Advocate: A study of people who had recently recovered from colds found that people who took cold medicine for their colds reported more severe symptoms than those people who did not take cold medicine.
Conclusion
Taking cold medicine is clearly counterproductive.
Evidence
In a study, there was a correlation between people who took cold medication and those who had more severe symptoms. (people who were X were more Y than those who weren't)
Evaluate
When authors see correlations, they over-eagerly jump to the conclusion that one thing causes the other.
In this correlation, took cold meds || more severe symptoms
which way does this author see causality flowing? took cold meds → more severe symptoms
That's why she's saying that taking cold meds is counterproductive (made the cold even worse).
When we see authors pondering Curious Facts and deciding on their Explanation, we think: - could there be some OTHER WAY to explain this curious fact? - how plausible is the AUTHOR'S WAY?
When it comes to correlations, we immediately think of two alternate ways to explain them:
curious fact: took cold meds || more severe symptoms
author: took cold meds → more severe symptoms
us:
took cold meds ← more severe symptoms? (reverse causality)
took cold meds || more severe symptoms / (third factor?)
In this case, reverse causality makes a lot of common sense: if you had mild symptoms, you probably wouldn't be as likely to bother taking cold meds, whereas if you had severe symptoms, you're more likely to seek medicinal relief.
Goal
Look for an answer pointing out the more likely interpretation of that correlation (that the severe symptoms came first, and THAT'S why those people took cold meds), or an answer saying there could be some Third Factor accounting for both (maybe older people are more likely to take meds and more likely to experience bad symptoms), or an answer just calling out the author's illegal inference from a correlation that one of the things is causing the other.