What is the risk of ecological fallacy in cohort studies using group-level data?

• A. Inference about individuals based on group-level data can be misleading
• B. It provides precise individual-level estimates
• C. It eliminates confounding completely
• D. It applies only to case-control studies

Answer: (A)

Ecological fallacy shows up when you try to draw individual-level conclusions from group-level data. In cohort studies that use data summarized by groups (like regions or communities), the average exposure and outcome of a group don’t dictate what happens to each person inside that group. People within the same group can vary a lot, and other factors that influence risk may differ within groups even if the group averages look the same. So a pattern seen across groups can be misleading about individuals—an association at the group level might not apply to individuals, or the direction of the association could even be different.

That’s why the option stating that inference about individuals based on group-level data can be misleading is the best choice. It also isn’t true that these designs give precise individual-level estimates, nor that they fully eliminate confounding. And ecological concerns aren’t limited to case-control studies; any study using aggregate data can fall prey to this bias.