Do elections bring out the worst in research?

Two weeks ago, I received a call on our landline phone from a local number and woman’s name I didn’t recognize. When I answered, a man said that he was conducting a nationwide political poll. Without identifying himself or the company or organization he was representing, he asked me two simple questions:

  1. Who do you plan to vote for in the upcoming presidential election?
  2. Do you feel that the country is headed in the right direction or going down the wrong path?

Then, he thanked me for my responses and hung up.

This is the first presidential election in which I’ve been of voting age and had a landline so I’m not particularly knowledgeable in what telephone polling entails but this struck me as very odd. The caller didn’t identify himself, didn’t ask me to identify myself – neglecting to ask even if I was over 18 years old – and asked a very subjective question. I also found it strange that the number was local and linked to a residential name but then again, our home phone shows up on other people’s Caller ID as “Lana Stark” so I can let that one slide.

Wanting to give the benefit of the doubt, I thought it might have been a company I’d answered a survey with before that already had my demographic information on file, saving me the trouble this time around. In that case, kudos!

But then I got a call last Friday from the same number. I answered, thinking maybe it was a follow-up call. It wasn’t. It was the exact same call as before. I was asked two questions and then abruptly hung up on without a chance to get a word in edgewise to say, “Hey, I’ve already answered this poll.” So not only does the company really not know who I am, it also has my responses counted twice in its “nationwide poll.” If this company can’t even keep track of who’s already answered the exact same poll, how much stock can we put in the results? I would argue not much.

My first telephone polling experience was a laugh and a half and I’m wondering: Is this a one-off or par for the course? What’s the point in conducting a poll like this if no one takes the time to ensure that the data is valid and unduplicated? Are there companies so desperate to get a piece of the election-time pie that they will be lax on quality control to get the job done? Has anyone else had a similar situation? What should a sound political telephone poll look like?

This entry was posted in Political Polling, Public Opinion/Social Research, Quantitative Research. Bookmark the permalink.

2 Responses to Do elections bring out the worst in research?

  1. Annie Pettit says:

    Well, this clearly wasn’t a call from a registered market research company. There are very strict guidelines on how they must identify themselves and this didn’t meet those criteria. So, as a measure of people who aren’t in the market research industry attempting to conduct research, it’s probably fair to say that the quality just isn’t there. Oh yeah, ANYONE can do research. #Sarcasm

  2. Rob Bixler says:

    Pollsters typically only call registered voters who voted in the last election. Their goal is to predict the election. They only care about your prediction of how you will vote. There are certainly telephone surverys conducted in elections that are used to test messages and whether messages shift your intentions to vote for a candidate.

    No telling what this actually was, but that you were just asked one or two questions is not that unusual.