A vaping survey reports one subgroup. Which records inform the uncertainty?
Read how the ITC study distinguishes a subgroup estimate from the data needed to calculate its standard error.
No Vape · Updated

Find the group and the file separately
The International Tobacco Control study team describes two approaches in a technical note for Wave 1 of its Four Country Smoking and Vaping Survey. The first physically keeps only the records meeting specified conditions. The second indicates which respondents belong to the subgroup while retaining all sampled respondents for the domain analysis. The note explains why that distinction can matter for variance.
Write down the exact inclusion rule used for the subgroup estimate. The file name may refer to a shared set of wave identifiers rather than the specific analytical subset. The ITC note explicitly distinguishes those uses of “core.” Match the documentation to the analysis being described.
Picture a subgroup in only part of a sample
Imagine an invented survey that sampled households in several areas using groups of addresses. A report wants the answer for adults who had used a service. Some sampled groups of addresses happen to contain no eligible service users. Those groups provide no answers to the target proportion. Their place in the wider sample design can still matter to the uncertainty calculation.
In this example a list of only service users could lose some design structure. A domain indicator lets an analysis identify who contributes to the subgroup estimate while retaining the information needed by the stated variance method. The example is a diagram in words, without actual survey results or a formula for your software.
Read the standard-error footnote
Look for a statement about domain analysis, weighting, strata and the actual variance procedure. Keep the point estimate and standard error in separate columns. A result can give a narrow description of the subgroup and still leave the calculation of uncertainty insufficiently explained.
The ITC note says the difference between its approaches can be negligible when most respondents are included and all strata are represented. It recommends the domain approach for smaller subpopulations in its described circumstances. It also notes an exception when the relevant group already has a separate survey weight. Retain those qualifications when reporting the guidance.
Describe who contributed to which question
A concise summary can name the subgroup estimate, the wider design retained for uncertainty and the procedure authors reported. If a paper omits the method, mark the standard-error calculation as unresolved. The existence of a subgroup table alone gives too little information to tell how it was produced.
Questions about this step
Does retaining the wider survey file include everyone in the subgroup percentage?
The domain rule still identifies which records contribute to the subgroup statistic. The wider file supplies design information for the stated analysis.
Sources and scope
Sources inform the guidance. The worksheets and examples are No Vape editorial suggestions, not tested treatments.
Written with AI assistance and checked against the cited sources. No clinical review is claimed. This guide covers practical planning, not individual medical treatment. How we research and write