Hello, I’m Pang Ying.

Have you received advice like this? My mother tells me to practice delayed gratification because experiments prove children who can delay gratification become more successful.

A friend showed me a photograph from ten years ago and said I looked so thin and ten years younger, so I should lose weight. I understand their concern, but the advice does not persuade me because it confuses correlation with causation.

Why begin the series here? Confusing correlation with causation is among the most common logical errors in life and work, and distinguishing them is central to science: smoking and lung cancer, low estrogen and heart disease, political systems and economic development, or whether quarterly sales rose because of advertising or coincidence.

Begin with a simple real example.

A severe polio epidemic affected many children in the early twentieth century. Ignorant of its cause, scientists analyzed data and identified what they thought was the culprit: ice cream.

Ice-cream sales and polio cases rose and fell along almost identical curves. Some blamed cold temperature and others sugar, frightening parents into prohibiting ice cream.

The conclusion was wrong. Both increased in summer: children ate more ice cream and encountered the virus more often. Their timing correlated, but ice cream did not cause polio, and banning it could not prevent disease.

Correlation says A and B occur together, not that A causes B. If they are only correlated, changing A cannot reliably change B; restricting ice cream cannot solve polio.

Correlation can nonetheless predict. Rising ice-cream sales might forecast rising polio cases and the need to prepare medical facilities.

You have probably heard of the Stanford marshmallow experiment on delayed gratification.

Children were promised a second marshmallow if they waited fifteen minutes before eating the first. Those who waited appeared more successful later, leading to the claim that delayed gratification caused success.