Ethical constraints Toby Walsh 4C, Cork, Ireland 4c.ucc.ie/~tw/ethics
-
Upload
christiana-harkin -
Category
Documents
-
view
231 -
download
2
Transcript of Ethical constraints Toby Walsh 4C, Cork, Ireland 4c.ucc.ie/~tw/ethics
Ethical constraints
Toby Walsh
4C, Cork, Ireland
4c.ucc.ie/~tw/ethics/
Outline
Why bother with ethics? Ethics in the research cycle
Funding Experiments Publication
When things go wrong Whistle-blowing
Resources
Web 4c.ucc.ie/~tw/ethics
Journals Science and Engineering Ethics
www.opragen.co.uk/SEE
Conferences Computer Professionals for Social Responsibility, Oct 5 2002,
Cambridge (MA)
Courses Check out your university Ethics must be taught in all UK Masters by Research programs
Why me?
I am not sure what qualifies me
I have faced a number of ethical dilemmas But then so will you
I am neither angel nor hopefully great sinner
A Day @ The Artificial Ethics (AE) Lab
Dr. MacIavelli arrives from Scotland to start a postdoc Email arrives asking him to
review a paper Checks progress of his
experiments Goes for lunch with interview
panel Drafts new conference paper Looks for a suitable
conference to submit to Clocks off early
Why bother?
To protect and benefit You Science Society
Why bother?
To protect your own reputation 3 most important qualities
of a scientist Reputation, reputation,
reputation
A blackened reputation is rarely restored Science is unforgiving
Why bother?
Treat others like you would like to be treated Get ahead by being the
“nicest” person in your area!
Science is very “social” Networking at conferences Job offers, PC membership,
…
Why bother?
To protect science’s reputation Science is largely self-
regulated We therefore enjoy
considerable freedoms But this requires us to
apply high ethical standards
Why bother?
To protect society’s interests Society invests in research
despite many other pressing needs
Society delegates many ethical issues to scientists
In return, society expects scientists to act in society’s best interests
Ethics in the research cycle
Funding Experiments
Data collection & presentation
Publication Authorship Plagiarism Citation Reviewing
Ethics & funding
Who do you take money off? Military or arms industry? Tobacco companies? Nuclear power industry? “Nasty” multi-nationals? Microsoft?
What do they require of you in return?
Ethics & experiments
Human & animal experiments Fortunately rare in CP Many ethical safeguards in
place Your university will surely
have an ethics committee to oversee such experiments
Nonsense, as long as it is done ethically, animal testing in an invaluable scientific tool
Ethics & experiments
Data collection & presentation Fraud Omission Manipulation Theft
First three almost always discovered Science requires results to
be reproducible
Panel is currently investigating possible fraud in claims by Hendrik Schon (Bell Labs) to have built an organic transistor
Ethics & publication
Where many of us face most of our ethical dilemmas! Not surprisingly, science is all about being the first to
publish an idea
Areas of concern Authorship Plagiarism Citation Reviewing
Ethics & authorship
Who do you make co-author of your paper? Colleague Supervisor Lab boss Your friends
They’ll do same in return!
Ethics & authorship
There exist a number of guidelines for co-authorship European Science Foundation“… In the case of joint authors, each should have made a
significant contribution to the creative or analytical process and each has to accept shared responsibility for the content of the resulting article or book. The concept of honorary or “ghost” authorships is inconsistent with good scientific practice…”
Ethics & authorship
Yes, but does this mean I put my PhD supervisor down as co-author or not?
No hard and fast rules It is often (but not always) the case that:
At the start of your PhD, you do. Your supervisor asked the questions, pushed you in the right
directions At some point into your PhD/postdoc, you don’t.
For some, this comes before the end of their PhD For others, this does not come till after their PhD work is in press
in journals
Ethics & authorship
How much does a colleague have to do to become a co-author?
