This is not a poll.

If it were, i would take it to IMHO. My question is about polls, though. Why is there an option in many polls “Undecided” or “Not sure”… what’s the point of having that as part of a poll?

This is not a response.

you aren’t trying to tell me something, are you…?

OK. This is a response. You include “undecided” or “don’t know” in a poll for the fact that often people don’t know enough about the subject to have an opinion. I suppose you could put a gun to their heads and make them choose, but that tends to skew the result.

No one’s ever asked you for an opinion on something for which you didn’t feel comfortable enough (or knowledgeable enough) to have one? Or presented you a choice of responses, and none of those choices seemed adequate or correct or complete?

Wow, you must be a genius!

Maybe someone could just respond by saying, I don’t know enough about the situation, so either educate me, or don’t count my vote…? Seems to me it would be more accurate…

On-the-spot education would probably be biased by the surveyer, so that’s out. The undecided/not sure category is to show how many people were polled without skewing the rest of the categories. It also shows how many people don’t know about the poll subject matter.

To give an extreme example, let’s say I poll 100 random people on which of my children they like better. 2% say my son, 3% say my daughter, and 95% are not sure. In my final report, would it look better to show that I only polled 5 people by listing all but the “not sure” results, or by showing all the results?

In most polls, you do not have an option of your own personal “custom” response. You must select from their choices, or refuse to answer at all. That’s because they cannot meaningfully quantify answers outside their pre-chosen options.

That wouldn’t be a fair poll then. If the objective is to gain data on people’s opinions, then not having an opinion is a very important part of it.

For example, if it’s a poll on which party leader you want to run a country, and 60% say “don’t know” or “don’t care” then that’s hugely important to the people trying to get voters to feel part of the political process. It’s like spoiled ballot papers; knowing who didn’t want to take part tells you about the level of dissatisfaction, disenchantment or ignorance of the issues involved.

I should have added: not counting negative responses produces far more inaccurate results than demanding either-or answers. Additionally, as well the problem of bias in getting pollsters to explain the issues, they probably don’t have the time and knowledge to explain them fully either. I’d like to see the MORI chap try to explain the details of economic and monetary union in Europe to me in five minutes!