Skip to content
No Vape

Michigan’s 2018 BRFSS vaping module used two weights

Follow the CDC example for Michigan’s split e-cigarette questionnaire before accepting a pooled 2018 estimate.

No Vape · Updated

Disposable vape resting on concrete
Illustrative image · Markus Winkler / Unsplash

Find the question in the right files

In the CDC’s guide to preparing 2018 BRFSS module data, look for Example 2. It identifies both Michigan questionnaires. The state’s records for this particular module come from the version 2 and version 3 datasets, LLCP18V2 and LLCP18V3, using their associated weights. You can trace that instruction in the guide’s worked code. A table heading that only mentions Michigan gives you the location. The code tells you which files represent the respondents answering this module.

The guide’s treatment of other states helps explain the reason. It lists places that asked the module as a common version and others that used only one split version. A single shortcut such as “always take the main file” misses those choices.

Centers for Disease Control and Prevention: Complex Sampling Weights and Preparing 2018 BRFSS Module Data for Analysis

Why the example halves both weights

The reason is proportional allocation. The CDC example compares the numbers of observations available from the two Michigan versions. Since the counts are close, it uses half of each original version-specific weight. If the sample sizes differ materially, the guide shows multiplying each version’s weight by that version’s share of the combined sample.

Here is a fictional illustration of the second case. Imagine a module appearing in two files with 1,000 observations in one and 3,000 in the other. Those sizes make the shares one quarter and three quarters. The example gives you a way to interpret a methods note, using rounded shares for illustration. It is unrelated to how many participants in either file reported vaping. It also does not supply the full statistical procedure for calculating uncertainty.

Centers for Disease Control and Prevention: Complex Sampling Weights and Preparing 2018 BRFSS Module Data for Analysis

What to record beside a published estimate

Note the survey year, state, module, questionnaire files and what the authors did to scale their weights before combining them. If the paper names a different year or pools several states, look up each relevant year and state in the corresponding CDC documentation. Michigan’s 2018 instructions establish only this worked example.

If you can verify the variable label but cannot tell which questionnaires entered the estimate, you can still describe the label accurately. Leave the basis of the estimate open until the methods or code show it. That distinction is especially useful when two reports claim to use the same survey but reach different numbers.

Sources and scope

Sources inform the guidance. The worksheets and examples are No Vape editorial suggestions, not tested treatments.

  1. Centers for Disease Control and Prevention — Complex Sampling Weights and Preparing 2018 BRFSS Module Data for Analysis
  2. No Vape — No Vape — app features and store links

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