As many scientists know, experiments sometimes (often) don’t work out as expected. You just have to deal with it, because that’s how science works. You record what you observe and come to conclusions based on that. Oh, and if you find a method that doesn’t work, you look for a new one
I learned many things in that experiment back in July. (Yeah, I started this blog post a long time ago, during a frustrating lab experiment.) One very important thing: Tyrophagus putrescentiae eggs disintigrate in 70% ethanol. Important lesson! Next, we tried freezing, since it was extremely difficult to sort through the food in all ten vials in one afternoon (every other day). Fortunately, freezing worked! Refrigeration probably would have also worked.
Anyway, science does not always go as planned, and you have to adjust to that.
Now I’m running a different experiment. It’s always interesting to find a different organism in your arena, but sometimes it’s hard to avoid. Those moths are sneaky! Fortunately, some of the little snafus aren’t likely to mess up your experimental results. Also, that’s why you use replication.
Okay, those are my musings for now. Any interesting experiment stories?
Tags: experiments, mold mites, science
Here’s one I heard recently: Last year, an experiment was developed to measure ATP activity in some tissue that had been prepared in a certain, very complex way. The basic hypothesis was simple: The pattern of ATP activity would be the same if the tissue was prepared this way as if it was not prepared that way, minus a constant. If this was true, then the preparation would be OK for a further experiment to work. In other words, this experiment was one of those nuts in the nuts and bolts of a much larger very detailed study.
So, the experiment was optimize last year, and then, recently, an attempt was made to replicate it and get a few more data points by doing a few more runs.
But it didn’t work. And it didn’t work again. And again.
Everything was checked. Including the freshness of the reagents. The reagents were, indeed, brand new bottles fresh form the manufacturer.
It turns out that one of the reagents being used was different in its preparation by the manufacturer now than it had been last year. Same name, same product ID, different stuff in the bottle.
Always check your reagents.
Quite right, jaf! The only kinds of experiments that aren’t productive are the poorly documented or outright fraudulent ones. Everything else is helpful, even if it’s only checking all of the dead alleys so that others (or yourself later) can research a productive path.
“Most scientists spend most of their time trying to discover why their experiments won’t work.”