Additional resources
Further research
References & bibliography
Health Services Research Unit
So what else helps?
•
•
•
•
Foster relationship with key policy makers to
understand their needs – then target message
Delivery should be in interactive forum if possible
Focus on implications rather than information
Policymaker more likely to act on research
evidence if packaged as an actionable message
Health Services Research Unit
Words of wisdom
“Lay the fundamentals bare, make the logic
clear and get rid of the jargon”
UK Government Scientific Adviser, Sir Mark Walport
Health Services Research Unit
A rallying call!
•
Speak up!
−
•
Stand up!
−
•
Communicate your science. If you don’t,
it’s as good as non-existent. Making
science just for yourself does little good.
Make yourself heard. Meet people.
Introduce them to yourself and the
work you do. Be proud of achievements.
Modesty in moderation!
Gang up!
−
Consensus is important. Even if it exists
at moderate levels, build on it. It makes
the scientist’ voice stronger in the eyes
of the outside community.
Anne Glover
Chief Scientific Adviser to the EU
Health Services Research Unit
Useful references
•
www.researchtopolicy.ca
•
Canadian Institute for Health Information. You say ‘to-may-to(e)’ and I say ‘to-mahto(e)’” Bridging the Communications Gap Between Researchers and Policy-Makers. CIHI,
2004
•
Canadian Health Services Research Foundation. “Reader-Friendly Writing—1:3:25,”
Communication Notes (2001) ww.chsrf.ca/knowledge_transfer/pdf/cn-1325_e.pdf
•
Food and Agriculture Organisation of the UN. Communicating with policymakers Presenting information to policymakers. www.fao.org
Health Services Research Unit
Contact details
Marion Campbell, HSRU, University of Aberdeen
[email protected]
@MarionKCampbell
Health Services Research Unit
Download

Teaching statistics to meet the needs of policy makers