Good research
shows its work.

Who we are trying to understand

Before fieldwork begins, Australian Poll will define the population the research is intended to represent.

Depending on the study, this may include:

• Australian adults
• Enrolled Australian voters
• Residents of a particular state or territory
• People within a particular age group or demographic
• Consumers within a particular market
• People living within a defined geographic area

The target population will be disclosed with every published study.

For the planned Western Downs study, age and council residence are self-declared and eligibility-screened. Email confirmation is separate from identity verification. Response quality review does not verify residence or guarantee one response per person. Accepted responses alone do not establish a representative sample.

Back to contents

How participants are selected

People may register with Australian Poll to be considered for future research.

Joining Australian Poll does not automatically make someone part of a poll.

Our registered research community is a pool of potential participants. It is not, by itself, a representative sample of Australia.

For individual studies, participants may be recruited through methods including:

Telephone research

Participants contacted for a specific study.

Australian Poll research community

Selected members invited based on study requirements.

External recruitment sources

Where appropriate, additional recruitment methods may be used.

The recruitment method used for each study will be disclosed.

Back to contents

Sample design

For each study, Australian Poll may establish sample targets based on characteristics relevant to the research.

• State or territory
• Age
• Gender
• Geographic area
• Metropolitan or regional location

Additional variables may be considered where methodologically appropriate. Sampling targets and relevant limitations will be explained in the study methodology.

Back to contents

The way a question is asked matters.

Australian Poll aims to use questions that are clear, neutral and understandable.

We will publish exact question wording and response options. Where relevant to interpreting a study, we will also disclose:

• Question order
• Whether response options were rotated or randomised
• Treatment of undecided respondents
• Relevant introductory wording or prompts

For political voting-intention research, Australian Poll will take particular care to minimise unnecessary priming before voting-intention questions.

Back to contents

Why we weight data

Survey samples rarely match the population perfectly. Where appropriate, Australian Poll may statistically adjust responses using independent population benchmarks.

Possible weighting variables may include:

• Age
• Gender
• State or territory
• Geography

For every study using weighting, we will identify the variables, benchmark sources and general weighting approach used.

Weighting does not automatically remove selection bias or make a sample representative.

Back to contents

Federal and state election research

For voting-intention studies, Australian Poll will clearly disclose:

• Target population, such as enrolled or eligible Australian voters
• Primary-vote question
• Treatment of undecided voters
• Preference methodology: whether preference figures are respondent-stated, based on previous-election preference flows, or modelled
• Fieldwork dates
• Sample size
• Recruitment mode
• Weighting methodology

Back to contents

Polls are estimates, not exact counts.

All survey research contains uncertainty.

Where Australian Poll publishes a margin of error or other uncertainty estimate, we will explain how it was calculated and whether it applies directly to the sampling method used.

Other sources of survey error can include:

• Non-response
• Coverage differences
• Question wording
• Recall error
• Sampling and recruitment bias
• Weighting

A reported margin of error should not be interpreted as capturing every possible source of survey error.

Back to contents

Our transparency standard

Every Australian Poll research release will include:

• Who conducted the research
• Who commissioned or funded it
• Target population
• Fieldwork dates
• Survey mode
• Recruitment method
• Sample size
• Question wording and response options
• Weighting information, including whether weighting was used
• Treatment of undecideds where relevant
• Preference methodology where relevant
• Relevant limitations

Back to contents

Corrections policy

If Australian Poll identifies a material error in published research, methodology, calculations or presentation, we will correct the published material and clearly disclose what changed.

We will not silently alter material findings.

Back to contents

Participant privacy

Australian Poll will handle participant information in accordance with its privacy obligations and research requirements.

Public research results will generally be reported in aggregate rather than identifying individual participants.

Back to contents