They used statistics!!

Posted on April 11, 2012


My previous post dealt with what i suspect was overzealous reporting by a reporter from the Daily Mail. Luckily, making fun of journalists never gets old. Apologies for linking to a newspaper article in a Dutch newspaper, but on the upside – it is only a short post…

Today the Dutch “Algemeen Dagblad (AD)” presented an article (link) with the title “15 percent of medical doctors knows about scientific fraud” (freely translated). Following last years stories that two Dutch Professors (Diederik Stapel and Don Poldermans) were guilty of scientific misconduct and subsequently fired, Medisch Contact a newssite for Dutch medical doctors, conducted a survey among 800 of their members to ask about their experiences related to scientific misconduct. 15% of those interviewed claimed they have ever been witness of scientific data being ‘made up’. Although Medisch Contact notes that of those interviewed only 18% had a PhD and only very few were (still) active in research (so we don’t know to what time period this number refers), this is serious business and deserves our combined efforts to root these practices out completely.

However, what I wanted to get at is the following. The AD journalist subsequently added (again,freely translated): Additionally, … nearly 25% of medical doctors have noticed that a scientist selected data to his benefit or used statistics on the data.” Wow, things must be pretty bad if scientists used statistics! I bet things can’t get much worse than that! What bothers me here is that apparently, in the mind of science journalists the use of statistics is equivalent to fraud. And since they write for the general public, this image will get stuck in everyone’s mind: scientists use statistics to manipulate the results of their studies to get what they want. Statistics = fraud.

This is a fairly small survey. but the rationale above has been exploited beautifully (for lack of a better word) by the anti-climate change lobby. Especially the word manipulated works especially well in lay articles. For those interested, here are some links to get the idea: 1,2,3,4,5.  Although interestingly, this got a lot less attention (Scientists Are Cleared of Misuse of Data).

Right, we’re getting a bit off track here. So was this really what the Medisch Contact said? Scientists committed fraud because they used statistics? Actually, what their communication (link) says is the following:”22% (of medical doctors in the survey) say they have ever been witness to selection of data or use of statistics to obtain significant results.”

Ah right! Now that makes a lot more sense. That is indeed scientific misconduct. Seems like a certain AD journalist missed a critical part of that sentence….

Posted in: General