04-18-2018 10:16:00 AM -0700
04-16-2018 01:32:51 PM -0700
04-16-2018 09:59:36 AM -0700
04-12-2018 09:53:41 AM -0700
04-10-2018 11:19:03 AM -0700
It looks like you've previously blocked notifications. If you'd like to receive them, please update your browser permissions.
Desktop Notifications are  | 
Get instant alerts on your desktop.
Turn on desktop notifications?
Remind me later.

Stupid Science Tricks


Of course, regular readers of this column seeing politics and ugly discussions in the context of science immediately think of the climate change debate; sure enough, that is the topic of today's stupid science trick.

Undoubtedly, you've heard quoted the result from Cook and others that "97 percent of scientists agree that global warming is caused by humans". (If not, have a look at this Google search.) This has been a subject of a succession of stupid science tricks.

The Cook et al study basically worked like this: a collection of papers were selected by computer, and rated by readers on a 1 to 7 scale from "Explicit endorsement with quantification" to "Explicit rejection with quantification."

Richard Tol examined the paper statistically and found:

The claim that 97% of the scientific literature endorses anthropogenic climate change (Cook et 15 al., 2013, Environmental Research Letters) does not stand. Numbers are padded with irrelevant papers. A trend in composition is mistaken for a trend in endorsement. Reported results are inconsistent. The sample is not representative. Data quality is low. Key results cannot be reproduced or tested as data disclosure is incomplete.

Stupid Science Trick #1: Don't Understand the Results

This one is a favorite of journalists. If you read the headlines in, say, the Washington Post

97 percent of scientific studies agree on manmade global warming, so what now?

they read as if 97 percent of all studies agree. Unfortunately, that's not even what Cook et al say: the real result was that 97 percent of studies that they classified as taking a position on climate change agreed -- but 66 percent of the papers they examined took no position at all.