Two vaping measures correlate: that does not mean their values agree

Identify the measurement claim you are reading

A report might compare two ways of recording the same quantity or relate two different quantities. Those questions differ. Find whether both methods use the same unit, cover the same period, and concern the same paired observations. Similar labels such as use or intake do not establish that the measures are identical.

The original Bland and Altman methodological paper warns against using correlation as a substitute for agreement in measurement comparison. The abstract supports that narrow distinction. This article does not reproduce its full analysis, set an acceptable measurement error, or validate any vaping device, service, or No Vape function.

Bland and Altman — The Lancet; abstract indexed by PubMed: Statistical methods for assessing agreement between two methods of clinical measurement (1986)

See how fictional pairs can move together but differ

Imagine one invented method giving values 1, 2, and 3 while another gives 11, 12, and 13 for the same three observations. The values rise together in a perfectly linear pattern, yet the second is ten units higher each time. That relationship does not make the values numerically identical.

These are toy arithmetic values, not actual puff measurements, nicotine amounts, or a study result. The example shows why a strong relationship can coexist with a systematic difference. It does not establish which method is accurate; that question would require an appropriate reference, design, and assessment for the intended use.

Look for an assessment matched to interchangeability

If the claim is that one method can replace another, look for how the authors evaluated differences and their practical acceptability. Retain the population, range of observations, missing-pair rules, and intended use. A method performing one way in a limited sample need not behave the same way in a different setting.

Do not calculate a correlation from your own few notes and call the result a validation study. Nor should you demand private participant data to resolve a public reading question. Use the reported methods and seek qualified clarification when the assessment is too technical or absent from the summary you have read.

Keep measurement and benefit as separate claims

Even justified agreement for a quantity would not show that using the tool helps people quit, establishes nicotine absorbed, or guides individual treatment. Those claims require their own evidence. A sensor-comparison paper cannot lend a clinical outcome to a tracking app simply because both concern vaping records.

A useful reading note names the paired methods, unit, relationship statistic, actual agreement assessment, and remaining limitation. If the report gives correlation alone, say that agreement has not been established by that statistic. This guide cannot rule on a named technology; it helps you repeat the measurement claim without silently expanding its meaning.