Opinion polls (Teaching kids to think: Part II)

So apparently some significant proportion of people are in favour of same-sex marriage. Or the same significant proportion are against. Or 3/4 of people surveyed think it's inevitable. What does it all mean?

In our democracy, public opinion polls are tempting and persuasive, I think because of our democratic basis of government. The numbers matter, because we are meant to make decisions based on what the majority wants.

But in another sense, the numbers don't matter. A position can be irrational, even when many people favour it. A decision can be a bad decision, even when the majority votes for it. Reality does not respect democracy.

We should demand some answers before we accept the results of any opinion polls. Who commissioned the poll? Do they have a bias that is more likely to skew results? (Rarely, some biases may lead to a more open mind about results) Who conducted the poll? Were they operating according to accepted statistical methods, or was it a "just click here" newspaper-type poll? What questions were asked?

If we're satisfied about these issues, the poll might be a reasonably accurate reflection of public opinion. If so, it can be helpful in the democratic guidance sense. But even if this is so, they cannot help us determine if a position is rational or a decision is sound. Why do respondents support a position? Are they worried about how they'll look if they don't support the position? Do they just not care, and will run with the majority? Are they scared or intimidated into support? Have they thought through the issue and come to a reasoned conclusion, or are they parroting slogans?

What matters are the reasons behind the opinion poll answers - if we can possibly get at those reasons, and assess whether they are sound and rational (or not!), then we are getting some way towards good decision-making.