The data on smoking habits in the UK come from the Annual Population Survey (APS). The data on smoking is collected on the Labour Force Survey, which forms a component of the APS. In 2017, there were 158,889 survey respondents to the question on smoking habits. Interviews are carried out either on a face-to-face basis or on the telephone.
Surveys collect information from a random sample of the target population to make generalisations (reach 'findings’) about everyone within that population.
For those findings to be reliable, the sample of people should ideally contain the same mix of age, gender and regional location as the target population.
Where this isn’t the case (because some people haven’t responded, for example) analysts use statistical tools to ‘weight’ the data. Weighting rebalances the survey responses so they represent the target population more accurately. They can then be used to reach meaningful conclusions.
The APS datasets are weighted to reflect the size and composition of the general population, by using the most up-to-date official population data. Weighting factors take account of the design of the survey (which does not include communal establishments) and the composition of the local population by age and gender. The weights for other sample members are then adjusted to compensate for this.
Confidence intervals for each ethnic group are available if you download the data.
The APS data is based on the responses of a sample of adults in England rather than all adults in England. This measure makes a reliable estimate of the percentage of adults in England who were current smokers at the time of the survey, but it’s impossible to be 100% certain of the true percentage.
Based on APS data, it’s estimated that 14.9% of adults were current smokers in England in 2017.
It’s 95% certain, however, that somewhere between 14.6% and 15.1% of all adults in England were current smokers in 2017. In statistical terms, this is a 95% confidence interval. This means that if 100 random samples were taken, then 95 times out of 100 the estimate would fall between the upper and lower confidence interval. 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 adults from the Chinese ethnic group responded to the survey than White adults, so we can be less certain about the estimate for the smaller group. This greater uncertainty is expressed by a wider confidence interval, for example of between 6.4% and 10.9% for Chinese adults in 2017.
Observed differences are considered statistically significant when the 95% confidence intervals for an ethnic group don't overlap with those of the reference group.
The ‘linearised-Jacknife’ method for calculating confidence intervals has been used. Previously the normal approximation method was used, however this was not able to take into account the design of the survey and how this can affect the precision of the estimates. Data from 2012 therefore has been revised using the new method. For further details of how the calculation is carried out please see the ONS website. For further details of the sampling method and weighting see the APS quality and methodology information.
Estimates are rounded to 1 decimal point in the charts and tables. You can see the unrounded figures if you download the data.