Key figure
Weighting removed about 30% of bias in Pew's opt-in test
A stat card from Pew Research Center's benchmark study of online opt-in samples. Across 24 benchmarks measured against high-quality government data, unweighted samples were off by 8.4 percentage points on average, and the most effective adjustment brought that to about 6, removing roughly 30% of the bias. It shows why weighting cannot stand in for representative recruiting, and why the choice of weighting variables matters more than the method.
Where this appears
What's in this key figure
30% of bias removed by the most effective weighting of online opt-in samplesAverage bias across 24 benchmarks fell from 8.4 points to about 6. Seventy percent survived.
Use this image
Free to use in your docs, decks and posts under CC BY 4.0, as long as you credit Koji with a link to this page. To embed it, paste this into your page:
<a href="https://www.koji.so/visuals/pew-weighting-removes-30-percent-of-bias"><img src="https://www.koji.so/visuals/pew-weighting-removes-30-percent-of-bias/pew-weighting-removes-30-percent-of-bias.png" alt="Stat card: the best weighting strategy removed roughly 30% of bias in online opt-in samples, from 8.4 points to about 6, in a Pew Research Center study." width="1200" height="641" /></a>
<p>Source: <a href="https://www.koji.so/visuals/pew-weighting-removes-30-percent-of-bias">Weighting removed about 30% of bias in Pew's opt-in test</a>, Koji</p>
