Date: 5/7/19 7:55 am From: <clearwater...> Subject: [obol] Re: Ebird--avoid using the "x" in checklists because...
Tom & All,
Ah, yes, the old ad hominem method of debate, followed swiftly by invocation of personal opinion ("I don't give jack ..." in place of rational argument, and finally a straw-man argument (insinuating that I was somehow calling ebird users "closet Trumpistas").
But let's break this down.
First, anyone who signs up for an ebird account already gets the regular e-newsletters from Cornell, so they've seen the exhortations about the claimed importance of "complete checklists." I know this may come as a shock to some of you, but I get them too. Hence I know that those exhortations have already reached the most relevant target audience, namely ebird subscribers.
So what was your purpose in amplifying this one by re-posting it on OBOL? Apparently as a recent "convert" (your word) you felt a need to get up on a stump and preach to the wicked who have been ignoring the message. That's OK, lots of us love to get up on the stump and preach now and then. But when you're up there, kindly don't be offended if you encounter some skeptical reactions.
I suggest that the concept of "marginal utility" as used by economists would be useful for considering proposals like this.
Certainly, every bit of data could potentially have value, including even rough estimates of siskin or starling flocks. However, there are costs to acquiring and recording each bit of data. It's reasonable to ask if the marginal benefit outweighs the marginal costs.
Since most ebird contributions are done by volunteers, the main marginal cost is field time. I'd guess that many birders who use "X" on their checklists are posting their checklists from memory after they get home or get back in the car, because they had better things to do while the birds were right in front of them.
They might have used that time to notice other birds that they would have missed while punching numbers into their smart phones. Most people only have two eyeballs and they can only truly focus those eyeballs on one thing at a time.
They might have used that time to observe details of bird behavior and use of habitat, which could add to our general understanding of these species and their habitat needs. For example, they might have taken time to watch the movements of a particular Grasshopper Sparrow to see what size territory it was using, and then relate that the condition of the habitat.
They might have used that time to observe how land management practices were affecting the habitat, and considered how to approach the owners or managers to help promote bird conservation on a direct and local scale.
They might have used that time to talk to curious passersby and point out some of the birds, with the hope of sparking an interest that might eventually get those folks to engage in bird conservation.
Those are just some of the marginal costs. The marginal benefits of counting or estimating every starling flock have yet to be explained in a way that's satisfying to me as someone who *does* work with statistics as a regular part of my day job.
To give just one example, here are eBird starling reports from this January through March, at a popular local gull-watching spot (Coffin Butte Regional Landfill):
Is our understanding of the numbers of starlings using this site diminished in any way by those three X's? No, it isn't. In a statistical analysis you could just exclude those checklists (or incorporate them in a fancier way) and you'd be left with two orders of magnitude in variation of the estimates.
In reality, all of those estimates are well below the actual number of starlings that visited the landfill on a daily basis. They just reflect how many those observers happened to notice while they were mainly focused on something else (namely, trying to pick out rare gulls). You would almost certainly get a higher starling count if you spent all day there waiting for some commotion that caused the whole flock to take flight. Or you could stake out their evening roost spots around the neighborhood (I know of a few). If you did that you'd get a count of many tens of thousands, perhaps even a hundred thousand or more.
The folks who try to estimate bird populations based on ebird's typically short-duration observer samples have various tricks in their bags for coming up with more realistic estimates that might come close to what you'd get from a roost survey. If those methods are robust, they shouldn't be affected by a few X's.
Ultimately you can end up with some numbers. But then what do you do with them in terms of native bird conservation, beyond saying, "Wow, that's a lot of starlings!"?
In cases such as this, estimating starlings on every last checklist has negligible marginal utility for bird conservation. You can do it if you want to, but there's no reason for people to think that their effort on such details is serving any appreciable conservation purpose.