Issues and debates — Edexcel A-Level Psychology

Issues and debates in Pearson Edexcel A-Level Psychology explore the overarching philosophical, ethical, and methodological dilemmas that cut across all core approaches. Mastering these synoptic debates enables students to critically evaluate psychological theories and empirical research across social, cognitive, biological, and learning perspectives.

Issues and debates in Pearson Edexcel A-Level Psychology synthesize the philosophical foundations, ethical considerations, and methodological assumptions across all psychological disciplines. This synoptic topic challenges students to analyze how conflicting stances on determinism, reductionism, scientific status, nature versus nurture, and social control shape empirical discovery.

Mastering this component is critical for securing top marks in Paper 3 (General and Practical Applications), where students must contrast disparate theories and assess real-world implications. It provides the analytical toolkit required to evaluate whether psychological research benefits human welfare or imposes coercive normative values.

• →Nature vs. Nurture: The debate over whether human behavior is predetermined by genetic and physiological processes (nature) or shaped by postnatal environment, socialization, and conditioning (nurture), culminating in the modern interactionist approach and epigenetics.
• →Reductionism vs. Holism: The contrast between analyzing complex psychological phenomena by isolating their simplest constituent units (biological or environmental reductionism) versus studying the integrated system as a whole (gestalt or humanistic holism).
• →Psychology as a Science: The evaluation of psychological paradigms against classic scientific canons, including empirical observation, Karl Popper's falsifiability, hypothesis testing, objectivity, replicability, and Thomas Kuhn's concept of paradigm shifts.
• →Social Control: The application of psychological theories (e.g., token economies, aversion therapy, clinical diagnoses) to regulate, shape, and coerce individual behavior in institutional, criminal, and societal domains.

