3 ms·
> So alarmists are extrapolating from two data points (years 1989 and 2013) for a 220-hectare plot of German farmland to the entire world. Actually the study i
by titzer 9y ago
> So alarmists are extrapolating from two data points (years 1989 and 2013) for a 220-hectare plot of German farmland to the entire world.
Actually the study is couched in a lot of careful caveats and they _don't_ extrapolate to the rest of the world. I didn't see any extrapolation. And, if you read the actual article that you linked above, a decline of 2/3 from 1970 was measured in Scotland. That's corroborating evidence that ought to be carefully considered.
> Actually they're not even measuring that, they're measuring the weight of dead bugs' soaked in 70-80% alcohol.
Are you f'in serious? You don't even trust them to be able to properly factor out the weight of alcohol? That's not being critical, that's just disbelief.
> I could go on and on about controls in statistical experiments but I think you get the idea.
I think it's easy to throw out such statements on HN but I think you get the idea.
- giardini 9y agotitzer sez: "Actually the study is couched in a lot of careful caveats and they _don't_ extrapolate to the rest of the world. I didn't see any extrapolation." Others are using the article as support for far-fetched claims. AFAICT a 220-acre study in the hinterlands of Germany with two data points (two samples, one in 1989 and one in 2013) is causing worldwide headlines claiming a collapse of bug population: https://www.google.com/search?&q=75%25+decline+over+27+years https://www.google.com/search?&q=75%25+decline+over+27+years titzer sez: " You don't even trust them to be able to properly factor out the weight of alcohol?" You missed my point: My point was that they measure the weight of a mass of unclassified insects + alcohol + whatever else was in the trapping fluid and that they weren't presenting statistics based on insect count or insect species or insect age, health, etc. Beetles weight more than mosquitoes for example but we aren't being presented statistics for specific insect species/age/health/etc, we're being presented statistics for a mass of unclassified insects + absorbed alcohol + whatever else. Any gross measure of insect biomass is highly variable depending on bug species, population, location, predation, food supply, season, disease, etc. If you do your harvesting immediately after , say, the cicadas emerge, you'll get lots of insect biomass. Do it a six weeks later and the biomass will plummet precipitously. So you must design your experiments to take such possibilities into account. titzer sez: "That's not being critical, that's just disbelief." And that is a poor attempt at use of a broad brush to tar someone. I _am_ skeptical of the studies, as any scientist should be. You may, like others, read confirmation from your car's front bumper, but I prefer experiments with proper controls and measurements when we're speaking of good science. And, of course, I am always wary of field experiments that may involve the use of ethyl (drinking) alcohol: Hans: Sorry, Hauser, I yust panicked and ran. Does it hurt much? Hauser, Hauser, are you drunk??? Hauser: My hands are on fire. Zat's the last time I try to throw a hornets nest! Good thing we have a handy pain-killer available! Zis bug juice is hardly working though; zese stings are killing me! Pass me zat other bottle. Hans: Scheiss, Hauser, you're drinking ze bugs from ze data collection!! Vat the hell we gonna use for data? Hauser: Zis trip is a catastrophe! Scheiss! I know what! Ve can give the data to Schneider. He said he can prove anything with almost any data. Once he told me, Schneider told me, "ze less data, ze better!" He says no one pays attention to real statistics anymore.