There is no better feeling in science than to see weeks of planning a new experiment come together in one day’s work to produce data. To be more correct, one datum.
We’ve been working for weeks building a prototype, and developing the experimental conditions in which we could test the prototype. The test conditions were as close to real life as we could get in a system scaled down in size to keep costs reasonable and manageable electronics in the prototype.
We had 5 scientists involved in overseeing the electronics, probes, monitors and the model system being examined. As a first ever experiment none of us knew what to do, what our roles really were, what to expect or even when things would be “done.” I think everyone felt we knew what a success would look like and that there were a thousand ways it would fail, but as with most very new things the failures are expected but provide information for the next try. For example on an earlier dry run to test the equipment we found out that the prototype seemed to work better with the $50.00 components as opposed to the $1000.00 parts. Give me a call if you need some slightly used $1000.00 parts.
Much to our astonishment we saw a huge, consistent and importantly explainable response with the prototype. Explainable is a key aspect to the success in that we need to summarize and report the results to make them useful. Random unexplainable results are not useful, our results need to be useful and not random, which they are.
What we saw is a simple numeric change. Picture a bunch of eyes looking at a slot machine and the wheel comes up with three $$$; a jackpot. We saw a number come up on the computer screen that represented a jackpot. Computer programs examined and confirmed the results and the model performed admirably.
What we saw is hard to explain without getting too technical and is also being patented, so it is confidential. But one little datum represented weeks of work, many hours of detailed planning and much excitement. Well, I say excitement but that is relative for us stuffy science geek types.
The postscript to one datum is that one experiment is not significant. We need to reproduce the results multiple times, determine prototype tolerances, specs and optimize performance. That will take months of work and eventually seeing the same response as today’s will become routine. But it was a good day and something to be savored.
That experiment has been repeated with nearly identical results. So the Datum is now data.
January 17, 2010 @ 9:21 pm