Saturday, June 13, 2009

Eeeek, a quandary

Hmm. What do you do when you realize (based on your own work and some deduction related to part of the molecule described) that something someone showed in a talk and published in a paper is extremely likely to be completely artifactual?

It's a minor issue, and unlikely to be field-defining or anything, so it's not like some major dogma needs to be unseated here. But the poor scientists who are working on this will likely continue to study this artifact, and it makes me sad because they are very nice people who are genuinely excited about this thing they think it novel and useful.

I mean, maybe it's not an artifact--we'd have to do the experiment to find out. But if it is...

This sucks.

8 comments:

Odyssey said...

Tell them. They'll be disappointed, but not as much as when they inevitably find out for themselves.

But be sure of what it is you're telling them...

Comrade PhysioProf said...

Tell them. If they look into it and realize that you are correct, then they need to correct the publication record. Depending on the details of the nature of the "artifact" this could be either by retraction, or by publishing a clarification as a letter to the editor or "correspondence" or "matter arising". if they fail to do this, and you yourself have confirmed it's an artifact, then you ought to send such a correspondence to the journal.

Candid Engineer said...

Tell them nicely for the good of science. It will undoubtedly save them pain down the road.

Arlenna said...

Yeah, I did talk to them about it at the meeting--the challenge is that they had not considered the specific things I noticed and thought they had addressed them, but had not. So, it might take showing them some evidence and smashing their happiness about their project. :(

ScientistMother said...

that sucks, but it also is part of science...

Ambivalent Academic said...

Tell them! I was on the receiving end of this recently, before I was too far into investigation of the artifact. I was really grateful that this person told me so I didn't waste anymore time.

Anonymous said...

This is a very, very important consideration. I did a late-career PhD on an esoteric topic on the crossover of biology and chemistry, and realised halfway through that the commonly used chemical techniques were most likely producing artifacts that didn't mean anything biologically. I changed the methods and got different results - which were not so 'nice', but seemed to make more sense. This seemed to be not a 'publishable' result, but I presented at a local conference where there were a lot of grad students, and a couple of years later everyone was asking students in similar fields 'are you sure you are not producing artifacts?' So, one can make a difference with informal conversations. I repeat, it is a VERY important consideration. It is up to those who work in non-traditional crossover areas to spot the glitches. More focussed people just don't think about them. (They are not better or worse scientists, it is just that we are all different).

Anonymous said...

Sorry, re the previous post, I don't like to be completely anonymous I am sometimes known as "d.".