Wednesday, February 9, 2011

surveys and data gathering

Every now and then I get a call to my mobile, asking me to take part in some survey or other. There's something intrusive about cold calling, so I invariably tell them I'm don't have time, and if they're too slow or persistant to realise i'm just trying to be polite (would you have time tomorrow?) then of course I rachet up my refusal!

However, when I think about it, as someone who loves reading the results of the latest pscychological or societal study, which often rely on mass surveys, this isn't reasonable behaviour on my part. After all, if everyone had my attitude, and there would be no one to answer the needed questions. So perhaps I should view it as one of those necessary tasks we all should undertake to help society at large - things which sometimes seem to be not just an inconvenience but an irrelevance to us, and yet which may have real value to the community as a whole.

Of course, another mental excuse which often comes into play is the idea that it is probably anyway commercial, and why should I help some marketeer in his targetting of me and my disposable cash. But this also doesn't make a lot of sense, since in essence what I'm saying is, how dare they try to figure out what I might want! Surely it is in general a good thing not only that the products we want are available, but (to reduce waste and energy consumption) those we don't want, aren't!
I suppose the source of hostility arises from an association of such practices with advertising, but this can't really be the case. Advertising is about creating a want without a need, but this doesn't work if you do the asking first!

This is I think related to the more general practice of questionaires and data gathering. While normally this is again just at the level of inconvenience, there are also times when it involves issues of privacy.

For example in Kindergarten they now I think do do some questioning/assessment of the kids and their behaviour and abilities. The problem which some people then see is, such data sets can be used to build up a profile/history of a child over its school life. And the problem with that is, maybe something from a child's past will influence how it's later actions/behaviour are viewed. If a child had learning/behavioural problems at an early age, then this might colour later interpretations. In an extreme view there might be the idea that a child is pre-judged based on supposed 'early indicators'. I.e. because he/her exhibited such and such in pre-school, then he/she *must* have such and such a disposition or disability.

However, being honest, I think the potential for good is much more than the potential for harm. Even if such a link between early traits and later behaviours was possible, it could never be a conclusive one. So its use in a negative manner is limited. However, the link only needs to be probable for it to justify 'pre-emptive' actions at an early age - which even if over-applied based on such data analysis, would mean it would reach the kids who really need it, and would be very unlikely harm the one's who didn't (who would be harmed by extra tuition etc.?).
Of course if this was used as a basis of 'streaming' kids at too early an age then this would be another matter - but again the point is the scientific evidence wouldn't be strong enough to make a conclusive judgement, which would be required to impose such measures. But for supportive extra measures no one is going to require it to be shown 'beyond reasonable doubt'.

The point is such data gathering really only has 'forward-looking' uses - what things will be sold, what behaviours might appear in the future etc. It is statistical prediction, and as such only has proper validity, and use, at the group level. At the individual, specific level, what someone will really do or why they really did, can't be extrapolated from that

So I guess it's a case of don't ask what your society might want to do to you, but what you can do for your society, when it asks you.