Differences in polls can be attributed to many things, including question wording and order.

Some modes require different question wording from others, in order to suit the features of the mode.

Differences in question wording across modes may cause different data to be collected by different modes.

Changes in the question wording used also make the meaningful interpretation of time series health data problematic.

Variations in question wording or the order of questions, for example, can lead to somewhat different results.

Therefore, nuances of question wording and the placement of questions within the interview can affect results.

Whether closed or open questions are used there are many problems of question wording.

Also variations in question wording or the order of questions can lead to somewhat different results.

Motivations were measured in the second fortnight of the campaign, but the question wording was not time-specific (see Chapter 2).

On some issues, question wording can result in quite pronounced differences between surveys.