Methodology & limitations

How Civic Pulse weights poll results to be representative, and the limitations of an online, voluntary sample.

Last updated 5 October 2026

Sampling

Responses are collected via the Civic Pulse app. We use quota sampling to ensure representation across age, sex, and state. Adjusted results apply post-stratification weighting to match ABS population benchmarks.

Raw vs adjusted

Raw results show unweighted responses from user participation only. Adjusted results apply ABS/Census demographic weighting to correct for over- or under-representation of certain groups.

Limitations

This is an opt-in panel, not a probability sample. Results reflect the views of app users and may not perfectly represent all Australians. Margin of error varies by question and sample size.

Confidence intervals

For national results with 12,000+ responses, the margin of error is approximately +/-1% at 95% confidence. Subgroup results have larger margins of error.