research fraud: how to do it and how to stop it

1
SHOULD RESEARCH FRAUD BE A CRIME? Research fraud: how to do it and how to stop it Luca Turin research scientist Theoretical Physics, Ulm University, 89069 Ulm, Germany The cliché states that extraordinary claims require extraordinary proof. A less well known corollary is that ordinary claims often require little or no proof at all. 1 It is possible to produce substantial amounts of research that is entirely made up provided that one keeps it plausible and boring so nobody is tempted to replicate it; one adds the regulation amount of “noise” to the data; and one varies the noise occasionally to avoid getting caught by identical statistics or duplicate noise traces. The main difficulty is to make sure nobody else in the laboratory notices, and this can be done either by stealth—on the part of foot soldiers faking the data—or coercion—on the part of the boss. Both methods are unreliable. Nevertheless, fraud mostly succeeds even when found out because there is no winning strategy for a whistleblower other than leaving the laboratory and keeping his or her mouth shut. Whistleblowers are loathed even when they are right. In countries like Holland, Denmark, and the United States that have central research integrity authorities, report the fraud to them and let them do their work. Otherwise, get the hell out and shut up. As for penalties, people found to be frauds should be made personally liable for misappropriated funds. This would focus people’s minds wonderfully, and refund the cost of expensive investigations. Competing interests: None declared. 1 Bhutta ZA, Crane J. Should research fraud be a crime? BMJ 2014;349:g4532. (15 July.) Cite this as: BMJ 2014;349:g5221 © BMJ Publishing Group Ltd 2014 [email protected] For personal use only: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe BMJ 2014;349:g5221 doi: 10.1136/bmj.g5221 (Published 21 August 2014) Page 1 of 1 Letters LETTERS

Upload: l

Post on 18-Feb-2017

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Research fraud: how to do it and how to stop it

SHOULD RESEARCH FRAUD BE A CRIME?

Research fraud: how to do it and how to stop itLuca Turin research scientist

Theoretical Physics, Ulm University, 89069 Ulm, Germany

The cliché states that extraordinary claims require extraordinaryproof. A less well known corollary is that ordinary claims oftenrequire little or no proof at all.1 It is possible to producesubstantial amounts of research that is entirely made up providedthat one keeps it plausible and boring so nobody is tempted toreplicate it; one adds the regulation amount of “noise” to thedata; and one varies the noise occasionally to avoid gettingcaught by identical statistics or duplicate noise traces.Themain difficulty is to make sure nobody else in the laboratorynotices, and this can be done either by stealth—on the part offoot soldiers faking the data—or coercion—on the part of theboss. Both methods are unreliable.Nevertheless, fraud mostly succeeds even when found outbecause there is no winning strategy for a whistleblower otherthan leaving the laboratory and keeping his or her mouth shut.

Whistleblowers are loathed even when they are right. Incountries like Holland, Denmark, and the United States thathave central research integrity authorities, report the fraud tothem and let them do their work. Otherwise, get the hell out andshut up.As for penalties, people found to be frauds should be madepersonally liable for misappropriated funds. This would focuspeople’s minds wonderfully, and refund the cost of expensiveinvestigations.

Competing interests: None declared.

1 Bhutta ZA, Crane J. Should research fraud be a crime? BMJ 2014;349:g4532. (15 July.)

Cite this as: BMJ 2014;349:g5221© BMJ Publishing Group Ltd 2014

[email protected]

For personal use only: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2014;349:g5221 doi: 10.1136/bmj.g5221 (Published 21 August 2014) Page 1 of 1

Letters

LETTERS