Statistics I don’t believe in

Oslo is one of the candidate cities for hosting the 2022 winter Olympics, and of course there is a lot of discussion going on whether it costs too much etc. Opinion polls are conducted regularly, and the current status is that the majority of people don’t want Oslo to host the games. A couple of weeks ago the office dealing with the application put up the data from all polls conducted this year. What is weird is that they choose to illustrate the article detailing the data with the results from the question Do you wish that Norway should continue to participate in the winter Olympics and Paralympics? where only 73% of the respondents answered yes and 19% answered no.

First of all I thought the question was silly. Of course no one want Norway not to participate, but then the results show that almost one in five don’t want Norway to participate. At first i thought that it might be just a fluke, but then I downloaded the data and saw that this question had been asked on three occasions, and the results were pretty similar:


Even after seeing the consistent trend, I still don’t believe the numbers are correct. I think this shows one of the potential pitfalls in opinion polls. Since it is really hard to get the nuances of peoples beliefs in questions with just a few possible answers, it may be tempting to ask more detailed questions. The danger is that the questions become more confusing, or they even become irrelevant. I think both these things are can explain what is going on here.