Pollsters, the media and opinion polling

Accuracy and usefulness

Letter of the Day
Letter of the Day

A chara, – Justine McCarthy’s assertion that opinion polls are self-fulfilling prophecies (“Do polls merely reflect public opinion – or do they shape it?”, Opinion & Analysis, June 7th) would appear to be flatly contradicted by the results of the recent referendums held here. And, as she notes, the Brexit referendum. However, in the case of Brexit polling, data always showed a large number of “undecided” voters. These swung more towards “Leave” than “Remain” when faced with an actual ballot paper.

Newspapers and other media outlets commission opinion polls because they fill large amounts of column inches or minutes of airtime. This, naturally, leaves those outlets poorly placed to question their accuracy or usefulness. For example, the “margin of error” is always presented as the possible difference between the results of an opinion poll and what might happen in the event of an actual election being held at the same time.

It is in fact the sampling error, that is to say the in-built error that occurs when 1,000 respondents are scaled up to reflect the views of a few million. Other and, as we have seen, larger errors can arise. Such as respondents being reticent to admit that they will vote in a way regarded as being “wrong”, being reluctant to state they do not intend to vote at all or even just wanting to end the call with the pollster as quickly as possible without being rude and so just giving any answer that comes to mind.

There is also the very real possibility that those who intend to vote a certain way are less likely to engage with pollsters in the first place.

READ MORE

My prediction is that “far right” candidates will fare much better in these elections than opinion polling expects and for the reasons outlined above. – Is mise,

DAVE SLATER,

Kilkea,

Co Kildare.