ANNEX 1 STATISTICAL RELIABILITY
As this survey involves only a sample of the adult population of Scotland, the results are subject to estimated sampling errors. The figures shown below allow for the impact of the weighting applied to the sample of 2600 interviews. The calculations are done on the basis of the 95% confidence interval, i.e. there is only a one in twenty chance that the 'true' result lies outside the stated range. Other than the figure for the total sample of 2600, the examples are taken from the percentages who were very satisfied with the service provided by this particular service or health professional.
Example | Sample size | Finding | Estimated sampling tolerance |
Total sample | 2600 | 50% | +/- 2.2% |
Questioned about seeing …own GP | 685 | Very satisfied with service provided
= 62% | +/- 4%
(i.e. a range of
58%-66%) |
…another GP | 165 | 59% | +/- 8% |
…the Practice Nurse | 859 | 65% | +/- 3.6% |
Out-of-hours service | 317 | 54% | +/- 6.4% |
Telephone consultation | 360 | 61% | +/- 5.6% |
Home visit | 246 | 69% | +/- 7.3% |
Outpatient appointment | 873 | 53% | +/- 3.8% |
Inpatient admission | 454 | 62% | +/- 5.4% |
The table below is a wider guide to confidence intervals, given different overall sample sizes. For example, for a sample size of 1000 interviews, where 50% give a particular answer, the chances are 19 in 20 that the "true" value (which would have been obtained if the whole population had been interviewed) will fall within the range of +/-3.5 percentage points from the sample result.
| Estimated sampling tolerances applicable to percentages at or near these levels |
10% or 90% | 30% or 70% | 50% |
Sample size |
200 | 4.5% | 6.9% | 7.5% |
400 | 3.3% | 5.0% | 5.4% |
600 | 2.6% | 4.0% | 4.3% |
1000 | 2.0% | 3.1% | 3.4% |