Tonight, I went onto my other computer. On that computer I keep some very large volumes of down-loads, such as Feynman’s ‘Lectures On Physics’, and so I rarely use it for stuff like this. I also use it for printing and for personal emails.
Tonight, there appeared a YouGov survey. YouGov surveys do not tell you what the survey is about before you start the survey. Nor do they tell you who has commissioned the survey.
Normally, I am just too busy to bother, but on this occasion I decided to complete it. The survey started off with questions about how your view dairy-farming. Questions such as: Are cows well treated? Does their farting contribute to global warming [sorry, 'climate change']? Is bovine TB spread by badgers? Should badgers be culled? How many cows should be described as a big herd? [This being a particularly stupid question, I put 'a million']. And then, for no apparent reason came two questions about smoking.
Normally, these YouGov surveys are designed in such a way that you can go back to the previous question, but this one wasn’t so designed. I did not really pay much attention since I was more interested in seeing where this series of questions was going. It was only after I had answered the two questions (or was it three?) and wanted to go back and see what exactly the questions asked did I realise that I could not go back. Questions then moved to other subjects such as the use of ‘price comparison’ sites.
The smoking questions were odd. I cannot remember precisely. You know how oddly these questions are sometimes phrased. In this case there was one which asked: ‘Would you marry a smoker? Would you insist that your spouse stops smoking? Or words to that effect. The other question was similar – I cannot remember.
But the important point is this. If I am busy elsewhere, then I might just ignore the survey because I am given no clue as to what the survey is about. Or, as in this case, I might start the survey, see that it is about the care of cows and, since I know nothing about the care of cows, abandon it. Now… how came it that these questions on smoking appeared right in the middle of this survey about cows and price comparison sites? Why? Why? Why?
Can I help it if I feel that ‘there is something going on’? This isn’t the first time that YouGov have slipped questions about smoking into the middle of surveys on other matters. What is also odd is that YouGov have a running survey about people’s opinion of the smoking ban anyway.
So why were these questions inserted in the middle of a survey about cows????
Now…the boss of YouGov is also a big ASH advocate. If he slipped the word to Arnott that these two questions were in this survey, who would know? And if Arnott slipped the word by email to many others, who would know? But it does not need to be as specific as that. If ASH followers paid particular attention to YouGov surveys, expecting, every now and again, that smoking questions would appear in the middle of questions about cows, the perception of ‘public opinion’ could be seriously affected and deliberately distorted.
Tomorrow, I shall complain to the commission. I shall demand that YouGov publish, at the beginning of the survey, whom they have been commissioned by and give a list of the subjects. I shall also demand that I can skip those questions which I have no knowledge of or do not wish to be bothered about, since, at the moment, one has to answer all questions, even if one’s answer is ‘Don’t Know’. That will not do.
I shall make this demand, and see where it goes because I am fed up with being led by the nose.
We smokers are not protesting enough to the right people, but it is difficult. I feel sure that there are many people who would like to protest but feel a lack of confidence. It is a common phenomenon.
The Government Organisation which controls public opinion polls is called MRS. At this moment I know nothing whatsoever about it. I shall find out tomorrow, and I shall complain and complain and complain. Before I complete a survey, I want to know who commissioned it, and then I want a list of the subjects. I feel that I have the right to know.
20/01/2012 at 01:30
Isn’t Arnott a cow?
l’m sure if l hooked up with Angelina Jolie my first thoughts would be ‘Does she still smoke?’
20/01/2012 at 02:10
I wonder if Arnott could possibly be suffering from bovine TB? Especially if she has been familiar with badgers?
Angelina? A pretty trumper I think.
20/01/2012 at 02:23
I got a junk survey today,I wadded it up tossed it in the wood box and it produced heat………email really cuts back on my ability to get cheap heat for my home,but today I guess I got a few btu’s of heat from that survey and a few other flyers……..going redneck green works everytime.
20/01/2012 at 12:08
I agree, YouGov seems to be an important anti smoking resource. I did one a couple of years ago specifically designed to further their agenda. Some questions were ‘loaded’ making (my) honest responses nigh on impossible. I complained via the comment option at the end of the survey. Unsurprisingly, I haven’t been asked to do a tobacco related one since. I think this is how YouGov weed out dissenters, thus skewing results of following surveys.
