Read on Twitter
Very quick thread on using evidence to back up arguments, how it can go badly wrong, and how to spot it - prompted by the latest Morning newsletter from NYT. 1/
Let’s take apart the sourcing for the claim, today, that “Social distancing leads to the isolation and disruption that have fed so many problems over the past two years” such as “vehicle crashes” 2/ https://www.nytimes.com/2022/03/09/briefing/covid-precautions-red-blue-states.html
This a big claim! The piece links to a source - this is where we might expect to see the reasoning or evidence for it. But wait! The source is: a piece *by the same author*… 3/ https://www.nytimes.com/2022/02/15/briefing/vehicle-crashes-deaths-pandemic.html
Linking to yourself as a source is bad practice. But let’s see if the linked piece has some hard evidence in it. 4/
It starts by noting a trend: “Crashes—and deaths—began surging in the summer of 2020.” And again proposes an explanation: “an increase in aggressive driving more than made up for the decline in driving.”
Where’s the evidence? 5/
Where’s the evidence? 5/
This earlier piece links to NYT reporting, so let’s see what evidence *that* article uses to establish the connection between pedestrian deaths and a surge in reckless driving that its headline suggests 6/ https://www.nytimes.com/2022/02/14/us/pedestrian-deaths-pandemic.html
This article should be able to tell us how a surge in reckless driving was measured. It should also establish on what basis someone feeling “reckless” can be measurably linked to more crashes. 7/
But it doesn’t cite a study or any kind of data set. It quotes a Professor of Psychology, Human Dimensions of Organizations and Marketing. This is his opinion 8/
That is how one person’s opinion can form the basis of three pieces in a national newspaper. This does not seem to me ideal operating procedure. 9/
