Statistical Reliability

The sampling tolerances that apply to the percentage results in this report are given below. This tableshows the possible variation that might be anticipated because a sample, rather than the entire population,was interviewed. As indicated, sampling tolerances vary with the size of the sample and the size of percentageresults. The confidence intervals take no account of design effects and, of course, there were certain areasin each region we did not conduct interviews.

Source: MORI

For example, for a question where 50% of the people in a sample of 850 respond with a particular answer,the chances are 95 in 100 that this result would not vary more than 3 percentage points, plus or minus, fromthe result that would have been obtained from a census of the entire population using the same procedures.Tolerances are also involved in the comparison of results between different elements of the sample. Adifference, in other words, must be of at least a certain size to be statistically significant. The followingtable is a guide to the sampling tolerances applicable to comparisons.