Of course, we know that antis in general routinely do this kind of thing regarding ‘public’ consultation and epidemiology.
21/01/2012 at 22:18
I remember seeing a survey about smoking in cars (I think) on Mumsnet (how I came to be there, God only knows!). One question asked was blatantly loaded. It was something like: ‘Should children ever be exposed to harmful tobacco smoke?’ They got a 97% “No” answer, which is not surprising since very few would say “Yes – it is ok for children to be exposed to harmful tobacco smoke”.
But I don’t think that ‘the rules’ would allow YouGov to block you personally. Can they simply delete you from the panel without telling you? That is more likely.
21/01/2012 at 05:49
Excellent thinking on the strategy behind the insertion of those questions! I don’t think I would have given it a second thought beyond annoyance, but you may have hit the nail on the head. A survey about cows? 95% of respondents would simply close the window after the first two questions. So the total number going on to see the smoking questions would probably be very small. Loading even a dozen or two extra ASH types into it could make for quite some impressive antismoking percentages in the results.
Meanwhile, if they counted the number of people who simply BEGAN the survey they could make the numbers sound impressive. E.G.:
HEADLINE: 75% OF BRITS SUPPORT SMOKING BAN IN PUBS!
“Last month YOUGOV commissioned a survey that over a thousand British citizens responded to. On the questions about smoking 75% believed the current ban was a good thing!”
Now, let’s suppose my 95% estimate is actually correct for a moment and that the “over 1,000″ was actually 1,001. That would mean that only 50 people would ever normally get to the questions about smoking. Maybe about half of them would like the ban and about half of them wouldn’t. But now lets throw in fifty organized ASH staffers/supporters. This now gives us 75 favoring the ban vs. just 25 against it.
Quite a significant change for news headlines, eh? Even if the fine points of the stat math are a bit dodgy.
- MJM
22/01/2012 at 01:05
To be honest, thinking further about it, I am not sure that one can normally go back over previous questions – I may be getting mixed up with other surveys. But what I did notice was that at the end of each ‘section’, the little rotating circle appeared, which I assume means that your answers to that section have been transmitted. Thus, if you answered the first couple of cow questions, those answers would have been transmitted. Would you them be counted as having taken the whole survey, even if you abandoned the survey at that point? I don’t know. It may be that each ‘section’ is treated as a separate entity. For example, the cows section answers may be recorded so that, if you abandon the survey half-way through, those answers will still count, but only in that ‘entity’ – the cows section. The smoking section, however, may have only a fraction of respondents that the cows ‘section’ had, if people stop the survey after, or during, the cows ‘section’.
We can thus see that survey results can be skewed when smoking questions are placed in the midst of questions about the welfare of cows and battery hens and such, provided that X number of people with ‘special interests’ are aware that smoking questions can pop up at any time.
I do not know how many questionnaires YouGov send out on any particular survey. It may be that many people, like me, mostly don’t bother. It may be that, to get a response of 1000, they need to send out 10,000 questionnaires. Now, YouGov claim to have 360,000 panelists. Bearing in mind that ‘Common Purpose’ have trained 16,000 people and add, say, 20,000 other ASHites of one sort or another, one can see that it is likely that 10% of surveys are sent to ASHites randomly without anything underhand occurring. That means that it is possible that, of the 10,000 questionnaires sent out, 1,000 could be to ASHites. Thus, if X number of panelists are not interested in cattle welfare and battery hens, and stop the survey after the first few ‘sections’, only 10,000 minus X persons will be left. X could easily be 8,000 of the 10,000, which would leave a ratio of 1,000 ASHites to 1,000 non-ASHites, IF all the ASHites ALWAYS complete the surveys, regardless of the initial, apparent subject.
In view of the fact that Tobacco Control spent 10 years in preparation, nothing would surprise me. But, of course, impossible to prove. A couple of whistle-blowers might help.
24/01/2012 at 01:08
[...] Aha, another pressure group. The Meat Men are fighting back with Righteous techniques of their own. As for the credibility of YouGov polling, I think Junican has some clues for us there. [...]