If the estimated readership of any title falls below a certain level the estimate becomes unreliable and must be withdrawn from publication.
This level is defined as the lower limit of the 95% confidence interval of a theoretical all-adult readership estimate of 0.3% of the adult population.
For example, on a total sample of 36,000, with an estimated adult population of 49 million, an estimated adult readership of 109,000 (0.224%) would fall below this level.
If the estimated readership of any title falls below this level for 2 consecutive 24-month periods ending in June, the last estimate released will be for the period ending that December.
The intention of this rule is to avoid any title being removed from the survey if a shortfall against the required minimum could be the result only of sample variation, rather than because of any real diminution of readership below 0.3% of all adults.