The importance of proper control experiments

Get your science fix here: research, quackery, activism and all the rest
Post Reply
FlammableFlower
Dorkwood
Posts: 1508
Joined: Mon Nov 11, 2019 1:22 pm

The importance of proper control experiments

Post by FlammableFlower » Tue Dec 14, 2021 3:24 pm

https://www.chemistryworld.com/news/the ... 01.article

Publication claimed to be able to do palladium-catalysed reactions using an amine catalyst instead. Turns out palladium was there all along...

User avatar
jimbob
Light of Blast
Posts: 5276
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: The importance of proper control experiments

Post by jimbob » Tue Dec 14, 2021 4:18 pm

Palladium has form in the topic of this thread.
Have you considered stupidity as an explanation

FlammableFlower
Dorkwood
Posts: 1508
Joined: Mon Nov 11, 2019 1:22 pm

Re: The importance of proper control experiments

Post by FlammableFlower » Wed Dec 15, 2021 3:00 pm

It gets everywhere, and it turns out is a bl..dy amazing catalyst...

The Whitesides group a while ago rebutted a paper that claimed iron catalysed cross-couplings were actually down to the trace Pd in the iron (even stuff that was supposed to be 99.999% pure). Most commonly it ends up getting impregnated on teflon-coated magnetic-stirrer bars and then carried from reaction to reaction.

Allo V Psycho
Catbabel
Posts: 734
Joined: Sat Nov 16, 2019 8:18 am

Re: The importance of proper control experiments

Post by Allo V Psycho » Wed Dec 15, 2021 8:43 pm

Has happened to me twice. Second time was grafting tumour cell lines, grown on a boring living substrate. The experiments were very difficult and time consuming to do, so you couldn't run the control at the same time as the experiment. The first run gave a fantastic resulting biological effect, and since it was a cell line, we knew we could scale it up and do proper biochemistry. Celebrations, and champagne all round, and we starting drafting the paper for Nature. Doing the control with the boring living substrate was just a matter of conventional routine.

Guess what. The activity was all in the substrate, which wasn't scalable, so no biochemistry, and no fame and fortune. Got a paper out of it of course, which has had (checks) precisely 10 citations in 20+ years. But just glad we did the controls.

First time was even worse. I was analysing the effect of a possible mitogen on cells in culture - the hard way: I continuously video-recorded the cells, and counted the actual cell divisions: none of this artefact filled tritiated thymidine uptake for me (besides, we had done that already). Up all night for weeks on end, re-focussing the bl..dy microscope. The results were exactly as the literature predicted for a mitogen - the inter-mitotic time shortened. I didn't NEED to analyse them again blind, of course. Cell divisions are really obvious on video, and there is a timer visible on the videotape. No doubt about it all.

But my conscience smote me. I got a colleague to re-label the tapes and I re-analysed them without knowing which were experiments and which were the blank controls.

I still have no idea how it was possible to mis-interpret the original tapes.

(It wasn't a bust - on the anonymised-tapes-analysis the inter-mitotic time was the same, but the proportion of cells dividing had increased, so it was still a mitogen, but in a different way from expected).

Post Reply