Psychologie sociale
Introduit| English | Sens |
|---|---|
| conformity/kənˈfɔːmɪti/ | conformité |
| fundamental attribution error/ˌfʌndəˈmentl ətrɪˈbjuːʃn ˈerə/ | erreur d'attribution fondamentale |
| cognitive dissonance/ˈkɒɡnɪtɪv ˈdɪsənəns/ | Dissonance cognitive |
Their character, my traffic?
- Imagine someone explains a classmate's lateness as “disorganised” but their own lateness as “traffic”. This illustrates a possible difference in explanation; it does not show that everyone makes this comparison or identify the actual cause of either delay.
- Social psychology examines attribution, conformity 从众, obedience, attitudes and group processes. Ask what people did, what reasons are reported and which comparison supports the explanation.
Attribution terms describe different questions
- A dispositional explanation concerns the person; a situational explanation concerns circumstances. The fundamental attribution error 基本归因错误 describes underweighting situational constraints when inferring others' dispositions. A self-serving pattern credits oneself for success while explaining failure externally.
- The actor–observer hypothesis proposes relatively more situational explanations for one's own actions and more personal explanations for others'. A large meta-analysis did not support a broad, context-free asymmetry. Do not treat this shorthand as a law or assume every personal explanation is mistaken.

Match each attribution label with its usual description, without treating it as a universal rule.
Difficulté 2/5 · 140 pièces + un petit bonus lors de la première réussite aujourd'hui
The labels concern related but distinct comparisons. The actor–observer shorthand has important empirical limits; a vignette illustrates a description rather than proving a general bias.
The actor–observer hypothesis is a context-free law that every person applies to all actions.
Difficulté 1/5 · 100 pièces + un petit bonus lors de la première réussite aujourd'hui
A broad unconditional asymmetry is not supported by the cited meta-analysis. Context, comparison and the kind of behaviour matter.
Kelley: compare people, targets and occasions
- In the covariation model, consensus concerns how others respond to the same target; distinctiveness concerns whether this person responds similarly to other targets; consistency concerns the response to this target across occasions. The three comparisons are not synonyms.
- The model predicts how patterns of information guide attribution. High consistency alone does not select a cause; examine the other comparisons and the evidence available. Covariation in observations is not equivalent to a controlled causal experiment.
Worked comparison. Many people laugh at a particular comedian (high consensus). One person laughs mainly at this comedian (high distinctiveness) and does so repeatedly (high consistency). Within the model, this pattern favours an attribution to the target rather than simply to a person who laughs at everything. It still does not exclude every alternative explanation.
Which comparison most directly concerns consensus in Kelley’s model?
Difficulté 1/5 · 100 pièces + un petit bonus lors de la première réussite aujourd'hui
Consensus compares people; distinctiveness compares targets for the person; consistency compares occasions for that person–target relation.
Separate acceptance, information and authority
- Normative influence concerns gaining acceptance or avoiding rejection; informational influence concerns treating others as evidence about what is correct. Following the same crowd could involve either motive, or both. Behaviour alone may not identify the motive.
- Obedience concerns responding to an authority's instruction. A direct instruction is different from matching peers, but the processes can overlap. State what the vignette says rather than inferring private belief, approval or motivation from compliance alone.
Worked distinction. A person copies an answer because they think classmates understand the problem: informational influence. A person privately thinks the answer is wrong but agrees to avoid rejection: normative influence. If only “copied the answer” is reported, the reason remains unknown.
Quelle influence est réellement décrite ?
Distinguez l'influence normative, l'influence informationnelle et l'obéissance.
In an unfamiliar place, a person follows others toward an exit because they believe the others know the correct route. The stated reason most directly illustrates
Difficulté 1/5 · 100 pièces + un petit bonus lors de la première réussite aujourd'hui
The reason explicitly treats others as information. Following a crowd without a stated reason would not by itself distinguish informational from normative influence.
Observing someone copy a group answer, without any other evidence, uniquely identifies a normative motive.
