OK, so who left the silly tablets on the underside shelf?

I have been studying KellyWatch and now I am confused about what’s lifelike.🙂 pic.twitter.com/3kenqZeI5M

— Actual Indy Loun 🏴󠁧󠁢󠁳󠁣󠁴󠁿 (@IndyLoun) April 25, 2023

Is that immediately? That is completely determined stuff from him, it is an apples-and-rotavators comparability.

— James Kelly (@JamesKelly) April 25, 2023

Many thanks (after a vogue) to ‘Actual Indy Loun’ on Twitter for alerting me to one thing that I am too blocked to in any other case concentrate on, ie. that my obsessive fan in Somerset has been performing some extra obsessing. (By the way, I’ve one or two suspicions about who ‘Actual Indy Loun’ could also be in actual life. Beneath all of the mischief-making, there is a sharp intelligence and breadth of political, cultural and historic data that units him aside from in any other case related accounts. I could also be improper, however….hmmmm.)

The apples-and-rotavators comparability Stuart has provide you with on this event goes like this.  Final Thursday, I discussed in my blogpost in regards to the new YouGov ballot that 34% of respondents suppose Humza Yousaf might be a greater First Minister than Alex Salmond, and 28% suppose he might be worse than Mr Salmond.  I identified that this six-point hole barely reaches the extent of statistical significance. And some days earlier, I identified that the yearly polling averages for Sure assist since 2014 have assorted wildly between 45% and 53%, and never “flatlined on 47%”, as Stuart and his fan membership have wrongly claimed many times and once more.

So is there a hilariously embarrassing contradiction, as Stuart appears to consider, between me saying {that a} six level hole in a person ballot barely reaches the extent of statistical significance, and me saying that an eight-point hole between the yearly averages of Sure assist is a wild variation?

Er, no.

The purpose a few six-point hole in a person ballot barely being statistically vital is simply an apparent assertion of truth.  The usual margin of error in a ballot of 1000 individuals is round three share factors, so if the 34% who suppose Yousaf might be higher than Salmond was overestimated by three factors, and if the 28% who suppose Yousaf might be worse than Salmond was underestimated by three factors, you’d attain an actual tie of 31-31.  However yearly averages of Sure assist are, by definition, not based mostly on a person ballot of 1000 – they’re based mostly on dozens of polls which between them comprise actually tens of 1000’s of respondents.  So, y’know, the three-point margin of error does not apply to the yearly averages.  Fairly easy level, I would have thought, however apparently it must be defined to some.  The one manner the error in a yearly common would strategy three factors can be if there was a systemic downside in polling methodology.

There is a neat little irony right here, as a result of all of this is kind of similar to the rationale Stuart first blocked me on Twitter round six or seven years in the past.  He saved claiming that Hillary Clinton’s modest lead over Donald Trump within the polling averages meant that it was all inside the margin of error and both candidate might be forward within the well-liked vote.  I and a number of other others (together with some who’re usually slavishly loyal to him) identified that the margin of error solely applies to particular person polls and that if Clinton was sustaining that small lead constantly throughout a variety of polls, it meant she really was forward until there was a systemic error in polling methodology.  (As you will recall, Clinton went on to win the favored vote, though she misplaced within the electoral school – however there was certainly a systemic polling error and her lead was overstated.)  Stuart, as is his manner, refused to pay attention, referred to as us all idiots, and resorted to the block button.  Plainly two-thirds of a decade later, the penny nonetheless hasn’t dropped for him – however I am afraid he is simply plain improper, and that is that.

If there every other bleedin’ apparent factors I can clear up for you, Stu, you’ll want to give me a shout, will not you.

*  *  *

Scot Goes Pop depends on donations to proceed:

Direct funds may be made through Paypal.  My Paypal electronic mail tackle is:  jkellysta@yahoo.co.uk