You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
Conformity is a form of social influence involving a change in belief or behaviour in order to fit in with a group. It is sometimes called majority influence, because it results from pressure — real or imagined — from the majority of a group. Although the Edexcel Social Psychology topic is anchored by obedience and Milgram's research, conformity is examined as a closely related process, and the two are frequently contrasted. This lesson builds the foundational knowledge: the three types of conformity described by Kelman (1958); the two explanations for conformity offered by Deutsch and Gerard (1955) — normative and informational social influence; and the landmark experimental evidence supplied by Asch (1951, 1956), including his variations. Mastering the link between type, explanation and evidence is what separates a descriptive answer from an analytical one in the exam.
Key Definition: Conformity is a change in a person's behaviour or opinions as a result of real or imagined pressure from a person or group of people.
By the end of this lesson you should be able to:
Edexcel 9PS0 — Paper 1, Topic 1: Social Psychology. This lesson maps to the specification's treatment of conformity as a form of social influence set alongside obedience. It covers the three types of conformity — compliance, identification and internalisation — the two explanations (normative social influence, NSI, and informational social influence, ISI), and Asch's classic study together with the variables affecting conformity: group size, unanimity and task difficulty.
Assessment Objectives exercised:
Connects to…
Herbert Kelman (1958) identified three distinct levels of conformity, which differ in how deeply the change is internalised. The crucial idea is that conformity is not all-or-nothing: it varies in depth (how much the private opinion changes) and in permanence (whether the change outlasts the immediate group setting).
| Type | Definition | Depth of change | Permanence | Example |
|---|---|---|---|---|
| Compliance | Going along with the group publicly while privately disagreeing | Shallowest — behaviour changes but opinion does not | Temporary — lasts only while the group is present | Laughing at a joke you do not find funny because everyone else is laughing |
| Identification | Conforming to the opinions or behaviours of a group because you value membership of that group | Moderate — you adopt the group's views while a member, but may revert when you leave | Lasts while you value membership | A new employee adopting the dress code and attitudes of their workplace |
| Internalisation | Genuinely accepting the group's views because you have examined them and find them convincing | Deepest — both public behaviour and private opinion change permanently | Permanent — persists even outside the group | A person who converts to a new religion after studying its teachings and genuinely believing them |
The key examinable distinction is between public and private acceptance. In compliance there is public compliance but no private acceptance; in internalisation there is both. Identification is an intermediate case: the person genuinely accepts the group's position, but the acceptance is conditional on continued group membership, so it can be lost if the person leaves. A useful way to remember this is to ask two diagnostic questions of any scenario: Has the person's private belief changed? and Will the change survive once the group is gone? Compliance answers "no" to both; identification answers "yes, conditionally"; internalisation answers "yes, permanently".
Exam Tip: In Edexcel applied items you may be given a scenario and asked to identify which type of conformity is being shown. The key distinction is whether the change is public only (compliance), tied to group membership (identification), or genuinely believed and permanent (internalisation).
Examiners reward answers that show the types and the explanations are intertwined rather than learned as separate lists:
Edexcel's applied questions typically give you a short vignette and ask which type of conformity is shown, so it pays to practise the diagnostic reasoning. Suppose a student privately thinks a set text is overrated but praises it in a seminar because everyone else is enthusiastic; the moment the seminar ends, they revert to their true view. Here the private belief has not changed and the behaviour lasts only while the group is watching — this is compliance, most plausibly driven by NSI (the desire not to seem out of step). Now suppose the same student joins a debating society, adopts its confident speaking style and its shared opinions while they are a keen member, but drifts away from those views after leaving for university — the change was genuine but conditional on membership, which is identification. Finally, suppose the student reads widely, is persuaded by the arguments of a campaign group, and continues to hold and act on those beliefs years later, wherever they are — the change is genuine and permanent, which is internalisation, most plausibly driven by ISI (they became convinced the position was correct). The discriminating questions are always the same two: has the private belief changed? and will the change outlast the group? Answering those two questions in order will reliably steer you to the right type, and naming the explanation alongside the type (NSI or ISI) is what lifts an answer from identification to analysis.
Morton Deutsch and Harold Gerard (1955) proposed two explanations for why people conform. They are called the two-process model because they identify two distinct psychological needs that drive conforming behaviour: the need to be right and the need to be liked.
Key Definition: Normative Social Influence (NSI) is conforming to the expectations of others in order to gain approval or avoid rejection. It is about the desire to be liked and accepted.
Key Definition: Informational Social Influence (ISI) is conforming because you believe others have better information or knowledge than you. It is about the desire to be right.
