00:00
The next kind of bias we're going to talk
is response bias and that's when subjects
tell you what you want to hear, it's similar
to interviewer bias, but interviewer bias
it's the interviewer who is eliciting the
response, when response bias is when the interviewee
is telling you what you want to hear, whether
or not you elicited that response or not.
So there's an example, let's say you are doing
a study on the Haiti earthquake that happened
a few years ago and you asked people prior
to the earthquake in Haiti, 'Did you know
the name of Haiti's capital city?', it's entirely
likely that they didn't know the name of Haiti's
capital city, but they're afraid of sounding
uninformed and so they are more likely to
say, 'Oh yes, I know the name of Haiti's earthquake
and I've always known it', because they want
to please the interviewer, that's when the
respondee has a tendency to want to give you
the answer that you want to hear. Some people
confuse response bias with reporting bias
and they are quite distinct concepts. Reporting
bias is a real problem in science communication,
this is when we as scientists have a tendency
to only report positive results and publish
positive results, sometimes it's called publication
bias. Again it's a tendency for us to only
report statistically significant results.
01:23
Related to reporting bias is confirmation
bias, that's when doing our research we tend
to only cite other research that agrees with
us, that's not a good thing in epidemiology,
we want to assess the entire universe of information
and get to the objective truth of the results
that we're seeking.