• The British Psychological Society code of ethics and conduct (2009) outlines four core principles: respect, competence, responsibility, and integrity.
• In obedience research, deception is often used to prevent demand characteristics, as seen in Milgram's rigged draw, which breaches the principle of informed consent.
• Protection from harm is a key ethical issue; Milgram's participants experienced severe distress, whereas Burger (2009) implemented a 150-volt stopping rule to mitigate this.
• The implications of findings refer to how society uses the research; for example, obedience research might be misused to excuse atrocities, while prejudice research can inadvertently stigmatise vulnerable groups.
• Questionnaire design must avoid leading questions, jargon, and double-barrelled items to ensure validity, with pilot studies used to identify and fix these flaws.
• Social desirability bias is a systematic issue where respondents answer in a way that presents them favourably, which damages validity; anonymity or filler items can reduce this.
• In interview design, structured interviews provide quantitative data that is easy to replicate and analyse, whereas unstructured interviews yield qualitative data that is harder to analyse but richer in detail.
• Interviewer effects occur when the interviewer's age, gender, or tone influences the respondent's answers, acting as an extraneous variable in self-report data.
• Recording decisions in interviews present practical trade-offs: taking notes can distract the respondent and miss details, while audio recording provides a full transcript but may make the respondent self-conscious.
• Define the debate through levels of explanation, running from the social and cultural through the psychological to the biological, rather than as a synonym for oversimplification.
• Give an example from this topic and state what it leaves out, such as social impact theory predicting how much obedience occurs without explaining why.
• Offer the holistic counterweight, for instance Sherif keeping the group rather than the individual as the unit of analysis at Robbers Cave.
• Argue the case for parsimony: simpler variables can be operationalised, measured and tested, which is how social impact theory generates predictions at all.
• Compare on stated criteria such as the proposed cause, the supporting evidence, the predictions made and what each implies for reducing prejudice.
• Use the minimal group result as decisive evidence that competition is not necessary, and the superordinate goal result as evidence that competition is sufficient.
• Note what the two share, that both are situational and so say little about why individuals in the same setting differ.
• Reach a judgement, for example that they operate at different levels and are complementary rather than rivals.
• Name the criteria of science and test the method against each, instead of asserting that psychology either is or is not a science.
• Keep reliability and validity apart: standardised wording raises reliability while leaving social desirability free to damage validity.
• Use falsifiability as a discriminating test, for example that the agentic state is inferred from the very obedience it is meant to explain.
• Credit replication as evidence of scientific status, citing Burger's repetition of Milgram's design and the closeness of the two figures.
• Quote a cross-cultural replication with its figure and then note that the procedures differed, so the comparison is not a clean test of culture.
• Separate a real difference between groups from bias in the research, for instance a sample of American men whose results were generalised to everyone.
• Use alpha bias and beta bias accurately, as exaggerating and as minimising differences between groups.
• Identify ethnocentrism where a paradigm designed in one culture is transported into another, which is an imposed etic.
• Set both explanations against the same behaviour, so a variation that shifts the obedience rate becomes evidence that the person alone cannot account for it.
• Point out that the authoritarian personality is itself an environmental account, tracing the trait to harsh parenting, so dispositional and inherited are not the same claim.
• Apply the limits of correlation: the personality findings are correlational, so a high score may follow from obeying, or both may follow from a third variable.
• Argue for interaction rather than a winner, for example that personality matters most where the situation leaves room for a choice.
• Explain why the modern replication had to be partial, linking the stopping point to Milgram's own finding about who continued beyond it.
• Report the comparison honestly: the later rate was slightly lower and the difference was not statistically significant, so the claim is continuity rather than change.
• Show development in ethics and method as well as in findings, for instance screening, a lower sample shock and repeated reminders of the right to withdraw.
• Bring in a second strand, such as Reicher and Haslam's challenge to automatic role conformity, so the answer shows theory developing and not just one study repeated.
• Credit is given for naming the agent of control and the mechanism, so write about legitimacy of authority, uniform as a visible cue to that legitimacy, gradual commitment and the agentic shift rather than simply saying that people obey.
• A named study with its finding earns the knowledge mark: Bickman's guard condition producing higher compliance than the civilian condition, or Milgram's sharp fall in obedience when an ordinary man rather than the experimenter gave the orders.
• Answers on reducing prejudice score when they say what Sherif changed, that competition for a scarce prize created hostility and that only goals needing both groups reduced it, and when they tie this to the conditions of the contact hypothesis.
• Evaluation credit comes from weighing situational against dispositional explanations such as the authoritarian personality, and from naming the ethical implications of publishing a reliable recipe for making people comply.
• Name the application and the psychology underneath it, so the jigsaw classroom is credited as interdependence plus the contact conditions of equal status, a common goal and institutional support.
• Effectiveness evidence earns credit, so quote a finding rather than a hope, for example that jigsaw classrooms raised cross group liking and minority attainment without harming majority attainment.
• Credit for limits: contact based schemes work best where the groups already have equal standing, and they do not touch the structural inequality or the competition for resources that Sherif showed creates hostility in the first place.
• Linking obedience research to practice, such as aircrew and hospital training that authorises a junior to challenge a senior, shows the knowledge being used beyond prejudice.
• Define sensitivity by consequence rather than by discomfort: the test is whether publication could damage the standing, employment or treatment of the group the participants stand for.
• Using the stages set out by Sieber and Stanley to structure the answer earns organisation marks, above all the point that a researcher cannot control how findings are used once they are in the public domain.
• Credit for naming a real risk inside social psychology, such as an obedience difference between two countries being reported as a national trait, or prejudice research implying that racism is an abnormal personality rather than an ordinary group process.
• A balanced conclusion scores: refusing to study race, obedience or prejudice leaves policy to be made with no evidence at all, so the judgement is about how such work is done and reported, not about whether it happens.
• Naming the guidelines at issue, confidentiality, informed consent, the right to withdraw and protection from harm, and saying how each was or was not met in this particular case.
