Interesting Links for 29-11-2018

[andrewducker]

Brexit: More voters would prefer to stay in EU than accept Theresa May's deal

(tags: uk europe polls )

Wriggly, giggle, puffball: What makes some words funny?

(tags: language humour )

Teachers of Reddit discuss positive trends they have noticed in today's youth

(tags: children GoodNews viaElfy )

New research finds there is no “right thing” to say when you want to be supportive

(tags: empathy support psychology )

Dropping Acid - how playing with technology caused a whole new musical genre

(tags: music history Technology )

Brexit: Referendum may be inevitable - John McDonnell

(tags: UK europe referendum )

The First Labor Strike in History (in ancient Egypt)

(tags: history strike egypt )

Comments

[doug]

That would be hilarious if true, but it's a transcription mistake - see e.g. https://twitter.com/polisisti/status/1067926368673980416?s=21

The corrected chart is here: http://www.deltapoll.co.uk/steve-fisher-condorcet
and - if you read carefully, because the colour choices are terrible, horrible, no-good, very bad - it shows that the May Deal is preferred to Remain, No Deal is preferred to Remain, and the May Deal is preferred to No Deal. So the Condorcet winner is the May Deal.

[doug]

But the polling in TFA linked above says Remain is preferred to the May Deal (by some margin), and the May Deal preferred to No Deal, but doesn't give figures for a match up betwen Remain vs No Deal.

I bet you could find a fair number of people who would have non-transitive preferences here, though, because people are like that.

[doug]

There's also this Survation data, fieldwork 27 Nov, which gives:

Govt Brexit Deal: 35%
No deal: 41%
Don't know: 24%

Govt Brexit Deal: 37%
Remain: 46%
Don't know: 17%

No Deal: 40%
Remain: 50%
Don't know: 11%

That's clearly Remain as the Condorcet winner, and the Govt Brexit Deal/May's deal as the Condorcet loser. But note also that even if there's apparently intransitive preferences in the two-way ties, the number of 'Don't knows' might explain it without requiring any individual respondent to have intransitive preferences.

It evidently depends who you survey and how you survey them.

[calimac]

So it appears that two of the three comparisons were reversed in outcome in the first, incorrect, chart, then?

Someone should inform the blogger whom I cited, but it's not going to be me, because I got lost in a welter of tweets when I tried to follow this. I thought Usenet threads were bad, but Twitter is an even worse medium to read a conversational exchange in.

[doug]

Yeah, the incorrect chart got the numbers round the wrong way. (The 'corrected' one is still terrible: the colours change meanings in different pies, and it's three pie charts, and pie charts are almost never the right answer.) It was probably an honest mistake, and I think the original blogger has since corrected it (the corrected chart link I posted). But that was screencapped and tweeted and retweeted and cut-and-pasted all over the place and there is no way on ghods' green Urth I would want to chase down them all and correct them, even if people on Twitter universally welcomed being corrected on factual errors.