An association describes a pattern
The NIH guide to biomedical definitions distinguishes association or correlation from a causal relationship. Finding that two variables change together is relevant information, but does not identify the mechanism connecting them on its own. Several explanations may fit the same pattern.
In our example, rain is a third variable that helps explain both umbrellas and puddles. In a real study, the factors might be age, existing conditions, habits or other characteristics. Identifying them requires knowledge of the subject and an appropriate design. Imagining an alternative explanation does not establish it, but it does help us recognise that the first story is not the only possible one.
The order of events matters too
To interpret a relationship, it helps to know what happened first. A measurement taken at a single point in time can show coexistence without clarifying direction. Sometimes a condition changes people’s behaviour; sometimes behaviour contributes to an outcome. Confusing those directions produces headlines that are too certain.
Imagine a fictional survey in which people who read more articles about sleep report sleeping worse. Concern about their sleep might be driving them to seek information. Concluding that reading those articles worsens sleep would skip an obvious possibility. The example establishes nothing about health; it illustrates a problem of interpretation.
Adjustment helps, but does not make everything certain
Analyses can account for known variables. That work matters, although it does not guarantee that every relevant difference has been measured. When reading an observational study, look for the factors considered and the limitations acknowledged by the authors. A list of adjustments should not be treated as an automatic eraser of every bias.
Nor should we assume that knowledge exists only when a randomised experiment is possible. Many questions require other designs. Quality depends on how methods, timing and evidence are brought together. What we should avoid is assigning a causal conclusion to a single statistical pattern when its design cannot support that conclusion on its own.
A headline can preserve the finding
“A study observes an association” can be engaging if it explains which variables, in whom and why the relationship matters. It does not need an added causal verb to deserve attention. Good science communication makes the open question visible rather than hiding it behind a more emphatic sentence.
Before sharing a cannabis news story, compare the verb in the headline with the one in the study. If the paper discusses association and the article says cause, look for the justification for that change. Checking a single word can prevent a large misunderstanding. Careful reading sometimes begins with paying attention to verbs.
Take this with you
Two related variables do not establish a cause on their own. Retain the design, timing and verb that match the evidence.
Sources you can open
The source may be in its original language. English publication does not change the jurisdiction or date discussed.
Follow the thread
Open the cannabinoid encyclopedia
Check the identity, research, uncertainties and documented forms of each compound.