I watched a recording of this meeting, following the link posted by Dave, and was amazed to learn that 486 species of macro moth had been recorded in the NMRS in January! (At 7:52 mins into the recording).
I appreciate that this is UK-wide and possibly includes all life stages, but it still seems remarkable that more than half the UK's macro moths have been recorded in January. A similar story for micros, with 646 species recorded in January. Or am I completely misunderstanding what is being said?
Phil T
Hi Phil. I noticed these graphs for macros and micros when I attended the presentation live. The figures were so far from what I intuitively expected to see that I almost asked a question, but I decided that although an answer might tell me what exactly was being measured, it wouldn't change the fact that the charts were not something to which I could relate.
ReplyDeleteIf this had been a scientific paper, one would have expected a caption describing exactly what was being measured - but this was a general interest presentation, not a scientific paper, so perhaps an exact definition is not required in such contect. However there is clearly a disconnect between what many people might implicitly assume a chart of "the number of species per month" would look like, and what these charts were actually showing.
While writing this comment, I have discovered that Blogger has a limit on the length of comments, so I've had to split it. A second comment will give examples for this context of possible reasons why the species counts in these charts might be very different from what you and I might have implicitly assumed we would see.
For this comment, I will just boil it down to the following general statement:
I know from quite a lot of experience with data analysis that one first has to clearly define the question to which one is seeking an answer. Then it's important to understand the nature of the data that is being analysed. Next it's important to carefully structure the query/analysis so that it matches the question, given the nature of the data. Finally, it's important when discussing the results to re-state what the question was to which this is the answer.
It's possible to create a huge gap between what some figures actually represent and how people receive them, even if all of the underlying data is "correct". In this particular case, there is no real consequence of the gap as no-one would intend to take any action as a result of these charts and they are not intended to be a scientific paper. But it does mean that - whatever the detail is of what they represent - it's unlikely that they help general understanding.
Thanks for those thoughts Tim. I suspect that it is probably your fourth factor which contributes the largest element to the difference between the data presented and what one might casually expect. Even so it is still a big stretch from what I would intuitively expect, to the data presented.
ReplyDeleteBut it is also possible, I suppose, that somehow the wrong question was asked of the database in the first place!