Why a coffee statistic is not a vote against tea

Suppose a headline tells you that most people drink coffee. It is tempting to imagine the rest drinking tea, as if breakfast came with only two marked boxes. But the original question might never have mentioned tea.

Read the question before the conclusion

In July 2015, Gallup interviewed 1,009 U.S. adults. It reported that 64% drank at least one cup of coffee on an average day. That figure describes reported coffee consumption in that survey, not current global consumption and not a coffee-versus-tea election. Read Gallup’s report.

A coffee drinker may also enjoy tea. Someone who drinks neither is not automatically on the tea side. And a person may prefer a drink they have only occasionally. Frequency, preference and exclusivity are separate ideas.

Three versions of the same conversation

Compare these questions: What do you drink most mornings? What would you order during a long afternoon conversation? If you had to choose one today, which would it be?

Even one person could answer differently each time. The changing answer would not make the person unreliable; the question changed the task.

This is why a number should travel with its wording, date and population. Lose those details and a modest observation can become a claim it never supported.

A useful habit before you share a statistic

Look for three things: who answered, when they answered, and what they were asked. If one is missing, pause before building a story around the percentage.

A World Chooses result answers our displayed either/or question among its participants. It should not be substituted for a consumption survey, and neither should that survey replace a preference result. Both can be interesting when we let each answer its own question.