Why multiple-choice vaping survey percentages can exceed one hundred

Read the question before judging the total

A chart of reasons for vaping may summarize a choose-one question, a select-all-that-apply question, or separately rated statements. These create different summaries. Locate the wording, response instruction, base population, and footnote. Do not infer the instruction from a chart’s shape. A pie-style appearance does not make categories exclusive when the underlying question allowed several choices.

Health Canada’s original vaping-panel report explicitly notes that percentages may not add to one hundred because of multiple responses or rounding. We use that reporting warning, not its substantive findings. These are distinct explanations: rounding affects displayed precision, while multiple choices can make categories overlap. Checking the actual question tells you which explanation is relevant.

Health Canada / Environics Research: Vaping Panel Surveys and Interviews 2022-2023 — Final Report

Separate people from selections

For an invented example, ten respondents may each choose any applicable reason. Seven select habit and six select social situation. The bars are seventy and sixty percent of respondents, adding to one hundred thirty percent. There are still ten respondents. You cannot say thirteen people answered or that seventy percent chose habit alone, because some may also have selected social situation.

If a different display divides by all selections instead, its percentages answer another question: how the selection count is distributed. That denominator can make the bars total one hundred while no longer showing the fraction of people choosing each option. Copy the denominator stated by the report. The fictional ten-person example illustrates counting only and is not a vaping poll or prevalence statistic.

Do not invent the overlap

To know how many chose either option, you need the count choosing both or an appropriate combined table. Two marginal bars alone do not give a unique answer. Also avoid assuming that each respondent selected the same number of options. A report may provide a net any-category measure; keep its definition rather than calculating a substitute by summing overlapping bars.

If the question or denominator is missing, ask for them or leave the total uninterpreted. A large sum is not by itself proof of an invalid survey, and a neat one hundred percent is not proof of good sampling. Recruitment, missing answers, and measurement definitions remain separate issues. Your task is to understand the specific chart without turning a formatting impression into a methods verdict.

Keep the instruction in your summary

A usable sentence says these are percentages of the stated respondents who selected each option, and respondents could select more than one. If you compare charts, check whether both used the same instruction. A choose-one question about the main reason and a multiple-select question about all reasons cannot be treated as identical measures just because the category names match.

This reading aid does not identify an individual’s cause for vaping or show what a quitting program should target. Responses describe the survey’s question under its actual design. For your own plan, a short context note can remain useful without mirroring a public poll’s categories. Finish by preserving the denominator, selection rule, and unknown overlap rather than producing a stronger total than the published data allow.