Why Most Published Research Findings Are False - (Sep/20/2005 )

This is the what John P. A. Ioannidis a researcher at the University of Ioannina School of Medicine and Tufts Med School in New England tries to answer in his article of the same title? He goes on to state that, "It can be proven that most claimed research findings are false." Bold move, especially while the Evolution v. Intelligent Design debate rages on. This is a peer-reviewed publication...

What do you think?

Lack of replication = Majority of false findings?

It's probably true if you sharply limit the definition of "research findings" to mean "studies that try to find relationships among groups of things with use of statistical meta-analysis" or some such.

In broading his language to paint all "reasearch findings" with the same brush, he seems to me guilty of the same over-reaching he's chiding others for.

I'd find it extremely hard to believe that most molecular biology research finding are false, for example. Some perhaps, but "most"?

If someone shows an enzyme has a particular activity, using adequate controls, does the enzyme not really have that activity? It may not even be that enzyme's primary role in vivo, but if it is demonstrated to be capable of converting substance X to substance Y given conditions A, does it not really have that capability?

Or if a particular bacterium is shown able to degrade a particular compound, or cause a particular disease, can it not really degrade that compound or cause that disease?

In general, I find that I'm able to replicate the results in most published papers.

Otherwise, I'd be screwed. We'd all be screwed...perpetually stuck in mol. bio purgatory.

-Matt

come on can you really take this guy seriously! seems to me that this paper might be published for sensationalism rather than furthering scientific discovery.

hello,
lack of proper care in using regagents for research will cause false positive results n publications. which can not be reporducible by other researchers.

IF POSSIBLE LET ME KNOW YOUR OPINION:
[i]
is it possible that a researcher get fed up in a research project which he is working on for several years with lot of efforts and wants to get rid of it since he cant reach any end,

at this time will he try to produce false results?????????? OR close the project?


thanks
payeli

Is it possible? Of course -- any type of deliberate misdeed or unintentional mistake is possible and has probably, at one time or another, occurred. The scientific method and the peer review system are pretty good at keeping papers based on such erroneous data out, or in the worst case detecting them after the fact.

But, events such as these are outliers -- clearly not the norm, and most published research in fact reports real, factual data, despite what Dr. Ioannidis says in his journal article.

Interesting. I have heard atleast 2 PIs (professor) claim the same thing.

Those who manufactured data will repeat in their career and will certainly be caught eventually.

I don`t think u got the idea behind this

its not about ''cheating'' its about the pressure that is put upon the bioscientist to produce POSITIVE results for the sponsors.... I don`t think anyone can write an essay and claim the above in a straightforward manner but believe me he implies in a delicate way what is a common truth

same thing happens in bioethics articles where opinions are dictated by the status quo

anyhoooooooo who cares ....

when i was masters student most of my freinds cheated on their results. PIs didnt care (pretended they didnt know) they graduated the students and published the papers. just a small example.