Keep science honest 1

Tim Birkhead says there should be courses on "doing science" that address scientific dishonesty (Working Knowledge, July 22).

At St Andrews University, we already have one: a second-year undergraduate module in "science methods" that teaches the "how" of science (philosophy, methods, ethics) rather than the "what".

But there is a deeper malaise. We can and should teach students not to falsify, but we should also teach them that there are no particular "correct" results that they should obtain in their practicals and projects.

Such a commitment to honesty should apply higher up the science hierarchy, too. Yes, we need to have zero tolerance of "falsification" and "fudging", but we also need to accept that a "negative result" is just as good as a "positive result" and allow publication accordingly.

Also, if the science establishment's model of a good scientist is of one who gets results rather than one who tells the truth, what incentive is there for honesty? In the draft discussion of the research assessment exercise methods for biology, there is not one single mention of research integrity apart from a vague mention of "rigour".

World-class science without world-class honesty is not worth the paper it is so voluminously written on.

Charles Paxton and Morven Shearer, St Andrews University