Statistics for Experimental Biologists


Topic index

Key books

External links


Interested in learning more about the prevalence of questionable research practices and the psychological factors that affect research integrity? Chapter 1 discusses these in detail.

Scientific misconduct: just don't do it

The aim of this website is to "up the game" of biomedical scientists by providing useful information on the conduct and analysis of experiments. Some aspects of experimental design and statistical analysis could be poor due to a lack of knowledge, but they could also be rubbish because of the intention to deceive on part of researchers. While obvious, in the interest of completeness, it should be stated that scientific misconduct and fraud damages the quality of the literature and therefore reduces the scientific community's ability to makes sense of the world. A good rule regarding misconduct or fraud is: just don't do it.

Here are two blogs that make for depressing reading: Science Fraud, which posts dubious papers, and Retraction Watch, which tracks papers that have been retracted. A few things stand out after reading these: (1) The number of dodgy papers is far too common. (2) It seems that people use the same tricks more than once (getting away with fabrication, image manipulation or duplication, etc. makes it likely that they will do it again). (3) Journals often seem reluctant to set the record straight, and when they do, the reasons for a retraction are often vague (e.g. "a problem with the figures"). (4) The chances of getting caught are now higher than in the past. This is due to the ease of finding, retrieving, and comparing papers from a particular scientist, and putting them online for all to see. In addition, software to detect plagiarism and duplication is available.

If you are temped to advance your career by dubious means, remember, just don't do it!