Attribution: behavior under choice and constraint
Distinguish an observed action from a dispositional inference by examining choice, incentives and situational constraints.
Question and prediction
Does defending a position in an assigned debate establish the speaker’s private belief?
Predict how knowing that a position was assigned should affect confidence in an attitude inference.
Mechanism and evidence
Attribution is the process of explaining behavior. Dispositional explanations emphasize a person’s characteristics; situational explanations emphasize constraints and context. Correspondence bias refers to inferring a matching disposition from behavior despite situational pressures. The fundamental attribution error is a related tendency to underweight context when explaining others. These are research concepts, not proof that every observer always makes the same error.
Jones and Harris’s 1967 experiments examined inferred attitudes after participants encountered expressed positions, including assigned positions. The authors reported attitude inferences aligned with the expressed position even under no-choice conditions. That bounded finding concerns a particular design and sample; it does not establish that observers never use situational information or that all cultures show identical patterns.
Separate three variables: the person’s actual attitude, the position expressed, and the observer’s estimate. An experiment can measure the last two without knowing the first. In the fictional debate, equally strong speeches are either assigned or chosen. Inferring belief from assignment alone treats behavior as more diagnostic than the design warrants. A chosen position may be more informative, but incentives and audience still matter.
To evaluate a dispositional explanation, ask whether behavior recurs across relevant situations and whether others respond similarly under the same constraint. Consensus, distinctiveness and consistency can guide observation. They are not a magic formula that reveals intent. A careful conclusion calibrates uncertainty and proposes evidence that could distinguish choice from compliance.
Inspect the relationships
| Variable | Observed here? | Appropriate claim |
|---|
| Assigned/chosen position | Yes | Describes constraint |
| Speech content | Yes | Describes expressed position |
| Private attitude | No | Uncertain inference |
| Observer estimate | Yes | Measures attribution |
Worked example
A’s speech establishes that A delivered a defense under assignment. B’s choice supplies additional information about preference but not certainty. Equal observer ratings do not prove equal actual attitudes; they may reveal insufficient adjustment for the constraint.
Your investigation
Draw separate boxes for actual attitude, expressed position and observer estimate. Use the Jones–Harris evidence card to identify which relation was investigated. Write a 200-word correction of ‘Both speeches prove sincere support,’ then propose a low-pressure observation that would change your confidence.
Unfamiliar transfer
A quiet employee speaks readily when given equal turns. What does this suggest examining? Explain your reasoning before checking feedback.
Reading
Jones and Harris (1967): The attribution of attitudes. The records below are original fictional teaching cases, not research results from this source.
Fictional debate conditions Speaker A is assigned to defend policy X. Speaker B chooses to defend X. Both deliver equally strong arguments. Neither privately reports an attitude.
Fictional observer ratings Observers estimate both speakers strongly favor X. The scale measures observers’ beliefs about speakers, not speakers’ actual beliefs.
Authentic evidence and its limits
Published result: attitudes under constraint — Edward E. Jones and Victor A. Harris (1967). Across their experiments, the authors reported a tendency to infer attitudes corresponding to an expressed position even when that position was assigned.
Paraphrase of the original article’s reported finding. The design measured observers’ estimates; it does not make each estimate a verified private attitude. The fictional debate ratings are not original data.
Read the attributed source. Original LearnAI paraphrase of the attributed source; no reproduced figure, participant data or verbatim passage. Original publication retains its rights.
