What Do The Polls Say About Dinner?
link More than half of U.S. Latinos favor same-sex marriage: survey | Reuters.
Seems that everything is done by polls these days, especially when it comes to election time. Everyone wants to slice off a bit of demographic to imply significance from some bit of esoterica. The standard technique is to isolate some segment of society and then make conclusions about their preferences as it pertains to the larger group. This kind of segmentation has a lot to do with the messages being conveyed by the various political groups who would pander to whomever is likely to give the most votes.
An Internet search fails to find even an estimate of the number of polling firms operating in the U.S., but we are all familiar with the big name ones: Gallup, Rasmussen, etc. There are also the newspaper related polls such as the New York Times, Washington Post and Reuters to name a few. Suffice to say though, if you want an opinion about something, someone will provide it for a fee.
The ubiquity of polling has created a political environment in which factions are pitted against each other in order to secure votes, which of course encourages polling to measure the temperature of any identified demographic, which brings us into a vicious cycle. If you are catering to left handed people, the right handed people get disenfranchised and resentment between the two factions will escalate. When people bemoan the polarization of the political process, this is ground zero of the cause. Instead of saying anything substantive, politicians will employ strategies to rope in the desired demo as articulated by poll results.
The particular sample in the link above purports to show the proclivities of 26 million people via a sample of 1760. The conclusion is that more than half of U.S. Latinos favor same-sex marriage. I don’t know about you, but I think that this is not only a leap of logic, it’s a 3 day train ride of logic based on a sample of 1760 Latinos. I would treat this conclusion with as much skepticism as if a sample of college men in San Francisco implied that 60% of college guys everywhere harbored homo-erotic fantasies involving Brad Pitt. Angelina maybe. To imply that all persons of an ethnic slice are likely to hold homogeneous opinions is naive at best, racist at worst. I’m not convinced that all Chinese, all Germans or all Italians can be pegged by a sample of 1760 of them.
The real danger here is that in today’s attention deficit society, nobody reads the body of the polls. They read the headline and ka-bing, it’s now a fact. This is reminiscent of the early days of television advertising when marketers could say pretty much anything they wanted. It was not uncommon to hear that “4 out of 5 doctors preferred Parliament cigarettes” or that “4 out of 5 Dentists choose Crest”. Bottom line is, polling is as much marketing as science. My poll of 5 people confirms this.