Archive for January, 2020


Correlation≠ causation

January 16, 2020

My wife shared this photo with me today. It is part of an opinion story that appears in the Orange County Register and presents a good lesson for editors about correlation vs. causation.

What’s wrong with it? Well, note the sentence in the cutline:

The county says that people who receive county services are most likely to become homeless.”

It is probably an accurate statement in that someone from the county –though I would like to know who or which office– said. But it certainly is misleading. The average reader would look at this and say that maybe the county should not be offering services or that the county is somehow culpable in creating homelessness. The fact is not that those who receive services become homeless, but that those on the road to homelessness often end up seeking services from the county. Good editors would catch the correlation vs. causation error. Weak editors would not pay attention to what the words REALLY say.

Oh, and by the way, the opinion story does not even make a direct connection with the cutline, but it does support the misconception.