Thursday, January 28, 2010

Abusing Surveys

The very least they can do is be consistent. Every political mouthpiece in this country who mentions a survey in the media is either for or against the results. Those who approve will praise the professionalism and neutrality of the survey institution. Those who find the results unfavorable will imply bias and question the integrity of the survey organization.

Among the slurs hurled at survey takers is that someone "paid" for said survey. Which is entirely true: someone does pay for the survey. But that doesn't mean that they dictate what results will come out. It doesn't matter if Villar or Zamora or the Liberal Party commissioned the survey (or, more accurately, subscribed to the results), what SWS or PulseAsia presents are what the respondents answered, tabulated and processed.

What the opinionated barbershop tambay (who, in terms of quality of analysis, is at par with a lot of the partisan attack dogs) needs to keep in mind is that paying P400K for the survey results isn't the same as buying first place. What the subscriber gets is a detailed report and usually a couple of rider questions he can ask. That's all.

The tambays need to remember that these organizations don't just appear every three years. They conduct surveys all the time; it's how they make a living. And they can't make a living if their reputation is tarnished by allegations of selling results. For surveys to work, the surveyors need to follow protocol and be honest with what they find out.

Now, assuming the tambay knows this, there's still that other line of attack that calls into question the methods employed, most especially sampling. "Bakit 500 lang ang respondents?" someone would ask. Well if they paid attention to statistics in college they'd understand. Sadly, few people do.

Of course surveys work. They're never 100% accurate, but they work well enough for a lot of businesses, for example, to base decisions on. There are large organizations that make a decent living off doing surveys and studying various markets / audiences. The savvy political organization knows what the results mean and how to use that knowledge to their advantage. Sadly, a lot of them think that bandying their lead is the only use of surveys.

Surveys are tools that the public and politicians misunderstand and misuse all the time. To paraphrase David Ogilvy they use it the way a drunk uses a lamppost: for support instead of illumination.

No comments: