Research, by nature, is repetitive. The need for demonstrating real effects demands the repeatability and reproducibility of results. This can come at a cost of sanity when initial promising results fail to recur in subsequent samples.
This has been my experience of late when I attempted ‘one last experiment’ before writing a brilliant paper based on an important and surprising finding. That was a couple of months ago. Since then, all my efforts in repeating the experiment have failed to yield good data.
In fact, the only consistent and reproducible trend seems to be the increasing depth of the dint in the wall from where I have been banging my head whilst trying to figure out what’s happening.
So now I’ve taken to desperate measures. I have gone back to the literature to try to explain the phenomenon. Has that helped? Well, no. Not really. It has thrown up many more variables that may influence the results but not offered any solutions. This probably comes back to why I started investigating this topic in the first place.
Then I went on to further desperate measures. I went for a coffee break and then came back to review the data through fresh (-ly caffeinated) eyes. And suddenly, like one of those Magic Eye images, the data resolved itself into a hazy outline of a solution.
I have been studying these samples and this phenomenon for weeks and have accumulated a great deal of data on the topic. Now I know exactly what doesn’t work and, in knowing that, a solution to why it doesn’t work is becoming more obvious.
Maybe soon, with more coffee and squinting, a real solution will present itself and I will get my ground-breaking paper.
Failing that, I should at least be able to produce a strong correlation between the standard errors of averaged replicates and the depth of dints found in walls at head height.