Fabrication and Falsification in Social Research

SPEAKER 1: Imagine that you are a vegetarian.A 2011 study found that meat eaters aremore selfish than vegetarians.You assume the study probably involved dataabout the behavior of people with different diets.But it didn't.The author, Diedrik Stapel made it up.Dutch universities responded by examining Stapels' work.

SPEAKER 1 [continued]: They found he fabricated data for 55 peer reviewedjournal articles and 10 PhD thesis that he supervised.Fabrication is making up data that never occurredand reporting it, while falsification is manipulatingresearch results or data so the research is notaccurately represented.So just how many researchers do this?

SPEAKER 1 [continued]: Analyst are divided on the extent of research misconduct.So it's difficult to create an overall picture.Some serious cases have come to light in the social sciences.For example, there are questions surrounding several economistsand psychologists who manipulate data and hypotheses to fiteach other.

SPEAKER 1 [continued]: 20 percent of attendees at the 1998 American EconomicAssociation meetings were surveyed.From a 23% response rate, they estimatedthat 4% of respondents had falsified dataat least once, though not necessarily for publication.Research into the causes of these kinds of researchmisconduct is scarce.

SPEAKER 1 [continued]: Some see the researcher as a bad apple.A researcher's own ambition, vanity,desire for recognition and fame, and the prospectof personal gain may lead to behaviorthat crosses the limits.Some point to the culture that may prevailin certain disciplines or research groups,known as the bad barrel.While others identify the creation of a research

SPEAKER 1 [continued]: environment overwhelmed by corrupting pressurescalled the bad barrel maker.Many academics are under increasing pressureto publish, preferably in English,as their nation or institution seeksto establish its placement in international researchrankings.So individuals are forced to meet publication and citation

SPEAKER 1 [continued]: targets in order to obtain jobs, grants, researchcontracts, or sponsorship.In many countries there are also weak and ineffective oversightregimes inside institutions or within the research sector thatcan be thought of as a bad barrel checker.Whichever explanation you prefer,we may find some academics will continue to release findings

SPEAKER 1 [continued]: whether or not they actually have data to support them.