Opinion Polling Characteristics
Cost Scale: The costs of opinion polling is dependent on the number of participants and questions, sampling method, question format and survey-type. Purchasing questions on existing surveys costs in the region of £100-£300 per question per 1000 respondents on an omnibus poll.
Time Scale: the factors mentioned above for cost also influence timescale. They can range from 1 day -1 month to implement. Analysis of the data also takes time, with the time taken dependent on the scale and complexity of the data produced.
Opinion polls give policy makers the opportunity to receive a large number of responses to a number of questions, and are not designed to provide and ongoing engagement or dialogue. As a result, they are less suited to understanding the reasoning or values that inform responders views.
The wording and design of opinion polls also have the potential to influence participants’ responses, due to order effects, framing of questions and wording.
As policy makers are responsible for framing the questions, the scope of opinion poll participant’s responses are limited. Whilst open-ended questions can broaden this scope and increase the range of opinions heard, analysing the data they produce is relatively costly and time-consuming. Participants typically respond on an individual basis, meaning participants don’t have the opportunity to respond to or build upon others’ views.
Opinion polls can involve large numbers of participants to produce statistically significant statements about the sample group. As facilitators are not required in all cases and survey questions are the same for all respondents, the number of participants are not limited by such factors.
Opinion polls provide one-way feedback to policy makers for a range of pre-determined questions. Participants typically respond to surveys without deliberation with other participants or specialists, although relevant information may be included as a preface to some questions. Policy makers should bare in mind how such information may influence participant’s responses, and use impartial language and balanced information where possible.