Questions for the West Australian WESTPOLL Survey this month cover:
Survey conducted 3-6 November 2008
|
Our Services | Flashpoll | Westpoll | Other Services About Flashpoll | Accuracy | Case Study | Dates Sample Size | Random Selection | Answering Machines | Engaged Numbers | Small Subsets We use a telephone sample of 300 metropolitan and 100 country people aged 18 years or older in Western Australia. A sample of 400 across Western Australia gives a theoretical survey error of ± 2.9% at the 70% level of confidence. At the 90% level of confidence, the survey error is ± 4.1%. Patterson Market Research’s survey program enjoys an enviable record of accuracy in predicting electoral outcomes. The overwhelming majority of electoral predictions made on the basis of Patterson Market Research’s polling have proven to be correct. This is why the West Australian Newspaper continues to support the Flashpoll program (published as Westpoll) after an association of 18 years. A multi phase random selection process Due to the active lifestyles and high mobility of Western Australians, we have developed a survey "call back" routine, which we believe, provides the most accurate and effective reflection of the population profile possible. Essentially, this is a two-phase call back routine. Phone numbers that are unanswered on our first attempt are called again (up to a maximum of three times during the survey program) after a minimum of two hours has elapsed. Once contact has been made with the target household, a "within household" random selection is made, and up to two further calls are made to appointment in an attempt to secure an interview with that randomly selected individual. There has been a rapid growth in the use of answering machines in Western Australia, and we have addressed this by treating answering machines as numbers which are unanswered, and calling them back after a minimum of two hours has elapsed. As is the case with unanswered numbers, up to three call attempts are made to contact households which use an answering machine. Engaged numbers are called back on 15-minute intervals in an attempt to secure an interview. The fortnightly schedule makes it possible for smaller subsets to be accumulated over several waves of the Flashpoll program. We term this process "data banking". It enables you to build a sample of people over two or three survey waves, (four or six weeks). This can be very useful if you are interested in a subset of the community; for example people who have shopped in a particular store franchise in the last couple of months. |