Among all the comment on the way results of the May Election differred from the predictions of the polsters, one comment stood out to me as being the most interesting.
Was it because the sample was inadvertently biassed?
Think how the data is gathered. Do you ever get asked to complete a questionnaire, on-line, by ‘phone or in person? Do you respond? What makes you respond or decline?
What kind of people are more likely to spend time dealing with an opinion poll?
I suggest businessmen, professional and senior managers are all too busy to bother.
On the other hand, students, unemployed people, and people in boring jobs, looking for an excuse to take a break, are all more likely to reply. Which groups are more likely to be Labour supporters, which ones Tories?
Could this be why the polls got it wrong?
I wish I had included this in my book on statistics.
How To Avoid Being Misled By Statistics Don’t Be One Of The 60% Who Are Below Average
Go to: https://www.createspace.com/4767398
Perhaps I will include it if I write a sequel!