ASPB - Newsletter - The Bioethics Imperative V

Untitled Document

The Bioethics Imperative V - March/April 2002

Column from March/April 2002 issue of ASPB News.

"Mokita": The truth we all know and agree not to talk about. Papua New Guinea

Scenario: A postdoc presents a photomontage of wild-type and mutant roots as part of her manuscript to your lab group. The scale bars on the six-panel figure are identical, and the roots are all pointing in the same direction and are at the same angle. In complimenting her on the figure, several lab members ask her to show them how she made the montage. She replies that she used Photoshop to touch up the images and Canvas to make the montage. In the original single images she then shows, the roots are all of different sizes and point in different directions and angles and the backgrounds differ in color. Some of the root hairs have been cut off in the manipulations. Has the postdoc mishandled her data?

Data mishandling is an easy thing to do. In addition to the example above, I have witnessed the following, all sins of omission or commission:

Graphing sets of tissues in groupings so as to imply a relationship between the tissues, e.g., the same numerical values in y-axis
Where the slope is constant, filling in values where data were not actually recorded
Using a control from another experiment rather than repeating a control every time the experiment is run
Omitting a set of data without just cause or ignoring one analysis because it conflicts with another
Starting all of your repeats of an experiment on the same day, at the same time, with the same reagents
Making a standard error from two means
Changing the aspect ratio or orientation of a photograph
Omitting details of data purposely to confuse someone else who might want to repeat the work so that you don't get scooped.

There is an aspect of data handling that is mokita: The very act of trying to make sense of data can create a bias in the researcher or a tendency to look for certain outcomes. Sometimes you simply forget to take a step back from your data and be sure that you are looking at it from all possible angles before you submit it for review. Sometimes the pressure to publish or get a grant drives you to make a sin of omission. Although inadvertent, these too, are cases of data mishandling.

Issues of data mishandling become more difficult when it is not you, but a colleague, peer, or student for whom you are responsible who is at fault. These issues quickly begin to plumb a bioethical morass.

I have posted a few case studies concerning data handling at http://www.aspb.org/newsletter/bioethics.cfm. In addition, web resources on this and related issues have been compiled at http://adams.dom.washington.edu/bri/bri.htm.

Next issue: Scientific misconduct.

Dina Mandoli
University of Washington, Seattle
mandoli@u.washington.edu

Supporting materials can be found HERE.