Sunday, September 13, 2009

Does God exist?

I've just got back from Market Harborough, where I went to experience this year's Arts Fresco. But that's not what I want to talk about right now. We parked near Emma's school, and then walked into town, passing this billboard on the way:

That's what I want to talk about*.
Look at the picture above and choose one of the available answers to this question: Is this a rubbish, heavily biased survey question or what?
a) Definitely
b) Yes
c) Probably
d) Thursday
e) No.

I am many things. Just two of those things are Atheist and Mathematician. I feel qualified to take some slight umbrage at this piece of advertising from both, or rather a mixture of both, points of view. For those who don't know,
'Alpha is an opportunity for anyone to explore the Christian faith in a relaxed setting over ten thought-provoking weekly sessions, with a day or weekend away.' [In their words, from their website]
My own experience with the Alpha Course is limited, so my views about it are likely to be a little biased. I have been pointed towards the course on a number of occasions, usually when a debating opponent has reached the end of their reasoning tether with regards to the existence of a God or Gods**.

But anyway, back to the poster. It annoyed me. Not because it was an advertisement for someone's religion; if you've got a club you're welcome to recruit members as far as I'm concerned***. It annoyed me because the whole God(s) / no god(s) debate is a long running and often heated one, and I don't like it when one side tries present so-called statistics which have been manipulated or collected in such a way as to skew the results in their favour.

As a maths teacher, at some point in the next few weeks I will be teaching at least one lesson on what makes a good (or bad) survey question. It's part of the AQA Module 1 handling data syllabus, and there are a number of 'why is this a bad survey question; how would you improve it?' type questions evident in the last few years' past papers. This billboard is a prime example, and I may well make use of it in future lessons.

Why's it a rubbish survey question? It's biased. That's the most fundamental survey flaw that GCSE (and earlier) mathematicians are trained to look out for. Survey questions should not be leading. 'You do like football, don't you?' is the kind of thing that's demonstrated as poor questioning. But the available responses are also a rich source of bias and attempts to lead people into picking the 'right' answer.
Lets look at the question in, er, question:

Does God exist?

Taking things back to absolute basics, lets assume that someone answering this question does so by picking one of the available responses at random. That means that each response has a 1/3 chance of being picked. But in reality, most people won't pick an answer at random, particularly on a survey on this subject!
So assume that everyone will either pick yes or no. That seems fair, doesn't it? You either believe or you don't. Right? No, actually. In a brief, internally conducted straw poll of the perceived standpoints of the people that I know, most people appear to be undecided on the matter. So they don't want to vote for a definite yes or a definite no. But look, here's a third option: probably. Well, it's not committing myself to a definite, so I'll have to take that one...

But it's still in favour of the existence of the questioner's God. It isn't 'I don't know.' It certainly isn't 'probably not'. It's not even 'maybe'.

So, back to the idea that everyone votes randomly: that's a 2/3 (almost 67%) likelihood that you'll pick an answer in favour of the thing that the questioner undoubtedly wants you to vote in favour of. Even when you take into account that most people won't be voting randomly, the closest thing to a neutral answer will grab many people who are quite happily on the fence or undecided about the matter and claim them as positive responses.

I think the most important question that such an advertisement conjures up is this: Why would a considered, rational thought or belief system have to resort to such blatant bias in its advertising****?

Feel free to agree / disagree / debate / comment. But please remember that I'd like to keep this blog a place that's suitable for all audiences, so profanity will cause your comments to be deleted.

Don't forget to head over to Mathsqs to ask your maths related questions!

* The Alpha Course website has this very poll on its front page. Why not vote? Just for fun.
** Mostly it's just 'a' God, as it's a Christian course and, as far as my limited experience leads me to believe,  doesn't give the possibility of any of the world's many and varied, and just as sensible, other religions any airtime.
*** As long as it doesn't promote such pastimes as rape, genocide, misogyny and slavery. Any more.
**** Yes, I realise the irony in this post: I have supplied the Alpha Course with a little bit of free advertising. But I'm considering the fact that very few people read this blog, and those that do aren't likely to be here for the boobies,