(No, the title of this blogpost isn't about the BBC and Gary Lineker, but it easily could be.)
I received something very unusual about an hour ago - it was a private message via the Blogger platform, a facility that I wasn't even aware existed. It was warning me that the Redfield & Wilton poll that I covered last night was not as methodologically robust as it should be, and as evidence linked me to an article in The National entitled "Redfield & Wilton poll needs tightening up after errors, expert says". Now, without wanting to sound too cynical, I genuinely deduced two things from the title of the article before even clicking the link - a) The National's "expert" would be Mark McGeoghegan, because it always is, almost without exception, and b) he would be hostile to the poll because it shows Kate Forbes is the public's preference for First Minister. I don't literally know for a fact that Mr McGeoghegan is a Humza Yousaf supporter, because I long ago "removed him from my social media experience" (as the jargon goes) to get some peace from his incessant and infantile trolling. However, everything I've seen of him in the past suggests that the Yousaf campaign is his only possible home, because his main political motivation is hardcore identity politics zealotry.
My guesses were all but confirmed by the text of the article. Although he doesn't explicitly identify himself as a Yousaf supporter, he sounds very emotionally engaged with the fact that Redfield & Wilton misspelled Yousaf's name "in every tweet and graphic, every time he's mentioned in their blogpost, and throughout the tables". The obsession with identity politics isn't far away either, because he blasts Redfield & Wilton for using the word "transgenderism" which he says is "pejorative" and "non-neutral".
Hmmm. Let me tell you a little story. Eighteen months ago, I crowdfunded a poll that was partly about independence-related issues and partly about GRA reform. Although I hadn't raised enough funds to commission Panelbase to conduct the poll, I eventually turned to them in desperation, and plugged the shortfall in the funds with my own money. Do you know why I went to those lengths? Because other pollsters I had previously been in touch with had tried to make the GRA-related questions I submitted more "neutral" and "balanced", and one of the ways they did that was by inserting the word "cis" - which 95% of the population do not even know the meaning of, and most of the remaining 5% find deeply offensive. I was practically in tears trying to explain that the people who had crowdfunded the poll were trusting me to come up with a fairer and more balanced poll than had previously been managed on the subject, and that if I allowed their money to be used to ask questions using ideologically-loaded words like "cis", they would be quite simply flabbergasted. It would genuinely have been a betrayal and they would never have forgiven me.
Now, I don't doubt the sincerity of pollsters who truly believe that inserting words like "cis" somehow makes poll questions more "neutral" - but I would just note that the fact they do hold that belief demonstrates powerfully that the concept of neutrality is very much in the eye of the beholder, particularly on the trans issue.
As far as I can see, none of Mr McGeoghegan's quibbles actually cast doubt on the reliability of Redfield & Wilton's results. I drew attention myself last night to the sloppiness of misspelling the names of both Humza Yousaf and Alister Jack, but it's hard to see how that would make any difference to the way respondents answered the questions. The absence of a Yes / No crossbreak in the data tables is an unfortunate oversight, but again, that wouldn't actually affect the results of the poll. The "transgenderism" question was peripheral, and had no impact on the results about leadership preference or voting intentions - although, let's be honest, even if it was possible to come up with a universally-accepted "neutral" word to summarise what the Scottish Government have been doing on the trans issue, it's highly unlikely that respondents would have reacted any more positively than they did.