(nods) Surveys are problematic that way, in general. The only way I know of to get around it is to phrase every question several different ways and look for variation among the answers based on the phrasing.
The safest move is probably to simply discard any question where the answer depends too much on the phrasing, although in practice that probably means discarding all survey results ever.
Mostly, survey results are good for comparing results on the same survey over time.
(nods) Surveys are problematic that way, in general. The only way I know of to get around it is to phrase every question several different ways and look for variation among the answers based on the phrasing.
The safest move is probably to simply discard any question where the answer depends too much on the phrasing, although in practice that probably means discarding all survey results ever.
Mostly, survey results are good for comparing results on the same survey over time.