Long term monitoring data is hard to come by. It's expensive, time-consuming, difficult and usually requires an expert. When someone comes up with a new whizz bang method that promises to change your world, you should upgrade, improve your techniques, do cutting edge science right? Wrong. If new methods offer substantially lower cost and/or increased accuracy, it is all to tempting to throw the baby out with the bathwater when keeping up with the scientific Joneses.
I work with other people's data. Lots of it. The limiting factors for good inference in population monitoring is the number and timespan of viable datasets. After funding, the single biggest factor limiting the timespan of data is changed methods.
So, if I can offer you one suggestion it is this:
Don't change.
If you must, or it seems wise to change, then index the old methods to the new, otherwise you are rendering all that time effort and money spent previously useless*. (not totally, but it is much less valuable).
No comments:
Post a Comment