There are two kinds of surveys. Those in which you yourself
want to find out something and those which aim to confirm the
way you see things.

The requirement that surveys are carried out scientifically ought to go
without saying (empirical social research). But when it comes to using
the findings of a particular survey in your communications work, it is not
enough for the survey to measure the right thing correctly. The findings
must also be presented in a way they can be assimilated quickly.
The trick is to reconcile scientific conciseness with usability – e.g. for
media work.