A sample of approximately 300,000 young people were selected from the National Pupil Database (NPD) and contacted to take part in the postal survey, with the option to complete it online. Fieldwork ran between 22 September 2014 and 9 January 2015. A total of 120,115 out of 295,024 participants responded to the survey. In a minority of cases (1% of pupils selected for the 2014 survey), parents opted to take their children out of the survey. A £5 shopping voucher was used as a token of appreciation, which was conditional on completing the questionnaire. The response rate was 41%, when taking into account any undeliverable mail outs and opt-outs from the sampled pupils.
By using the NPD as a sampling frame, it was possible to stratify the sample to help ensure the sample was representative across a range of pupil and area characteristics. Data was weighted by gender, ethnicity, Free School Meal eligibility, quintile of the Index of Multiple Deprivation, and local authority to reflect the known population profile of 15 year olds in England (sourced from the NPD).
Both the numerator and the denominator were weighted to make them applicable to the population as a whole. Further details on the sampling methodology and weighting procedure can be found in the WAY survey technical report
Confidence intervals for each ethnic group are available in Download the data.
Based on survey responses, it’s estimated that 8.2% of all 15 year olds in England in 2014/15 were smokers. Because the WAY survey questioned a sample of 15-year-old smokers rather than all 15-year-old smokers, however, it’s impossible to be 100% certain of the exact percentage.
It’s 95% certain, however, that between 8.1% and 8.3% of all 15 year olds were smokers. In statistical terms, this is a 95% confidence interval, with a lower and upper confidence interval of 8.1% and 8.3%, respectively. This means that if 100 random samples were taken, then 95 times out of 100 the estimate would fall between 8.1% and 8.3%. But 5 times out of 100 it would fall outside this range.
The smaller the survey sample, the more uncertain the estimate and the wider the confidence interval. For example, fewer children from the Mixed ethnic group responded to the survey than White children, so we can be less certain about the estimate for the smaller group. This greater uncertainty is expressed by a wider confidence interval of between 8.2% and 9.8% for Mixed ethnic children.