Again, no hard and fast rules My pragmatic advice
Err on the side of caution Nothing more sure to end the working relationship
Ask their opinion In addition, ask for your name to be removed when
appropriate
Ethics & authorship
Some tests Have they read the paper? Do they understand the
paper? If you took sick in the
middle of the seminar, could they finish it?
Ethics & authorship
What order do you list authors?
By “merit” But this can be hard And what about “equal”
merit
Alphabetically Common in a number of
areas
Ethics & authorship
My recommendation: Invent an ordering scheme and stick to it
My scheme Used to be, alphabetical But recently [Colton, Bundy, Walsh 2000], … Now it is, “you argue over where your names go, mine is in last
place” My name is always last so you can tell nothing about how little
I did!
Plagiarism
The presentation of someone else’s work as your own
In exams, cheating In science, theft
Citation
Failure to cite In severe cases, plagiarism In less severe cases,
hinders careers In least severe cases,
hurts & offends
Citation is so easy & painless to do
1. D. Johnson : 10473
2. J. Ullman: 10087
3. A. Gupta: 7696
4. R. Milner: 7276
5. M. Garey: 6044
6. R. Rivest: 6038
7. J. Dongarra: 6024
8. R. Tarjan: 5875
9. L. Lamport: 5777
10. J. Smith: 5314
Ethics & publication
Your are responsible to protect your intellectual property For your funding body For your university/company For your own profit
Software patents are now possible But will annoy your colleagues, …
Ethics & reviewing
Blind reviewing Allows for criticism without
fear
Double blind reviewing “Safeguards” against bias
Open reviewing No one gets to hide!
Ethics & reviewing
You have unpublished work on the same problem You are obviously well qualified As a courtesy, mention possible conflict to
Editor/Program Chair
You already reviewed and rejected paper Look for changes Were previous reasons fatal? Do different standards apply to this
conference/journal/… ?
Ethics & reviewing
This journal submission already appeared at a conference Conferences don’t usually count as archival Does it extend previous appearance?
An almost identical paper already appeared Unless it was at a workshop, inform Editor/Chair If it appeared with a different author, treat very
seriously!
Ethics & reviewing
You have worked with the author in the past Recently
decline due to conflict of interest A long time ago
if people knew both your identities, would they raise their eye-brows?
You work in the same institution as the author Almost always decline
Ethics & reviewing
Papers are sent for review in strictest confidence You cannot share them
with colleagues You cannot admit to
knowing their contents You cannot work on
extending their results
Science is a race to publish, all the credit goes to the first to publish
UK Nolan committee
Standards for public officials (e.g. grant reviewers) Selflessness
Your decisions should be made solely in terms of public interest Integrity
You should avoid financial or other obligations that will conflict Objectivity
Your decisions should be entirely on merit Accountability
You can be asked to justify your decisions Openness
You should be as open as is possible, except where wider public intestest conflicts
Honesty Leadership
Whistle blowing
When things go wrong, what can you do?
Stop & be very sure of your ground People’s careers are in
balance here
Informally approach Editor/Chair Phone is often better than
email!
The perils of email
What you wroteI have concerns about Walsh’s submission. It appears that it borrows techniques from my recent AAAI paper that I sent him last month. I am sure it was an oversight on his behalf not to reference me …
How it was read
There are serious problems with Walsh’s submission. It steals techniques from my recent AAAI paper. Walsh is obviously a thief as I sent him my AAAI paper last month …
Whistle blowing
University level Formal procedures, committees, …
National level NSF Office of Inspector General NIH Office of Research Integrity (~10 proven cases/year) Danish Committee on Scientific Dishonesty Norwegian National Committee on Scientific Dishonesty German DFG Ombudsman (actually three people) …
Enough of me
Let’s open this up! What ethical dilemmas
have you faced? What situations concern
you?
Outline
Why bother with ethics? Ethics in the research cycle
Funding Experiments Publication
When things go wrong Whistle-blowing
Conclusions
To misquote T.J. Watson THINK ethically
Science depends on good ethics They are few black and white decisions Most are shades of grey Don’t be afraid to seek advice
Supervisor Mentor …