A Christianity activist group called the Alpha Course has been running an online poll to determine what proportion of those taking the poll believe in god or not. The poll was rather cynically loaded by having two positive options (“yes” and “probably”, the latter of which is intended to capture as many “maybe” votes as possible) and one negative option (“no”). Naturally, if this had worked the way the Alpha Course intended, they would be crowing about how only x-percent did not believe in god, etc..
Unfortunately for the Alpha Course, their assumption that the majority view aligned with theirs was spectacularly blown out of the water by the poll result which – at the time of writing – shows that 97% of respondents think there is no god.
Naturally, their response to this was that the poll result must have been rigged by a coordinated campaign to skew the results with the Alpha Course spokesman, Mark Elsdon-Dew, going on to say that “I don’t think this is indicative of people’s faith in this country”.
Well, yes – it is possible for any online poll as superficially implemented as this one to be deliberately skewed (let’s not forget the skewing factor of the loaded poll options, by the way).
On the other hand, it’s also possible that the general result is correct – even if you allow a fat margin of error for deliberate skewing votes, the results are so polarized that one would still end up with a negative majority.
If nothing else, this is a great example of confirmation bias – if the result agreed with their own beliefs, it would be lauded as a vindication. Now that it’s overwhelmingly contrary to their beliefs, it just has to be a deliberately skewed result that doesn’t represent the “true” sentiment.
Statistical research and analysis needs to be conducted with exacting discipline. One of the reasons why statistics are so notoriously unreliable is because they are, more often than not, gathered by people who just don’t know what they’re doing! Without truly understanding the science of statistics, it’s all too easy to bias or misinterpret the outcome (deliberately or inadvertently).
As for this fatuous poll by the Alpha Course, well…they only have themselves to blame 