Reading focus
Syllabus: How to Read Psychological Science
Connect the experimental design, observed helping responses, and proposed explanation. Ask what the manipulation establishes and what remains an interpretation.
Why you’re reading it. Apply the distinction between a manipulated condition, an observed response, and an inferred mechanism to an experiment about helping.
A question to consider. Which explanation does varying the number of supposed bystanders help rule out, and what additional evidence would establish diffusion of responsibility as the mechanism?
Why this reading
Darley and Latané turn a question about helping into an experimental comparison: does believing that other people can also hear an emergency change whether, and how quickly, someone seeks help? The study separates an observed response from a proposed explanation in terms of shared responsibility. Its design makes both the value and the limits of a situational explanation visible.
Before you read
The study was prompted by the 1964 killing of Kitty Genovese in New York, reported at the time as having been witnessed by thirty-eight people who did nothing. Know that the reporting was later shown to be substantially inaccurate; the experiment's value does not depend on the anecdote.
Have in mind the two explanations the authors set against each other: that bystanders are apathetic, a claim about persons, and that responsibility is diffused among those present, a claim about situations.
Read for
- How the number of bystanders is manipulated without any bystander actually being present, and why that matters for the inference.
- The dependent measures. Notice that speed of response carries as much of the argument as whether people responded at all.
- The authors' attempt to relate personality measures to helping, and what they conclude from the failure to find relations.
- The mechanism proposed. Is diffusion of responsibility observed, or inferred from the pattern?
Consider afterwards
- The participants could see neither one another nor the victim. Which features of a real emergency does this remove, and how might those features affect helping?
- What would you need to compare the contribution of personality and situation beyond the particular manipulation and measures in this experiment?
Editorial context
The Fischer et al. meta-analysis examines how the bystander effect varies with emergency conditions, including danger. Philpot et al.'s CCTV study asks a different question: whether at least one person intervenes in a public conflict. Frequent help at the event level does not by itself contradict a reduction in an individual's probability of helping. Distinguish the outcome and unit of analysis before treating a later study as a replication or reversal.