Statistics and Ethics: a powerful mix

1
Statistics and Ethics a powerful mix

• Paul Marchant
• Leeds Metropolitan University
• p.marchant_at_leedsmet.ac.uk
• Leeds Bradford RSS meeting 21 May 2008

2

• Statistics is a big subject cropping up in all
  areas of scientific enquiry.
• Ethics is a big subject cropping up in all areas
  of human activity.

3
Statistics is at the heart of science

• Science tells us what the world is like and
  statistics tells us about doing science.
• (The logical path from data collection to
  conclusion)
• Statistics is therefore a very important
  subject.. the reason Im keen on it and more
  people should be.

4
All Science is Social

• All Science is Social to some extent (the
  social structure of the science business
  itself.power, money, employer-employee
  relationships etc etc.)
• (We seldom seem to talk about the context in
  which we do our work of statistics..)
• We are human with all the human foibles and
  frailty that entails.and as humans, inhabiting
  the world, we are affected by what others give to
  us.

5
RSS code of conduct

• Speaks of
• public interest employers and clients
• the profession and standards professional
  competence.
• (Commended to all Fellows and Mandatory on all
  Professionally Qualified Fellows)
• Other codes exist

6
Other Codes

• American Statistical Association (1999)
• Ethical Guidelines for Statistical Practice
• www.amstat.org/profession/index.cfm?fuseactioneth
  icalstatistics
• International Statistical Institute (1985)
• http//isi.cbs.nl/ethics.htm
• These give more info and discussion on the
  issues involvedvery worthwhile reading.

7
Problems can arise

• Responsibilities may conflicte.g. employer and
  public interest.Someone in control might not be
  persuaded to give the time to do the job properly
  ( i.e. to make sure any conclusion is
  sufficiently good).
• The great power of statistics almost invites
  problems.as knowledge (even if its only pretend
  knowledge) is power.
• The banner of science . the badge of truth /
  the seal of approval .a very valuable
  commodity! (Think of advertising)

8
Serious scientific misconduct happens

• Note a previous RSS meeting on 29 Jan 2008 in
  Leeds (Medical Section) on Biomedical
  Controversies, included issues of serious
  scientific misconduct. (materials available at
  http//www.maths.leeds.ac.uk/statistics/rss/ )
• There is a large area between the sunlit summit
  of the pure scientific ideal and the murky depths
  of deliberate scientific misconduct..a twilight
  zone where the purity of science meets grubby
  humanity.e.g. people just not knowing what they
  are doing!

9

• There seems to be a great emphasis on doing
  research.helpful for career advancement..pressur
  e to cut corners (even/especially when the
  presence of a corner is not recognised!)
• A lot of statistics done by those without
  sufficient statistical educationdont have a
  mentor. (It seems that sometimes statistical
  advice is assiduously avoided).

10
Research Ethics Committees

• I was on an NHS REC for four years
• Interesting
• Rather like being on jury service
• The REC a diverse and conscientious group of
  people. ( a useful platform for statistical
  education of other members)
• There can be a tendency to want to accept
  'harmless' nonsense, i.e. if a study is not
  likely to do much harm, then it is acceptable
  I think we should be more rigorous.
• Needs to be an arms length between the judge,
  juror (or even barrister for the prosecution)
  and barrister for the defence.

11

• A REC is surely not to promote research rather
  to make sure that any research that is done is
  ethical. (An R and D department should promote
  the research and provide the necessary.).
• Statisticians need a general scientific awareness
  (as well as ethical sense) as statistics is not
  just maths. There are other things beyond
  sampling variation..
• It is important not to be intimidatedhold firm
  to scientific logic.

12

• The costs of poor research, a lot of chaff
  and potential harm. Potential loss of public
  confidence.bad for society, bad for statistics
  and bad for statisticians. It is important that
  sound statistical thinking is publicly recognised
  and valued.(and the opposite!)

13

• Need mates with whom to check, to clarify, to
  chew over.
• I would have liked to have someone to talk to
  when I started on the REC. I would also like to
  have some guidance on the dodges that those
  wanting to slip things through get up to.
• Need source(s) of support, particularly for
  isolated workers.How to facilitate this?

14

• I wonder
• How frequent it is that problems arisei.e.
  people face conflict.
• How frequently is RSS brought in to attempt
  resolution.
• What problems are occurring in subjects outside
  of the mainstream of penetration by statistics

15
Richard Feynmans thoughts

• It's a kind of scientific integrity, a principle
  of scientific thought that corresponds to a kind
  of utter honesty--a kind of leaning over
  backwards .. In summary, the idea is to give all
  of the information to help others to judge the
  value of your contribution not just the
  information that leads to judgement in one
  particular direction or another.
• from Cargo-Cult Science (Caltech address 1974)
  reported in the collection Surely Youre Joking
  Mr Feynman