Blending

There is a substantial body of international research and now some Australian research which demonstrates that the prevalence estimates obtained from non-probability opt-in online panels can be very inaccurate, relative to benchmarks, and very variable compared to well-designed probability surveys. To this end, the Social Research Centre is able to offer a sample blending approach whereby interviews obtained from in-scope members from our probability-based panel (Life in Australia™) are combined with the non-probability interviews obtained from a commercial nonprobability panel provider. As an example, the Social Research Survey was able to produce high quality probability-based estimates of the prevalence of image-based abuse for women aged 15-45 using online data collection.