Something that really bugs me is how the news will announce the findings from some research study but fail to draw the proper conclusions.
EXAMPLE: I once read an article about research that found a correlation between children who have nightlights and children who wore glasses. The article immediately jumped to the conclusion that having a nightlight causes vision problems in children — but that causation was not found in the research!
A correlation between A and B just means that they seem to be associated but more research is needed to determine causation. There are three relational possibilities:
- A causes B.
- B causes A.
- A and B are both caused by a ‘confounding variable’ C that was not originally considered.
In the example above, these possibilities would be:
- Maybe having a nightlight causes vision problems in children. The vision problems could be caused by eye strain from kids trying to read in low light or perhaps the light in the bedroom prevents their eyes from properly relaxing during sleep.
- Maybe wearing glasses causes children to want a nightlight. This seems perfectly reasonable: Since kids don’t wear their glasses while they sleep, they could want a nightlight to help them navigate their way to the bathroom with blurry vision.
- Or, perhaps, the glasses and nightlight are both caused by something else: over-protective parents. Doting parents may worry so much about their kids that they provide a nightlight for safety and take them to the optometrist ‘just in case.’
Of course, there is always one more possibility:
- There is no causal link between A and B — the correlation was just a coincidence.