HookThe computer dance that blew up its own hypothesis
In 1966 a team of Minnesota psychologists led by Elaine Walster invited 752 first-year students to a 'Welcome Week' dance, telling them a computer had used a personality questionnaire to pair them with their ideal match. The questionnaire was a prop; the pairings were random. What the researchers were really testing was the matching hypothesis — the tidy prediction that people seek partners of similar social desirability to themselves. They expected the plainer students to hit it off with plainer partners and the beautiful with the beautiful.
The data refused to co-operate. The one thing that predicted whether a student wanted to see their partner again was brutally simple: physical attractiveness. Everyone, regardless of their own looks, preferred the most attractive partner they were given; intelligence, personality and self-esteem barely registered. The study had set out to prove that we settle for our own level and instead caught us all reaching for the top. It took later, more realistic experiments — where rejection was a genuine risk — to rescue a modified matching hypothesis. That reversal is the perfect way into this topic, because Relationships is a topic of competing theories, each with named studies that support and undermine it. Your job in Paper 3 is not to recite one account of love but to weigh several against the evidence — which is exactly what Walster's students did to her hypothesis, without meaning to.
ModelEvolutionary explanations — love as a survival strategy
The evolutionary account starts from Darwin's sexual selection: characteristics that improve mating success get passed on, even when they cost survival (the peacock's tail). In humans the engine is anisogamy — the mismatch between plentiful, mobile sperm and rare, costly eggs — which, through Trivers' theory of parental investment, predicts different strategies for each sex. Because a woman's minimum investment (nine months of pregnancy) dwarfs a man's, females are choosier (inter-sexual selection, choosing quality) while males compete for access and favour quantity and youth (intra-sexual selection).
The landmark evidence is Buss (1989), a survey of over 10,000 people across 37 cultures. The pattern was strikingly consistent: women valued resources and ambition in a partner (cues to provisioning), men valued youth and physical attractiveness (cues to fertility), and both sexes prized kindness and intelligence. Clark and Hatfield (1989) add a sharp field demonstration of the same logic — when attractive strangers on a US campus propositioned passers-by, 75% of men agreed to casual sex and not a single woman did. The strength is cross-cultural consistency pointing to an evolved basis; the standing criticism is that these preferences may reflect economic and cultural conditions (women historically denied resources of their own) rather than genes, and that the theory struggles to explain homosexual relationships.
MechanismFactors affecting attraction — how strangers become couples
Three factors turn proximity into attraction. Self-disclosure — revealing personal information — deepens intimacy, and Altman and Taylor's social penetration theory describes it as a gradual peeling of layers, from shallow to deep, that works best when it is reciprocated: I disclose, you respond in kind, trust grows. Physical attractiveness matters more than we admit, partly through the halo effect (we assume good-looking people also have good personalities), and the matching hypothesis refines the story — over time we tend to pair with partners of roughly our own perceived attractiveness, balancing desire against the fear of rejection that Walster's realistic follow-up captured.
Filter theory (Kerckhoff and Davis, 1962) ties these into a sequence. From the wide field of everyone we could meet, a series of filters narrows the 'field of availables' to the 'field of desirables'. First social demography (proximity, class, education) removes those we will never realistically encounter; then similarity in attitudes and values matters most in the early, under-18-month stage; and finally, in longer relationships, complementarity — each partner meeting needs the other lacks — becomes the stronger predictor. The theory usefully explains why relationships pass through phases with different priorities, though modern research questions whether the 18-month cut-off, or the filters themselves, still hold in an era of online dating.
ModelWhy relationships last — and why they end
Four economic and process theories dominate this leaf. Social exchange theory (Thibaut and Kelley) treats relationships as a marketplace: we track rewards minus costs and stay while the profit is acceptable, judged against two yardsticks — the comparison level (what we feel we deserve, set by past relationships) and the comparison level for alternatives (whether a better deal is available elsewhere). Equity theory (Walster) corrects SET's crude accountancy: what matters is not maximum profit but fairness — a relationship feels stable when both partners perceive their ratio of rewards to costs as roughly equal, and both over-benefiting and under-benefiting breed dissatisfaction.
Rusbult's investment model is the most powerful because it explains a puzzle the others cannot: why people stay in unhappy, even abusive relationships. Commitment, Rusbult argues, depends on three things — satisfaction, the quality of alternatives, and, decisively, investment (the resources sunk in that would be lost on leaving: children, shared history, a home). High investment plus poor alternatives can produce commitment even when satisfaction is low. Finally, Duck's phase model maps breakdown as passing four thresholds: the intra-psychic phase ('I can't stand this', brooding privately), the dyadic phase (confronting the partner), the social phase (going public, friends take sides), and the grave-dressing phase (constructing a face-saving story of what went wrong).