The two processes can be summarised as a contrast between being right (ISI, cognitive, leading to internalisation) and being liked (NSI, emotional, leading to compliance). This distinction is the analytical backbone of almost every conformity question.
A powerful way to see the two processes in action is to compare studies that used ambiguous tasks with studies that used unambiguous ones, because the two designs recruit different explanations.
Muzafer Sherif (1935) exploited the autokinetic effect — an optical illusion in which a stationary point of light in a completely dark room appears to move, because the eye has no frame of reference. Because there is genuinely no correct answer, the situation is maximally ambiguous. Sherif first asked participants individually how far the light "moved"; their private estimates varied widely. When he then placed them in groups and had them call out estimates aloud, their judgements converged on a shared group norm — and, tellingly, when tested alone again afterwards, participants retained the group estimate rather than reverting to their own. This retention is the signature of internalisation driven by ISI: with no objective anchor, participants genuinely used others' estimates as information about reality, and privately adopted the group norm as their own.
Asch (1951) was deliberately reacting against Sherif's design. He argued that convergence on an ambiguous task was unsurprising and told us little about conformity to a majority one knows to be wrong. By using an unambiguous line-judgement task — where the correct answer was obvious — Asch removed the informational reason to conform, so that any conformity that remained had to be normative: driven by the desire not to stand out. The contrast is instructive: Sherif's ambiguous task isolates ISI and produces internalisation, whereas Asch's unambiguous task isolates NSI and produces compliance. Holding the two studies side by side is an excellent way to demonstrate, in an exam, that you understand why the two-process model needs both explanations.
Exam Tip: If a scenario describes a genuinely uncertain situation (a new job, an ambiguous emergency), lead with ISI/internalisation; if it describes an obvious situation where the person goes along anyway to avoid standing out, lead with NSI/compliance. Matching the explanation to the ambiguity of the task is a reliable route to the application marks.
Solomon Asch (1951) conducted one of the most famous studies in social psychology. His aim was to investigate whether people would conform to the obviously wrong answer of a unanimous majority on an unambiguous task — one where there was a clear correct answer. This design was deliberate: by removing ambiguity, Asch could isolate normative pressure, because there was no informational reason to doubt one's own judgement.
To examine the extent to which social pressure from a majority group could make a person conform, even when the correct answer was obvious.
| Finding | Detail |
|---|---|
| Overall conformity rate | 37% of responses on critical trials were conforming (incorrect) |
| Participants conforming at least once | 75% conformed on at least one critical trial |
| Never conformed | 25% of participants never conformed on any trial |
| Control group | Error rate was under 1%, confirming the task was genuinely easy |
People will conform to a unanimous majority even when it is clearly wrong, primarily because of normative pressure — the desire to avoid the discomfort of standing out and being ridiculed. In post-experimental interviews, most conforming participants reported that they had continued to privately trust their own perception but went along with the group publicly to avoid disapproval — a textbook description of compliance driven by NSI. A minority, however, said they had genuinely come to doubt their own eyesight, hinting at an informational component.
Asch varied the conditions to identify factors that affect conformity. These variations are examinable in their own right and provide the strongest evidence for the two-process model.
| Variation | Change made | Result | Explanation |
|---|---|---|---|
| Group size | Varied from 1–15 confederates | Conformity rose up to 3 confederates then levelled off | A small unanimous majority is sufficient; adding more has diminishing returns — three is the "magic number" |
| Unanimity | One confederate gave the right answer | Conformity dropped to 5.5% | A dissenter breaks the unanimity of the majority and provides social support |
| Unanimity (different wrong answer) | One confederate gave a different wrong answer | Conformity dropped to about 9% | Breaking unanimity reduces conformity even when the dissenter is also incorrect — the break matters, not the correctness |
| Task difficulty | Lines made more similar in length | Conformity increased | When the task is more ambiguous, ISI increases because participants genuinely doubt their judgement |
| Private response | Participants wrote answers rather than speaking aloud | Conformity decreased significantly | NSI was reduced because the group could not observe the response |
The variations are analytically powerful. The private response result isolates NSI (removing observation removes the normative pressure), while the task difficulty result isolates ISI (introducing ambiguity creates an informational reason to conform). Together they demonstrate that both of Deutsch and Gerard's processes operate and can be manipulated independently.
graph TD
A["Group exerts pressure<br/>(unanimous wrong answer)"] --> B{"Is the task<br/>ambiguous?"}
B -->|"No — answer obvious"| C["Normative pressure dominates<br/>(NSI)"]
B -->|"Yes — genuinely unsure"| D["Informational pressure dominates<br/>(ISI)"]
C --> E["Compliance:<br/>public change only"]
D --> F["Internalisation:<br/>genuine private change"]
Asch's procedure required participants to judge the length of lines — a trivial, artificial task with no personal consequences. This matters because conformity in everyday life typically concerns decisions of genuine importance (political opinions, consumer choices, moral judgements) where the stakes and social dynamics differ. Because the task bore little resemblance to real-world conformity situations, we cannot be confident the findings generalise beyond the laboratory. The 37% figure may overstate conformity (participants had no investment in the line answers) or understate it (real conformity dilemmas carry far higher social costs). Either way, the artificiality weakens external validity.
