Legal & DebateDUAL-USE

Narrative vs. Statistical Evidence

What it is

The strategic choice between proving a point with a story — one person, one case, told vividly — or with numbers, and the persuasive asymmetry between them: stories move belief through transportation and identification, statistics through base rates, and each can be used to hide what the other would show.

How it works

The folk wisdom that a story always beats a statistic is not what the evidence says. Allen and Preiss's 1997 meta-analysis found statistical evidence slightly more persuasive on average than narrative evidence, and Zebregs and colleagues' 2015 meta-analysis refined the picture: statistics do more for beliefs and attitudes, narratives do more for intentions. The two forms work by different routes. Green and Brock showed that narrative persuasion runs through transportation — absorption in a story reduces counterarguing and makes its claims feel experienced rather than asserted — and Small, Loewenstein, and Slovic showed that an identified individual draws far more sympathy and money than statistical victims, an effect that shrinks when people are prompted to deliberate. So the choice of form is strategic. A lawyer with a sympathetic client and bad base rates tells the story; one with good base rates and an unsympathetic client shows the chart. A public-health campaign leads with one child's face; an opponent of the same policy leads with a rare adverse event told as a story, because a single vivid case can neutralize a base rate in the listener's mind. The honest use matches the form to the question — stories for what something is like, statistics for how often — and offers both. The manipulation is choosing the form that hides the answer.

Real-world examples

  • •Since 1982, when Reagan seated Lenny Skutnik in the gallery, every State of the Union has used named guests to argue policy by story, a practice continued by presidents of both parties — an individual doing the work a table would do less well.
  • •Charity appeals: Small, Loewenstein, and Slovic's donors gave more to “Rokia,” a single named child, than to a description of millions facing famine; adding the statistics to the story lowered donations.
  • •Vaccine debates: one vividly told adverse event circulates further and persuades more than the adverse-event rate it is an instance of; the rate, offered alone, persuades less than it should.
  • •Workplace: a product manager arguing from “one customer told me” against a dashboard showing the opposite — or a dashboard used to dismiss the one customer who found the bug the metrics do not capture.
  • •Courtroom: the prosecution's narrative of a single night against the defense's statistics on eyewitness error, each side choosing the form the other cannot easily answer.

Ethical guidelines

Where the line is

Choosing a story or a statistic is legitimate when the form fits the question being answered and the other form is available on request — the story located in its base rate, the statistic illustrated by a real case; it becomes manipulation when the form is chosen because it hides what the other form would show, a vivid exception deployed to override a base rate or an aggregate used to bury a case the measure does not capture.

  • ●Offer both forms and let them check each other: the story for what it is like, the statistic for how often.
  • ●When you tell a story, say how typical it is. An anecdote presented without its base rate is a claim about frequency the speaker has not made.
  • ●When you give a statistic, be able to put a face on it honestly; if no real case fits the number, the number needs examining.
  • ●Do not use a vivid exception to neutralize a well-established base rate, and do not use a base rate to dismiss a case that reveals something the measure does not capture.

How to defend against it

  • ►After a story, ask the frequency question: “How often does this happen? Out of how many?” A story that cannot be located in a base rate is an emotion, not evidence about a policy.
  • ►After a statistic, ask the instance question: “What does a typical case in this number look like?” Aggregates can hide that the cases are not comparable.
  • ►Notice transportation in yourself: if you stopped counterarguing while the story was being told, that is the mechanism working, not the argument winning. Re-read it cold.
  • ►Deliberately request the other form. When someone gives you one, ask for the other; the honest advocate has it, and the manipulative one chose the form to avoid it.

From the Defense Playbook

Sympathy-versus-Law Self-Instructionminutes

When you have to apply a rule to a person whose story moves you, give yourself the instruction judges give juries: acknowledge the feeling, then decide each required element on the evidence, in writing, and check the result by imagining the same facts with a different person. (Educational; not legal advice.)

Every playbook entry states how strong its evidence is and when not to use it. Browse the full playbook.

References

  1. Allen, M., & Preiss, R. W. (1997). Comparing the Persuasiveness of Narrative and Statistical Evidence Using Meta-Analysis. Communication Research Reports, 14(2), 125-131
    Statistical evidence is slightly more persuasive on average than narrative evidence.
  2. Zebregs, S., van den Putte, B., Neijens, P., & de Graaf, A. (2015). The Differential Impact of Statistical and Narrative Evidence on Beliefs, Attitude, and Intention: A Meta-Analysis. Health Communication, 30(3), 282-289
    Statistics affect beliefs and attitudes more; narratives affect intentions more.
  3. Green, M. C., & Brock, T. C. (2000). The Role of Transportation in the Persuasiveness of Public Narratives. Journal of Personality and Social Psychology, 79(5), 701-721 · link
    Transportation into a narrative reduces counterarguing and increases story-consistent beliefs.
  4. Small, D. A., Loewenstein, G., & Slovic, P. (2007). Sympathy and Callousness: The Impact of Deliberative Thought on Donations to Identifiable and Statistical Victims. Organizational Behavior and Human Decision Processes, 102(2), 143-153
    The identifiable-victim effect and its reduction under deliberation, including the Rokia experiments.
Last reviewed
Suggest a correction

Detect Narrative vs. Statistical Evidence in any text

Paste any message, email, or article into our free Manipulation Detector to see if Narrative vs. Statistical Evidence or other techniques are being used on you.

Try the Detector FreeCreate Free Account