Traditional surveys – how reliable are they?
Here is an interesting article from the Economist about political polling in the US. The article discusses the increasing difficulties in conducting polls or surveys that assess voting intentions in the US.
Most polling companies, in the US and elsewhere, conduct their surveys by calling phone landlines (fixed lines). But less and less people are using landlines – the article states that some 25% of US residents only have a mobile phone these days. Polling companies often don’t call mobile phones for various reasons, mostly related to cost. So the conclusion is, be careful when looking at survey results based on this traditional approach.
Interestingly, the article did not mention the growth of surveying using the Internet – or the possibility to survey using smart phones.
This article from FiveThirtyEight blog provides more insight into the issue – mentions the growth of Internet polling and is not so pessimistic about the future of traditional surveys.
For evaluation, the debate is interesting as often we use surveying as a tool – and many of the points discussed are relevant to the surveying undertaken for large-scale evaluations.
Entry filed under: Evaluation tools (surveys, interviews..).