Perrin and Spencer (1980) replicated Asch's procedure with UK engineering, mathematics and chemistry students and found conformity on only one of 396 trials. They argued that Asch's original results reflected the conformist climate of 1950s America, a decade dominated by McCarthyism, when non-conformity carried real social and political risk. This is a serious challenge because it suggests conformity is shaped by the historical and cultural context in which it is measured rather than being a stable feature of human nature — the study lacks temporal validity. A nuance worth noting for top marks is that the Perrin and Spencer sample (science and engineering students) may have been unusually confident in making objective judgements, so the comparison is not perfectly clean.
Asch's sample consisted entirely of male American undergraduates, which is both androcentric (men only) and ethnocentric (one Western, individualist culture). Cross-cultural research by Bond and Smith (1996), a meta-analysis of 133 Asch-type studies across 17 countries, found conformity was reliably higher in collectivist cultures (which prize group harmony and interdependence) than in individualist cultures (which prize personal autonomy). This matters because it shows conformity rates are not a fixed human constant but vary systematically with cultural values — so Asch's figure does not directly generalise to women, to older adults, or to non-Western populations.
Participants were deceived about the central purpose of the study and about the confederates' identity, so they could not give fully informed consent, and they may have experienced discomfort when their judgement conflicted with the group. This critique must be balanced: the deception was arguably necessary, since revealing the confederates' role would have destroyed validity (a cost–benefit justification), and there is little evidence of lasting harm. The implication is that while the study breaches modern ethical guidelines, the breach was relatively mild and the scientific value substantial.
A major strength is that Asch's variations provide direct experimental support for Deutsch and Gerard's explanations. The private answer condition reduced conformity, supporting NSI (removing observation removes the normative motive); the ambiguous task condition increased conformity, supporting ISI (uncertainty creates an informational motive). This dovetailing of theory and evidence is a genuine strength. However, the model has limits: it implies NSI and ISI are independent, whereas in reality they often operate simultaneously and are hard to disentangle. The dissenter variation illustrates this — a dissenter may reduce conformity both by reassuring the participant they will not be the lone deviant (reducing NSI) and by offering an alternative source of information (altering ISI).
There is also direct evidence that conformity via ISI depends on the individual's confidence in the task domain, which qualifies a purely situational reading. Lucas et al. (2006) asked participants to solve mathematics problems of varying difficulty after seeing others' (incorrect) answers. Conformity to the wrong answers was greater on the harder problems, consistent with ISI — but crucially, participants who rated their own mathematical ability highly conformed less than those low in self-efficacy, even on the difficult items. This matters because it shows that ISI is not a uniform response to ambiguity: the same objectively difficult task produces more conformity in a person who doubts their competence and less in a person who trusts it. The implication is that even the informational process, which the two-process model frames situationally, is moderated by a dispositional variable (task-specific self-efficacy), reinforcing the wider lesson that conformity emerges from an interaction of person and situation rather than from situational pressure alone.
Twenty-five per cent of Asch's participants never conformed on a single trial. The two-process model, which emphasises the situation (group pressure, ambiguity, observability), cannot fully explain why some individuals are consistently resistant. This points to a role for dispositional factors such as personality, self-esteem and locus of control. A complete account of conformity must therefore integrate situational and dispositional explanations rather than privileging one — the behaviour is best understood as an interaction between the person and the situation, a theme that recurs across the whole Social topic.
The two-process model reaches into real behaviour, which raises its external validity beyond the laboratory. NSI helps explain why advertisers use "bandwagon" and social-proof tactics ("nine out of ten customers prefer…"), which exploit the desire to belong. ISI helps explain why people copy others in genuinely uncertain situations, such as evacuating a building when they see others doing so. The model has also informed social-norms campaigns that correct misperceptions about peer behaviour (for example, telling students that most peers drink less than they assume) to reduce harmful behaviour. A theory that predicts and shapes real behaviour has demonstrated value beyond artificial line-judgement tasks.