A model AO3 paragraph on Rusbult's investment model shows how to convert a study into evaluation. 'A strength of the investment model is that it explains why people remain in dissatisfying relationships, which social exchange theory cannot. Rusbult and Martz (1995) studied women at a domestic-violence shelter and found that those who reported the greatest investment and the fewest attractive alternatives were the most likely to return to an abusive partner. This is a genuine advantage because it shows the model captures a real and troubling pattern — commitment surviving the collapse of satisfaction — that a simple rewards-minus-costs account would predict should never happen. However, the model is largely correlational: it shows investment and commitment rise together, but cannot establish that investment causes people to stay rather than committed people choosing to invest more. This matters because it weakens the model's claim to explain behaviour rather than merely describe it, and suggests a more cautious conclusion — that investment is a reliable marker of commitment without necessarily being its engine.' Notice the build: a clear point, a named study as evidence (Rusbult and Martz), an elaboration of why the strength counts, then a developed methodological counterpoint that ends on a qualified judgement. Point, evidence, elaboration, counter-argument, conclusion — the chain, not the count of points, is what earns the top AO3 band.
MechanismVirtual relationships — does the screen change the rules?
Computer-mediated communication forced psychology to ask whether online relationships obey the same laws as face-to-face ones, and two theories give opposite answers about self-disclosure. Reduced cues theory (Sproull and Kiesler) is pessimistic: because online interaction strips out the non-verbal cues — facial expression, tone, physical presence — that build trust, it encourages deindividuation and less meaningful self-disclosure, leaving communication blunt and impersonal. Walther's hyperpersonal model argues almost the reverse: online relationships can become more intimate than face-to-face ones, because the sender enjoys selective self-presentation (editing an idealised version of themselves) and disclosure happens earlier and more intensely — a relationship can 'boom and bust' faster online.
The third idea is the absence of gating (McKenna and Bargh). In face-to-face life, 'gates' — a stammer, shyness, physical features judged unattractive — obstruct relationships before they start. Online, those gates are invisible, so the relationship forms on the basis of self-disclosed content rather than surface features, which can liberate the shy and socially anxious. But the same absence of gating cuts both ways: it also lets people build entirely false identities, so the freedom that helps some deceive others. The evaluation examiners reward is that these theories are not simply rivals — the truth depends heavily on which platform, and which relationship, you are describing.
CaseParasocial relationships — loving someone who never met you
A parasocial relationship is one-sided: the fan knows a great deal about a celebrity who does not know they exist. McCutcheon's Celebrity Attitude Scale identifies three deepening levels. The entertainment-social level is casual and social ('my friends and I love talking about her'); the intense-personal level involves obsessive feelings and a sense of a real bond ('I consider him my soulmate'); and the borderline-pathological level features uncontrollable fantasies and a willingness to act on them ('I would break the law to meet her'). Most fans never leave the first level.
Two explanations compete. The absorption-addiction model (McCutcheon) argues that people with a weak or deficient sense of identity become absorbed in a celebrity to fill the gap, and — like any addiction — need ever stronger involvement to sustain the effect, escalating up the levels, often triggered by personal crisis. The attachment theory explanation traces the tendency to early childhood: those with an insecure-resistant attachment type, who crave closeness but fear rejection, may find parasocial relationships attractive precisely because a celebrity can never actually reject or abandon them — intimacy with none of the risk. The two accounts are complementary, one describing the psychological deficit and the other its developmental root.
VocabularyKey terms the mark scheme pays for
TrapsMisconceptions that cost marks
ExamWhat examiners want
Relationships is a Paper 3 optional topic examined in a mix of short AO1/AO3 questions and the 16-mark essay (6 marks AO1 description, 10 marks AO3 evaluation), and its defining feature is that almost every leaf offers competing theories — SET versus equity versus Rusbult, reduced cues versus hyperpersonal, absorption-addiction versus attachment. That structure is a gift: the built-in counter-theory is ready-made evaluation, so your strongest AO3 is often to pit one explanation against another and show which the evidence favours. Keep a named study attached to every theory (Buss for evolution, Walster for matching and equity, Rusbult and Martz for investment, McCutcheon for parasocial), because a specific study is what turns a vague point into credited evaluation.
The essays are marked in four bands, and Level 4 (13-16 marks) is reserved for AO3 that is 'thorough and effective' — meaning elaborated, not listed. Build each evaluative point as a chain: state it, support it with a named study, explain why it matters, then develop a counter-argument, and let the essay argue towards a conclusion rather than trailing off. Watch the AO1/AO3 balance: on a 16-marker roughly a third of your words should describe the theory accurately and two-thirds should evaluate, so resist the urge to over-describe Buss's method at the expense of judging it. Where a question names a specific theory ('Discuss Rusbult's investment model'), stay on that theory — bringing in equity theory only as an explicit point of comparison, never as a substitute. Precise studies, elaborated evaluation and a visible line of argument are what carry this topic into the top band.