News for hackers. Stuff that matters.
Re:The great British Brexit robbery by dpwm on Sunday May 07, 2017 @08:18AM (#14285083)
There are many problems with drawing inferences about overall public opinion from programmes like BBC Question Time. The main one is sample bias. Sampling occurs in three stages:

1) Audience members are self-selecting.

2) Audience members are then selected from the group that self-selected to represent a cross-section of the population.

3) Not all audience members get to ask their question, they are selected by the host.

The problem with the second stage is that there is a person, called an "audience producer", who selects people. When that person reposts the social media output of Britain First, an extremely far right political organisation, it questions whether they may have a political bias that may affect the sampling [0].

Watching Question Time could quite easily convince you that you are in a majority or a minority and you could be wrong. Treating Question Time as any kind of meaningful measure of public opinion is a form of self delusion, albeit one that's hard to shake off even if you know it. The only thing a person speaking on Question Time could tell you is that there exists a person with those views, and even that might not be true. Same goes for any other audience panel.

[0] https://www.opendemocracy.net/ourbeeb/tim-holmes/is-question...

There are no comments on this comment yet.