Specimen question modelled on the Edexcel 9PS0 paper format
Evaluate the two-process model as an explanation of why people conform. (12 marks)
This 12-mark item is weighted toward evaluation, marked as roughly 4 marks AO1 (accurate knowledge of NSI and ISI and how they link to types of conformity) and 8 marks AO3 (evaluation — research support, methodological and ethical critique, the individual–situational debate, real-world relevance). Application (AO2) marks appear only when a question carries a scenario stem. Edexcel rewards a clear line of argument in which evaluation is developed rather than listed.
The two-process model says people conform for two reasons. Normative social influence (NSI) is when you conform because you want to be liked and accepted, so you go along with the group to avoid rejection; this usually leads to compliance, agreeing in public but not private. Informational social influence (ISI) is when you conform because you think other people know better, especially when a situation is unclear; this usually leads to internalisation, genuinely changing your mind.
Asch's study supports the model. He showed participants lines with confederates who gave wrong answers, and 37% of responses were conforming. When participants wrote their answers down privately, conformity went down, which shows they were conforming to be liked, supporting NSI. A weakness is that the study was artificial because judging lines is not like real life, so it might not generalise. It was also only done on American men, so it is biased. Overall, the two-process model is a useful explanation but it has weaknesses.
Examiner-style commentary: To reach top-band, this answer needs to explain why each weakness matters rather than just stating it, and to develop the evaluation with named follow-up research. It correctly links NSI to compliance and ISI to internalisation and uses the private-answer variation well, but the AO3 is a short list ("artificial", "biased") with no consequence drawn out — the classic Grade C ceiling. Naming Perrin and Spencer for temporal validity, and Bond and Smith for culture, and then saying what each implies for the model, would lift it.
The two-process model (Deutsch and Gerard, 1955) explains conformity through two motives. Normative social influence reflects the need to be liked: it is an affective process rooted in the drive to belong, and because the change is motivated by social approval rather than conviction it typically produces compliance — public agreement without private acceptance, lasting only while the group is present. Informational social influence reflects the need to be right: it is a cognitive process triggered by uncertainty, and because it changes what the person believes to be true it typically produces internalisation — a genuine, durable change of opinion. These map onto Kelman's (1958) types, with identification occupying an intermediate position.
The model's principal strength is that Asch's (1956) variations operationalise and isolate each process. The private-response condition reduced conformity, demonstrating NSI: once participants could not be observed, the normative cost of dissent disappeared. The task-difficulty condition increased conformity, demonstrating ISI: genuine ambiguity created an informational rationale for deferring to others. The same paradigm manipulates the two motives independently, which is powerful evidence that both are real and distinct. However, temporal and cultural validity are questionable: Perrin and Spencer (1980) found conformity on just one of 396 trials with British science students, implying Asch's findings reflected McCarthy-era America, and Bond and Smith's (1996) meta-analysis found systematically higher conformity in collectivist cultures — together suggesting the strength of each process is historically and culturally contingent. A further limit is that the model treats NSI and ISI as independent, yet the dissenter variation shows them entangled: an ally reduces both the normative cost of deviating and the informational pull of the majority. The most defensible conclusion is that the two-process model is a genuinely useful, experimentally grounded framework, but one that requires integration with cultural context and individual differences — recognising conformity as an interaction of person and situation.
Examiner-style commentary: This response would sit at the top of the mark range because the AO3 is sustained and each point follows a point–evidence–explanation–implication structure. The temporal and cultural critiques are synthesised into a single argument about contingency rather than listed separately, and the independence assumption is interrogated using the dissenter variation. The discriminator over a merely strong answer is the quality of the reasoning chain, not the volume of description — a further improvement would be to note the real-world social-norms application to show external validity as well as internal validity.
Contemporary research has refined the classic picture. Studies of conformity on social media show that NSI operates powerfully online even when the "group" is anonymous and physically absent — likes, shares and visible engagement counts create normative pressure to align with apparent majority opinion, suggesting the need for social approval does not require face-to-face observation in the way Asch's private-response variation implied. Neuroscientific work using fMRI has found that conforming to a group judgement is associated with activity in reward-related brain regions, and that deviating from the group activates regions linked to error monitoring and social pain — lending biological plausibility to NSI. A useful debate to pursue at undergraduate level is whether online "echo chambers" represent internalisation (people genuinely persuaded by repeated exposure via ISI) or mere compliance (people publicly signalling group membership via NSI) — and whether the two can even be distinguished when public and private behaviour increasingly merge. Turner's self-categorisation theory and the idea of referent informational influence offer a more group-based reframing of these processes for those reading beyond A-Level.
This content is aligned with the Edexcel A-Level Psychology (9PS0) specification.