Media, science & critical thinking
A story has a face, a beginning and a consequence. Statistics feel colder, which is why one individual fate can outweigh thousands of invisible cases. The answer, however, is not to ban stories. We need to identify precisely which question they can answer and which they cannot.
1. What the evidence shows
Case reports and case series have no control group and generally cannot reliably estimate frequency or treatment effect. They can, however, draw attention to an unusual or novel phenomenon and generate a hypothesis for further research.[1]
Hierarchies of evidence differ according to the question. A causal effect of an intervention requires comparison, while qualitative research can systematically examine experience, meaning and process that a simple average fails to capture.[2]
Research into identifiable victims shows that a specific person can elicit a different response from a statistical description. The psychological effect of a story does not itself confirm that the story is representative.[3]
2. How to read the claim in context
The error occurs when the scale changes: “this happened to this person” becomes “this is what happens to people in general” without an intermediate step. To make that move, we need a denominator, a comparison group and control for other explanations.
A story can legitimately show that a system-wide average overlooks someone, or describe a mechanism worth testing. Its strength is exploratory and explanatory, not automatically a measure of frequency.
Good journalism therefore combines human experience with data and openly states how the case was selected. Readers can then appreciate the significance of an event without mistaking one person for a population.
- An individual case establishes that the event occurred.
- Without comparison, it generally determines neither frequency nor cause.
- Case reports and qualitative research have legitimate but distinct roles.
- How representative a story is without knowing the sampling and population.
- A causal mechanism based solely on the sequence of events in one case.
- How strongly emotion distorted a particular reader’s judgment.
3. Five questions to ask
- What precise question is the story meant to answer?
- How was the case selected?
- Do we know the number of similar and contrasting cases?
- Is there a comparison group or an alternative explanation?
- Is the story an illustration of data, a source of a hypothesis, or being presented as general evidence?
4. Conclusion
A story is not the opposite of evidence. It is a particular kind of information. It can confirm that an experience exists and open a question; frequency and cause generally need to be established by broader and more systematic data.
— Jiný Kontext