Ethics’for’Astronomers’w.astro.berkeley.edu/~kalas/ethics/documents/Lecture-03.pdf · RM’...
Transcript of Ethics’for’Astronomers’w.astro.berkeley.edu/~kalas/ethics/documents/Lecture-03.pdf · RM’...
Ethics for Astronomers
Lecture 03
September 17, 2012
“Research Misconduct: RM=PF2”
READINGS: Bubble Fusion Scandal
US Federal Policy on Research Misconduct etc.
Does this kind of stuff really happen?
Misconduct: Rare in astronomy? (adapted from Broad & Wade 1994)
Hipparcos (~200 BC): Published star catalog taken from Babylonian sources as if it were the result of his own observa\ons (publica(on?) [Toomer, G.J. 1975, “Dic\onary of Scien\fic Biography” (Charles Scribner’s Sons, New York), p. 191]
C. Ptolemy (~200 AD): Claimed to have performed astronomical measurements that he did not. [Newton, R. R. 1977, “The Crime of Claudius Ptolemy” (JHU Press, Bal\more)]
G. Galilei (~1600 AD): Exaggerated the outcome of experimental results. [Koyre, A. 1968, “Metaphysics and Measurement: Essays in Scien\fic Revolu\on”, (Harvard U. Press, Cambridge)]
I. Newton (1687-‐1713): Introduced fudge factors into his magnum opus so as to increase its apparent power of predic\on (intent?). [Wesjall, R. S. 1973, “Newton and the fudge factor”, Science, 179, 751]
RM
Rhoads, L. J. 2002, “Beyond conflict of interest: The responsible conduct of research”, Science & Engineering Ethics, 8, 459.
• 38% are professors • ORI received 1752 allegations 310 investigations (18%) 124 misconduct (40% of investigated or 7% of initial) • Is that a lot? Numbers versus severity of one incident.
RM
Rhoads, L. J. 2002, “Beyond conflict of interest: The responsible conduct of research”, Science & Engineering Ethics, 8, 459.
• Falsification & fabrication alone or in combination = 86% of misconduct findings. • Plagiarism alone or in combination = 14% • Similar table from NSF gives plagiarism = 60% • Depends on authority of institution, PHS returns plagiarism case back to institutions for resolution. • Some plagiarism cases dominated by “authorship disputes”.
PF2
• Plagiarism: “the appropria\on of another person’s ideas, processes, results or words without giving appropriate credit, including those obtained through confiden\al review of others’ research proposals and manuscripts” [Federal Register, October 14, 1999] Words and ideas?
• Judged rela\ve to the norms of the par\cular discipline?
• Unnecessary to cite common knowledge [e.g. “gravity” (Newton)]
• How should plagiarism be resolved? (apology?, retrac\on?, sanc\ons?) • Collegial solu\ons among peers, formal pathways otherwise.
• Was it sloppy work, or inten\onal?
• Why is plagiarism unethical? What are the wider consequences?
PF2
• Fabrica/on: “making up data or results and recording or repor\ng them” [Federal Policy on Research Misconduct]
• Falsifica/on: “manipula\ng research materials, equipment, or processes, or changing or omimng data or results such that the research is not accurately represented in the research record” [Federal Policy on Research Misconduct]
• Judged rela\ve to the norms of the par\cular discipline? • How should F2 be resolved? (apology?, retrac\on?, sanc\ons?)
• Was it sloppy work, or inten\onal?
• What happens if it is NOT federal research?
• Why is F2 unethical? What are the wider consequences?
We are in a collabora\ve team, 3 subteams complete 3 preparatory observa\ons that need to be included into a grant proposal that is due tomorrow. The PI of the project explains that only 8 proposals will be approved out of 800 submioed, and the grant commioee will be looking for any possible reason to reject a proposal. The project has 6 students and 4 postdocs who are supported from the exis\ng project, but in approximately 10 months the funding will drop to 25% of its current level.
Team A presents the result below, but cannot explain the discrepant point. The PI notes that the methodology may appear flawed if they indicate an “unknown source of error”. Team B suggests that the point is removed. How would they jus\fy this ac\on, and how would you argue against it?
Team B presents the results from an observa\on performed iden\cally three \mes over. They suggest that the second trial is so low that it is wrong, and in its place the team could report the average of the first and third
trials. Is this falsifica\on of fabrica\on? Trial Data Suggested Value
#1 75.89 75.89
#2 23.33 74.33
#3 72.77 72.77
Team C gives the following data for a third experiment. The drop towards the laoer third of the experiment is consistent with a change in bias in the electronics of the detector. Since the bias correc\on is applied to all of the data, they suggest offsemng the curve upward by the median value of the previous points. Since bias correc\ons are not detailed in the grant proposal, this bias correc\on does not need to be men\oned in the proposal. Is this “cooking” the data?
1. Fairness: Everybody else is doing, fair for us to do it.
2. Benefit to science outweighs the cost, ends jus\fy the means.
3. Making up for errors. 4. We need the research support.
5. What else?
The PI instructs team A to send the plot with the discrepant point removed, team B to take the average value for trial #2, and team C to provide a plot with the “corrected” data.
Aser the mee\ng, student Arthur says to postdoc Lisa: “Dude, I can’t believe we’re just gonna make up some of this data.” Lisa says, “Oh, he has his reasons. Think of it this way:
What is the underlying assump\on on the first page that is driving the PI’s decision?
• ?
Should any of the students or postdocs leave the group, and with what consequences?
• A real life example: • Farewell Message from Alan Stern, Associate Administrator for NASA's
Science Mission Directorate – NASA Headquarters, 03 April 2008
– “As you probably know by now, I have announced my resigna\on as NASA's Associate Administrator for SMD and will be depar\ng at the end of this week… I assure you that my decision to resign came only aser several months of hard thought and reflec\on about the consequences of spiraling mission costs that SMD could not control. In the end, this became an important maoer of principle…”
– Analogy: Mike Griffin is the PI, Alan Stern is the lead “postdoc” driving the science forward.
– The PI risks losing his talented team when “principles” are trampled.
– How does the “postdoc” weigh the long term benefits versus overlooking the short term problems.
– Compare the alterna\ves for employment available to Alan Stern, a postdoc, and a student.
Bubble Fusion Scandal
• What are the essen\al facts of the case?
2002, Science, 295, 1868
Can it happen in astronomy?
Can it happen in astronomy?
Planet discovery not confirmed by Swiss Group...
Can it happen in astronomy?
Can it happen in astronomy?
What should be done?
Bubble Fusion Scandal
• List of APS guidelines violated (PF2 ?)
US Federal Policies on Research Misconduct
• ?
Should RM have a defini\on wider than PF2?
• Posn 1:
• Project: Create an astronomy data case study to illustrate fabrica\on and/or falsifica\on of research results (base it on astronomical data or theory), or plagiarism. If appropriate, a project may be disemminated on the course web site. Use your imagina\on or consider real-‐life projects.
• Present in class next week.
• Start reading “Conflict of Interest” material.