Ethics

While most academic training includes discussion and description of academic codes of honor - including outright fraud (cheating) and plagiarism - we have begun to include instruction in research ethics as part of our formal undergraduate curriculum. Research ethics are critical to the advancement of science as they provide a framework within which scientists can share their ideas and knowledge in a fair and professional manner, and include topics such as proper (and timely) reporting of data, authorship, and professional behavior. Traditionally, students learn research ethics through informal mentoring with their research advisors. However, a more formal approach is encouraged by funding agencies, professional societies, and common sense. At IU, we are responding by incorporating discussions of ethics into our undergraduate classes, as well as organizing special ethics seminars for research students.

Our ethics program is built around a "case study" approach. Students are given scenarios involving real life situations that they are likely to encounter as undergraduates or beginning graduate students, and are asked to discuss possible resolutions of the ethical questions involved. Discussion topics include reporting data, data rights, credit for ideas, and professional behavior. The answers are not clear-cut, and students must grapple with shades of gray to help them define what the limits of ethical behavior are. Students also read and discuss the booklet "On Being a Scientist: Responsible Conduct in Research," published by the National Academy of Sciences.

Scenarios for graduate students involve ethical concerns more appropriate for their career stage, including conflicts of interest, authorship, and collaboration.

Scenarios

Students discuss the following situations in small groups and then present the pros and cons of recommended actions. They use as a guiding principle what is best for science and the advancement of human knowledge.

Experimental Technique

In lab courses, students are trained to report and include all data. However, in real life, the situation is often more complicated. Answers aren't known in advance and often scientists are developing new techniques at the same time they are trying to measure some physical quantity. Consider the following situations:

• You are measuring the radial velocities of 50 stars in a star cluster using a cross-correlation technique. After measuring all 50 stars, you realize that the spectra contained telluric lines (from the earth's atmosphere) as well as spectral lines from the stars, which changed some of the results. You re-measure the velocities, avoiding the spectral regions with telluric lines. What values should you report - both sets of measurements, an average of the two sets of measurements, or just the second set?
• In a search for new excited states of the proton, you have made an important discovery. In three independent samples of the data obtained during one experiment, the results for the mass of the excited state are 2.015, 2.020, 2.017 GeV. A fourth analysis gives a value of 2.028 GeV. You want to report the existence of this new excited state at an upcoming conference. What value should you report?
• You are measuring the flux from stars in an optical image of an open cluster. In your first pass, you use a small aperture (2 pixel) to measure the counts associated with each star in the image. In a second pass, you use a large aperture (100 pixel) to measure the counts for each star. In a third pass, you examine the curve of growth and use a modest aperture (20 pixels) that corresponds to where the curve of growth levels off. What values should you report - all sets of measurements, an average of the three sets of measurements, or the results from just one of the sets?

Misconduct

When James was an REU student, he and his mentor, Dr. Smith, worked on a research project that is now being written up for publication. Dr. Smith has read James' first draft of the paper and has suggested that it is ready for publication after a few minor changes are made. In the meantime, James has moved on and is now a first year graduate student applying for a prestigious NSF Graduate Fellowship. James would like to list the paper as "submitted" on his application and he knows the paper will be submitted before his application will be reviewed by the Fellowship Committee.

• Is it ethical to list the paper as "submitted?"
• What should James do?

Conflicts of Interest

Research proposals (for funding and telescope time) undergo "peer review" where a group of scientists evaluate the scientific merit and feasibility of the research project. Obvious conflicts of interest, such as former students/advisors and current collaborations, usually mean that you cannot participate in the review. Less obvious conflicts of interest are sometimes more difficult to evaluate.

Consider the situation where you are serving on a panel evaluating proposals for telescope time on a major international facility. Your panel consists of seven other astronomers who have some expertise in stellar astrophysics, but you are the sole panelist currently working in the field of stellar spectroscopy. How would you handle the following situations?

• A proposal you are reviewing requests observing time to obtain radial velocities of stars in a particularly interesting star cluster. You know that a colleague obtained radial velocity data for that cluster 3 years ago, but hasn't published it.
• A proposal you are reviewing requests observing time to obtain observations very similar to ones you yourself have already received time to observe on another telescope.
• A competitor working on the same research problem you are has submitted a proposal for time to obtain a critical observation needed before they can publish their findings. Your own work has been delayed from publication, but you expect to submit the paper soon.

Colin has been allocated telescope time to obtain observations of a small sample of AGN. When he arrives at the telescope, he hears that the previous observer was also observing AGN. He notices that the observations from the previous run are still located in the top directory of the computer. Should he look at these files to make certain he doesn't duplicate data already taken? If he has not yet finalized his source list, should he add sources from the previous observer's list in order to complete his sample?

Social Interactions

Kyle is attending his first American Astronomical Society Meeting. After a long day of standing in front of his poster, he is ready to celebrate his success at making up brilliant answers to random questions from the passing crowds. He pulls out his invitation to the party of the meeting (everyone will be there), and starts to make plans. Should Kyle retain his professional demeanor during the party? If yes, how can it be a real party? If no, what is the distinction between personal and professional behavior? Does your answer change if Kyle is a distinguished member of the Society rather than a student? Does your answer change if Kyle is female (Carla)?

Authorship Practices

Professor Mendoza and her graduate students have been working for several years on a class of new materials expect to provide good candidates for high temperature superconductivity. Preliminary results on five materials have been presented as posters and abstracts at meetings, but it is now time to write up the results for publication in a refereed journal. They are considering whether to publish their results as a series of papers, or as a single, larger paper. Two students will be completing their dissertations and looking for jobs soon and will benefit from longer publication lists and from first-author papers, but the work may have greater impact if presented as a single, long paper. On the other hand, a single, long paper might be too long and complex. In addition, Professor Mendoza will be coming up for tenure soon, and would like to have a significant paper on her resume to document the importance of her work for the Tenure Committee.

• How should the group handle the publication of their results?
• Are they justified in publishing a series of papers?
• If they publish a single paper, how should order of authorship be determined?

Giving Credit

Debra is an advanced graduate student working on the origin of fluctuations in the cosmic microwave background. As a member of the WMAP team, she has access to an extensive set of satellite measurements. The results of her work indicate a significant anisotropy that she attributes to baryon number violation in the early universe.

Debra attends an annual APS meeting and presents an abstract and a poster on her work. At the conference, she meets Dr. Nelson, an eminent researcher in her field, and is flattered that Dr. Nelson takes an interest in her work, which she describes to him in detail. But six months later, Debra is surprised when she sees an article by Dr. Nelson in Physical Review Letters that suggests that anisotropies in the CMB may be created by baryon number violation in the early universe. Debra is surprised and disappointed, however, to find that Dr. Nelson has not cited her abstract.

• Should Dr Nelson have cited Debra's abstract?
• What recourse does Debra have if she feels her work has not been acknowledged?
• Should Debra have been less open about her work?

Cynthia is applying to graduate school. As part of the application, she is requested to write about her previous research experience(s) and her research interests. After completing a first draft, she asks a friend to read and comment. She incorporates many of the suggested wording and grammatical changes in a revised draft, which she then gives to her research advisor for additional comments. She incorporates the extensive changes recommended by her advisor and submits the statement with her application to graduate schools and fellowships. Has Cynthia behaved ethically? Would your answer change if the recommended changes she made were "minor"? Would your answer change if she had added new research ideas suggested by her friend and/or advisor?