Difficulté 1/5 · 100 pièces + un petit bonus lors de la première réussite aujourd'hui
Acceptance, accuracy, authority or several processes could matter. The behaviour alone does not identify the motive.
Keep classic-study conditions and denominators
- Asch's 1955 report describes agreement with a misleading unanimous majority on 36.8% of the relevant line selections. This is a proportion of judgments under those conditions, not the percentage of people who always conform. A supporting dissenter reduced conformity; the effect depended on the variation.
- In Milgram's published 1963 laboratory condition, 26 of 40 male volunteers continued to the maximum 450-volt switch, giving 65%. The learner was a confederate and received no such shocks. This describes behaviour in that deceptive procedure, not a universal obedience rate or proof that participants approved of harm. Distress and deception also raise ethical questions.
In Asch’s reported supporting-partner variation, a dissenter increased conformity to the misleading majority.
Difficulté 1/5 · 100 pièces + un petit bonus lors de la première réussite aujourd'hui
A supporting partner reduced conformity in the reported variation. Different dissent arrangements and changes over trials should not be collapsed into one universal percentage.
What does the 36.8% result in the stated Asch report describe?
Difficulté 1/5 · 100 pièces + un petit bonus lors de la première réussite aujourd'hui
The unit is a judgment under the reported conditions, not a fixed type of person or a general human rate.
In the stated 1963 condition, 26 of 40 participants reached the maximum switch. What percentage is this?
Difficulté 2/5 · 140 pièces + un petit bonus lors de la première réussite aujourd'hui
26/40 × 100 = 65%. This is the participant proportion for that condition.
Dissonance: distinguish ratings from mechanism
- Cognitive dissonance 认知失调 is an aversive state associated with inconsistency among relevant beliefs or between belief and action, which can motivate attempts to reduce that inconsistency. In Festinger and Carlsmith's 1959 procedure, participants performed dull tasks; paid groups then described them positively to another person.
- The 1-dollar group's later enjoyment rating was higher than the 20-dollar group's. This fits an insufficient-external-justification account under the study conditions. The observed group difference is not a direct measurement of every participant's private process or proof that low rewards always change attitudes.
Means from the enjoyment question in Table 1, on a −5 to +5 scale. Each analysed condition had 20 participants. No uncertainty intervals are shown; the graph alone does not identify a unique psychological mechanism.
In Festinger and Carlsmith’s reported 1959 analysed conditions, which group gave the highest mean rating of task enjoyment?
Difficulté 1/5 · 100 pièces + un petit bonus lors de la première réussite aujourd'hui
The one-dollar group had the highest reported enjoyment mean. This supports the study’s predicted group pattern, not proof of every individual’s private mechanism.
Using the reported enjoyment means, 1.35 for the one-dollar group and −0.05 for the twenty-dollar group, what is the observed difference, first group minus second, in rating points?
Difficulté 2/5 · 140 pièces + un petit bonus lors de la première réussite aujourd'hui
1.35 − (−0.05) = 1.40 rating points. This group difference is not a direct measure of every participant’s private process.
Attitudes and behaviour need matching scope
- A broad attitude and one specific action need not correspond closely. Compare the target, action, context and time: a general view about the environment is different from a stated intention to cycle to work tomorrow.
- Strong, accessible attitudes can be useful predictors, but opportunity, habits, competing goals and social constraints matter. Observing one action does not reveal a stable attitude with certainty, and an observed association does not alone establish causal direction.
Worked limit. A person supports cleaner transport but drives because no suitable public route is available. The action does not by itself prove the attitude was insincere. Examine the specific choice and available alternatives.
Audience effects and collective effort differ
- Social facilitation accounts predict that others' presence can strengthen dominant responses, sometimes improving simple or well-practised performance and impairing difficult or unfamiliar performance. Task demands, evaluation and the setting matter; an audience does not guarantee improvement.
