Map of life expectancy at birth from Global Education Project.

Friday, July 09, 2021

On the bias of science: part two

A favorite game of right wing cranks -- and we've seen them try it here -- is to read abstracts of government-funded research projects, distort their meaning, and make fun of them. This goes back to Senator William Proxmire and his "golden fleece" awards, and third party presidential candidate Ross Perot in the 1980s. Apparently the idea of studying fruit flies or gas emissions from cows strikes people as silly. Of course the investigators aren't given a chance to explain why it isn't silly after all. (FYI fruit flies are a common lab model used in genetics experiments. Gas emissions from cows make a substantial contribution to atmospheric CO2 and climate change. That isn't actually funny.)

Anyway, a knowledgeable and fair-minded person would not be expected to agree with every federal grant decision. How could they? The federal grant making process is competitive and there's usually little to choose between the proposals that get funded and the ones that come close. People may have different priorities, favor different methods, have differing tolerance for risk and reward, all sorts of reasons why they might make different choices. But the important thing for people to know is that for the most part, decisions about what research to fund are not made by government employees.


Congress establishes some broad priorities in the legislation authorizing the federal science agencies. Your elected representatives decide how much funding each of the National Institutes of  Health receives. NIH is very popular so Republicans vote to fund it, in fact they rejected The Former Guy's attempts to cut the NIH budget. NIH, the National Science Foundation, and Agency for Healthcare Research and Quality all work in a broadly similar way, although the details vary. Agency staff, who are scientists qualified in the areas they oversee and are not political appointees, prepare funding announcements, which are mostly fairly broad descriptions of areas of interest. Some of these stay open for years, others for a briefer time, and there are some one-time announcements which are usually more focused. However, investigators can submit an application for any research they want through what are called "parent" announcements. The more focused announcements are not limiting, just encouraging.


The staff then assign proposal to peer review committees, which consist of scientists with relevant expertise, almost entirely university based. They do not work for the government and they do not receive instructions from the government about how to score the proposal, except for broad criteria, such as innovation, significance, approach and qualifications of the applicants. Generally three people are assigned to read each proposal in full, and score it according to these criteria. The higher scoring proposals are discussed with the entire group, and get scores from everyone. Then of course the proposals that ultimately score highest may be funded.


Note that "government" does not make these decisions. No bureaucrat and no politician gets to decide or even to hint at which proposals ought to be funded. No-one is allowed to discuss any of the proposals with any of the review committee members. That said, are there biases involved in this whole process, or the conduct, evaluation and publication of the ensuing research? There may well be, which I will discuss next. But it is not in furtherance of any sort of nefarious government agenda, not is it likely to result in funding for research which is actually frivolous, even  if you don't understand it. We can have differences of opinion, but those are the facts.



2 comments:

mojrim said...

One tiny quibble here, estemado Cervantes. To wit: domestic cattle and their feces lagoons emit methane, which is 25x as potent a greenhouse gas. When you tally up the tonnage and factor in the delta they are responsible for around 1/3 of america's planet killing emissions. The really dismal thing is that this happens in fixed locales (esp dairy cows which are worse than beef cattle) that lend themselves to recapture but it's become something of a third rail to talk about. The EPA page on GHGs neatly obscures this reality by simply counting tonnage emitted and ignoring relative GHG value.

We could capture all that NH4 for a lot less money than going after CO2 but it might amount to a small increase in the price of beef production. That would be workable except that this is america:
1. Beef is america's sacrament.
2. The cost would fall on the ranchers, 80% of whom are small, family operations (<1000 head).

Remember all that madness with the Bundys in Nevada and Oregon? While They may individually be assholes, both standoffs gained traction because the beef industry has evolved into a near total monopsony, with McDonald's and Walmart topping the list. Because this is america it's the smallest producers that get squeezed and absolutely no one but small ranchers cares.

mojrim said...

Speak of the devil...
https://www.motherjones.com/politics/2021/07/heres-how-bidens-new-executive-order-could-shake-up-big-ag/

I'm starting to think this government may accomplish something after all.