Open Access
Future NRS
Data Smoothing
NRS’ own research amongst subscribers suggests that the volatility in the Survey’s estimates between one period and another is some cause for concern. This is especially so in circumstances where the direction of change in the NRS estimate is opposite to the direction of any change in the circulation data. The fluctuation in the NRS estimates between periods is nothing unusual, and is the inevitable consequence of sampling variation, something that is true of all sample surveys, even one as robust as the NRS.
It would be possible to minimise such differences by linking any short-term changes in readership estimates to circulation data, while maintaining long-term real changes in readers-per-copy. Such an adjustment procedure is being investigated and tested in 2007, and if approved, could be applied to the survey’s published estimates in 2008.