This episode of @c4dispatches should be used to warn journalists what happens when you don& #39;t take steps to prevent confirmation bias or out-group homogeneity bias in your reporting (or to understand the data you are analysing and concepts like & #39;average& #39;) #AfterDispatches https://twitter.com/C4Dispatches/status/1250833731738382337">https://twitter.com/C4Dispatc...
The @IRR_News formal complaint to Ofcom highlights some of the statistical issues (btw, I& #39;m assuming the data was sourced from http://data.police.uk"> http://data.police.uk & #39;s street level crime reports) https://twitter.com/IRR_News/status/1252357419084050434">https://twitter.com/IRR_News/...
More on the data aspect by @profcolinclark here, too: https://twitter.com/profcolinclark/status/1251846435898867712">https://twitter.com/profcolin...
The first part is a bizarre methodology: "We made a list of Traveller sites where there had been reports of crime, and looked to see if there was crime there. We found that there was indeed crime in the places that we had picked based on their crime reports"
You can follow @paulbradshaw.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: