Thu | Dec 13, 2018

Letter of the Day: Polling vs hunch?

Published:Saturday | March 5, 2016 | 12:00 AM


Whether the social scientist called her task polling or forecasting, our recent general election results (so far) have been intellectually damaging to many (most?). One damage-mitigating option is to take refuge in 'margin of error'.

Margin of error is not just an academic admission of built-in limitations of the task of political polling or forecasting, but a tacit and thus easily missed point, that polling/forecasting is tantamount to guesswork, though clothed in sophisticated academic language and qualifications, provisos and caveats.

My respected and academically superior friends in the social sciences will forgive me because I know precious little about the disciplines since I have never taken any such course despite my long years of study (undergraduate and postgraduate). But ponder this: On the morning before election day (February 24, 2016), at about 6:15 a.m., I sent the following email to the team doing 'The Morning Watch' on LOVE 101FM:

"I don't call the Lord's name on anything, so my hunch is that the JLP will upset the PNP and there will be a few seat shockers. If Mrs Holness wins a seat, it is all over, but for the actual general counting. That's my one-eyed story and I am sticking to it." I am recovering from retinal detachment surgery on my left eye.

Now you judge: No knowledge of the social sciences, no prophetic word from my God, but the sophisticated social scientists who dared to pronounce in public fared no better than my hunch.

I would recommend to all social scientists that they suck humble salt and find and read Chapter 8 of the 1984 edition of philosopher Alasdair MacIntyre's book After Virtue, which has the caption 'The Character of Generalizations in Social Science and Their Lack of Predictive Power'.

I close with one, hopefully inviting, quotation from that chapter: "... [T]he aim of the social sciences is to explain specifically social phenomena by supplying law-like generalisations which do not differ in their logical form from those applicable to natural phenomena in general, precisely the kind of law-like generalisations to which the managerial expert would have to appeal ... . The salient fact about [the social sciences] is the absence of the discovery of any law-like generalisations whatsoever." (Page 88)

Interestingly, the standard for a genuine prophet in the Bible allows no margin of error. It's 100 per cent accuracy every time or death!