Trying to estimate the population of birds in North America between now and 50 years ago seems like it would have an error rate of at least 29%. Are the counting and estimation mechanism even anywhere close to similar?
You don't believe the number, so they could possibly be wrong, so it's not a cause for concern? That's a lot of jumps to make to get to the case where we don't have to worry.
What techniques are they using to count and estimate the birds, and what objections do you take to their methods? Sadly I can't find out as the site is behind a paywall for me.
There's only one jump: I don't trust the data, so I'm not yet concerned by this click-bait title.
In this day of clickbait titles, studies that are not reproducible, and suspect motivation for studies, the idea that we can calculate the total number of birds with an error rate of less than 30%, including the fact that some data is 50 years old, doesn't pass my smell test.
I would prefer waiting for accurate data rather than panic over yet another headline that is not anchored in actual science and facts. For example, did they try to count the number of birds at the same time of the year both times? Do some birds have different multi-year migratory patterns leading to them being in different parts of the world? Did they try counting the birds twice using the same techniques but two different teams, and then compare the numbers?
There are too many ways this calculation could fail for me to be trust this data point.
> The analysis, published in the journal Science, is the most exhaustive and ambitious attempt yet to learn what is happening to avian populations.
> A team of researchers from universities, government agencies and nonprofit organizations collaborated on the new study, which combined old and new methods for counting birds.
> Kevin Gaston, a conservation biologist at the University of Exeter
> Hillary Young, a conservation biologist at the University of California, Santa Barbara
> Scott Loss, a conservation biologist at Oklahoma State University
> Kenneth V. Rosenberg, a conservation scientist at Cornell University and the American Bird Conservancy
> Dr. Young, of the University of California, Santa Barbara
> “If we have two data sets showing the same thing, it’s a home run,” said Nicole Michel, a senior quantitative ecologist at the Audubon Society who was not involved in the study.
> Europe is experiencing a similar loss of birds, also among common species, said Dr. Gaston, of the University of Exeter. “The numbers are broadly comparable,” he said.
So all these people are idiots and you know better?
Not once does the article state the expected error rate with the calculation. Even polls give an error rate when they are announced. Saying 29% decrease without having an idea of its accuracy is meaningless. This is simple science. We learned about giving error estimates in college in every single experiment, and now it's okay to throw those out because it concurs with your world view?
Sorry but I prefer to look at things scientifically. You can continue to believe things that fit in your world view, but I prefer science over religion.
This appeal to authority probably won't be very effective in convincing thr root commenter about the validity of the study. None of the quotes addres the concerns about counting methods, and the experts cited are more likely to be seen as experts in biology than statistics, inference, and logic.
If the experts do this sort of work often they'll be proficient in the application of statistics. We don't distrust engineers because they aren't the people who came up with the physics they are applying.
That's a valid assumption to start with, sure. If the results are called into questions on the basis of statistics, their field of work is no longer a convincing argument.
When someone calls into question your design choice, you don't say "trust me an engineer" unless you're joking.
