Cartoon from the Seattle Times

Times have changed since the reliable old rotary telephone joined the Model T and the locomotive cowcatcher. It used to be that people generally answered the telephone and cooperated with pollsters. Then the world of telephony changed. Nowadays it is not unusual for robocalls selling gonzo vacation packages, suspicious medical insurance, and such to outnumber the calls to which we actually like to respond. Moreover, with Caller ID “Spam Risk” notification, it has become downright difficult for pollsters to get a live respondent.

What happens when one gets through to me? I just say “I don’t respond to polls or surveys” and hang up before the caller can inhale.

Then, too, the multiplication of cell phones has made it chancy to poll a household with no landline and multiple cell phones. I have both a landline and a cell phone. The latter is off most of the time because I was annoyed by receiving numerous robocalls in Mandarin Chinese; so I just use my cell phone to call out when traveling.

There is an interesting PBS website called “The Problem with Polls” that gives you an idea of the problems faced by research organizations.

What surprises me is how polls that wildly contradict one another continue to be news. My assumption is that instead of a one-digit margin of error, it is probably closer to ±25% or more.