The reason testimony beats aggregate evidence

A table shows that a programme helped ten thousand people. A single beneficiary tells the story of how it helped her. The story raises more donations than the table, and the story is one data point from the table.

What everyone sees

The preference for the story looks like innumeracy or sentimentality: people who understood statistics would rely on the aggregate. The possibility that the story carries different information — not more or less, but a different kind that the table cannot deliver — is less commonly discussed.

What is actually happening

Slovic described psychic numbing: as the number of victims increases, empathy per victim declines. Small, Loewenstein and Slovic showed that an identifiable victim elicits more aid than a statistical one, and that combining the two actually reduces the effect of the individual. Haidt argued that moral reasoning often follows moral intuition, which is triggered by specificity and narrative rather than by aggregation. The story is not a worse summary of the same data; it is a different instrument that activates an emotional processing path the table cannot reach.

Why it stays hidden

The dominance hides because people who respond to the story believe they are responding to the facts, and people who present the table believe they are being rigorous. Neither is wrong: the facts in the story are accurate, and the table is rigorous. The difference is in which processing path each instrument activates, and the emotional path produces more action than the analytical one, which is a finding about the audience rather than about the data.