00:01
We're looking at a survey.
00:02
So it's about going through red lights.
00:06
And we want to start with explaining how undercoverage might bias this survey.
00:13
So what is undercoverage? undercovered is a problem in your sampling.
00:18
It's when you aren't representing the entire population.
00:22
So some of the population, i'll just put pop n, is not represented.
00:31
Why might this be happening here? well, we have to look at how they take the sample.
00:37
They take it by randomly sampling listed residential telephone numbers.
00:43
That might not be the entire population that they're sampling from.
00:46
So they are missing anyone without a residential telephone.
01:01
And that means they are missing out part of their population, who might not have a residential telephone, perhaps renters may be less likely to, people who are new to the area, younger people.
01:13
So this is biasing the sample.
01:21
Now, what about response bias? response bias is where people change their answers for some reason.
01:32
So it's where the answers are affected...