- Social loafing is reduced individual effort in a collective task relative to a suitable comparison. Identifiability, task value and whether one's contribution seems necessary can affect it. Lower group output can also reflect coordination problems, so output loss alone does not prove lower motivation.
Worked distinction. A practised typist becomes faster while watched: an audience-performance pattern. Several people work hard but duplicate each other's work and produce little: a possible coordination loss. Do not label every low-output group as loafing.
Doing a well-practised task better when watched is social ____.
Difficulté 1/5 · 100 pièces + un petit bonus lors de la première réussite aujourd'hui
The described improvement while watched fits social facilitation. Audience effects depend on the task, setting and response being measured; they are not always beneficial.
Low collective output proves that individuals reduced their effort.
Difficulté 1/5 · 100 pièces + un petit bonus lors de la première réussite aujourd'hui
Coordination loss, unequal resources or task differences can lower output. Social loafing requires evidence about individual effort relative to a suitable comparison.
Group decisions can shift in different ways
- Group polarisation is a shift toward a more extreme version of the group's initial average tendency. It can mean a stronger cautious position as well as a riskier one. It does not mean the group necessarily splits into two camps, nor that every discussion polarises.
- Groupthink concerns concurrence-seeking that interferes with critical evaluation, such as suppressing dissent or ignoring alternatives. Agreement alone is insufficient evidence of it. Examine how evidence, disagreement and alternatives were handled, not merely whether the final decision was unanimous.
Quelle description du processus de groupe correspond ?
Distinguez le processus mesuré et évitez les conclusions causales automatiques.
Helping depends on the outcome and setting
- Bystander research often finds a lower probability of intervention by a particular individual when other witnesses are present. Ambiguity, shared responsibility, danger, relationships and support from others can change the pattern. More witnesses does not guarantee that no one helps.
- Distinguish the probability that one witness helps from the probability that at least one person helps in the whole event. Counts of helpers, speed of response and effective assistance are also different outcomes. A claim about one denominator cannot automatically answer all these questions.
Worked evidence check. A report counts whether any witness intervened at each event. That can address event-level helping. It cannot, without individual-level information and a suitable comparison, establish each witness's probability of helping or prove one universal group-size effect.
A field report records whether at least one witness helped at each event. It concludes that each individual becomes less likely to help whenever group size increases. Identify the outcome mismatch and explain what further evidence is needed.
Le résultat mesuré est toute aide au niveau de l'événement, tandis que la conclusion porte sur la probabilité d'un individu particulier. La participation individuelle et des comparaisons appropriées entre tailles de groupe sont nécessaires, en tenant compte du danger, des opportunités disponibles, des liens avec les témoins et d'autres différences. Plus de témoins peuvent modifier différemment les résultats au niveau individuel et événementiel ; la mesure événementielle seule ne prouve pas un effet universel.
Recap
- Separate dispositional explanations, situational explanations and evidence of bias.
- Identify the reported motive or instruction before naming an influence process.
- State study conditions, measurement units and denominators.
- Keep an observed rating difference distinct from its proposed mechanism.
- Group effects are conditional patterns, not automatic outcomes.
- conformity: Alignment with a group's responses, norms or expectations.
- fundamental attribution error: Underweighting situational constraints when explaining others' behaviour through dispositions.
- cognitive dissonance: An aversive state associated with inconsistent relevant cognitions, which can motivate attempts to reduce the inconsistency.
A person publicly agrees with a group. A report concludes that the person privately changed belief, has a permanently conformist personality, and would obey any authority because of the classic studies’ percentages. Explain three unsupported inferences.
L'accord public ne révèle pas si la croyance a changé ou si la raison était l'acceptation, l'information ou une autre influence. Une action contrainte n'établit pas une disposition permanente. Les taux issus de jugements particuliers ou d'une condition d'obéissance unique ne déterminent pas la réponse de cet individu dans tous les contextes, et la conformité n'est pas identique à l'obéissance. Examinez les motifs, les instructions, les preuves répétées et les comparaisons pertinentes.