What are 3-day averages?
We publish data in daily, 3-day, weekly and monthly ‘cuts’. But we strongly draw your attention to the different levels of reliability associated with the different sample sizes. One-day figures represent only 125 responses to a particular measure (buzz, quality, value etc). The INDEX score comprises 6 measures, so the one-day sample size for that is 750. If one looks at the one-day score on one measure of, say, women only, then the sample size is around 60. That is why we strongly recommend you only consider the INDEX score on the daily cuts, and use the weekly averages for individual measures and demographic breaks. Indicative margins of error for the different sample sizes are given below: • 315 (demographic break on one-week single-measure score): +/- 5.5% • 625 (one-week single-measure score): +/- 4.1% • 325 (demographic break on daily INDEX score or one-week single-measure score) = +/- 5.4 • 1,500 (demographic break on monthly single-measure score): +/- 2.
