let’s talk about…polls.
not the long wooden things used to propel boats in shallow water, but the statistical tools used to determine how people feel about something or someone.
specifically: presidential polls.
in most of the presidential polls john kerry is losing to george bush.
which sucks.
but which is also statistically irrelevant.
these polling companies stay in business by convincing people that their work is significant and relevant.
but it can’t be, except in the most general sense. and here’s why…
polls are done over the phone.
over ‘land lines’, to be specific.
because the only phone directories that exist are for land-lines.
so here’s my question…how many of you primarily use a home land-line when using the telephone?
not many, right?
i’d say that 1/2 of my friends don’t even have a land-line, and the other 1/2 only rarely use their land line.
people under the age of 35 primarily use cell phones.
and if you primarily use a cell phone then you’re not going to be taking part in any of these polls, because they’re only conducted over home land-lines. example: in the last new york state senatorial election all of the polls had hillary clinton ahead by 4%, but she won by 13%.
why the big gap? because the polls were done with people primarily over the age of 35 who were at home when the pollster called.
but the companies that conduct these polls can’t say, ‘oh, by the way, our methodology is completely outdated thus yielding our results irrelevant’, because then they’d be out of business.
so todays lesson: polls are, statistically speaking, fairly meaningless.
yes, john kerry might be trailing george bush among voters over 40 who stay home during the day, but i have a feeling that john kerry is probably doing pretty well with voters under the age of 35 who have jobs and primarily use cell-phones.
-moby