• Using the evidence rather than the fame of the case: a normal digit span with no new long-term memories, and steady improvement on mirror drawing with no recollection of practising, which is why the case supports more than one memory system.
• Weighing cost against benefit, since decades of testing one man produced findings that shaped the field, an argument examiners expect to see answered rather than assumed.
• Noting that his capacity to consent was permanently impaired, so what was obtained was presumptive consent from a guardian rather than informed consent from the participant.
• Extending the point to the publication of his identity and the sectioning of his brain after death, which reopened the question of whether that permission was his to give.
• Separating internal validity, whether the measure captured memory and nothing else, from ecological validity, whether the task resembles remembering outside the laboratory.
• Naming the measurement choice and its effect, for example that recognition produces higher scores than free recall because retrieval cues are supplied by the test itself.
• Identifying the threats by name, demand characteristics, experimenter effects, participant variables and order effects, each paired with the control that answers it.
• Using field evidence properly: the real witnesses studied by Yuille and Cutshall resisted leading questions of the kind that misled participants in Loftus and Palmer, which challenges how far laboratory results generalise.
• Arguing that artificiality is the price paid for causal inference, so the defensible conclusion is that laboratory and field methods are needed together.
• Identifying the level of explanation in play, so machine reductionism, the computer analogy of input, storage and output, is kept distinct from biological reductionism about brain regions.
• Giving the strength in scientific terms, that simpler units allow operational definitions, controlled experiments and predictions that could be shown false.
• Giving the counter case with evidence, that memory tasks recruit distributed networks and that an episodic buffer had to be added to account for integration.
• Reaching an interactionist conclusion, that different levels of explanation answer different questions and work best as complements rather than rivals.
• Using Bartlett's reconstructive account, where schema and effort after meaning reshape recall of the War of the Ghosts, as the holistic alternative.
• Structuring by point of comparison, such as the nature of the short-term store, the evidence each account explains and the predictions each makes, rather than by taking one model at a time.
• Naming the decisive evidence, since the serial position curve and case studies of amnesia support separate stores while dual task findings and the selective deficit shown by KF support separable subsystems.
• Saying what each account cannot do, for example that a unitary short-term store cannot accommodate a patient with a poor digit span and intact visual memory, and that the working memory model says little about long-term memory.
• Judging against agreed criteria, testability, breadth of explanation and application to practice such as revision strategy or the cognitive interview.
• Noting similarities as well as differences, since both models are information processing accounts that treat remembering as flow through components.
• Listing the features of science and applying each to a named memory study, objectivity, control, replication, falsifiability, empirical evidence and hypothesis testing.
• Using replication as concrete support, since a standardised word list procedure lets another researcher repeat the study and check whether the finding holds.
• Raising the problem of inference, that cognitive processes cannot be observed directly and are modelled from behaviour, so a model can be wrong in ways the data do not reveal.
• Bringing in falsifiability, with the central executive as the example of a construct too vaguely specified to be tested properly.
• Balancing control against validity, because the very features that make the laboratory scientific make its tasks unlike everyday remembering.
• Distinguish culture bias from cultural difference: credit comes from saying that a short span in Spanish speakers is a property of the language, specifically longer number words and the word length effect, not evidence of a weaker memory.
• Use Bartlett (1932) to show reconstructive memory is culturally loaded, naming the transformations in recall of The War of the Ghosts rather than simply calling the account schema theory.
• Name a gendered finding and its limit, such as Loftus, Banaji, Schooler and Foster (1987) showing an advantage that depends on whether the item is female or male oriented, so neither sex is globally superior.
• Evaluate the methodology: an imposed etic arises when a test standardised on English speakers is exported unchanged, and a beta bias arises when gender differences are assumed away by testing only one sex and generalising.
• Link to nomothetic method, since culture and gender bias are what happens when a general law is drawn from an unrepresentative sample.
• State the dissociation precisely: surgery spared immediate span and procedural learning while destroying the ability to lay down new episodic memories, which argues for biologically separate systems.
• Give the nurture evidence with its mechanism, namely that post event information and cultural schema alter what is reported at recall rather than only how confidently it is reported.
• Argue for interaction rather than a winner, using evidence that experience changes the brain itself, so the structures the nature account depends on are themselves shaped by use.
• Credit the case study method for its richness and detail while noting that one surgical patient cannot be generalised and that the lesion extent was only mapped accurately by later scanning.
• Connect the debate to determinism, since a purely biological account implies memory failure is fixed while an experiential account implies it can be altered by how a witness is questioned.
• Show the logic of theory change: credit is given for naming the finding that a model could not explain and then the modification it forced, not for describing two models side by side.
• Use patient KF to argue that a single short term store cannot be right, since damage was selective to verbal input rather than to storage as a whole.
• Explain the dual task method and why it matters, because two tasks using the same component interfere while two using different components do not, which is how separate slave systems were demonstrated.
• Note the limits of the newer model, since the central executive remains poorly specified and is often criticised as a homunculus that is described by what it does.
• Offer a competing account such as levels of processing from Craik and Lockhart (1972), where durability follows depth of processing rather than time in a store.
• Identify the mechanism of control, namely that leading questions, repeated questioning and suggestion change what a witness reports rather than only how the report is judged.
• Name protective applications with their evidence, such as cautions on identification evidence following the Devlin committee and the cognitive interview raising correct recall without a matching rise in errors.
• Argue both directions, since the same research that protects a defendant can be used by an expert witness to discredit a truthful victim.
• Evaluate the ecological validity of the laboratory base, because watching a filmed crash carries no consequence and real witnesses to a violent event are aroused and personally involved.
• Link to ethical implications and socially sensitive research, since findings about false memory carry consequences far beyond the participants who produced them.
• Say what the application does and why the psychology predicts it will work, for example that cued recall supports a damaged retrieval process while storage itself may be relatively spared.
• Give effectiveness evidence honestly, crediting the small improvements in mood, communication and quality of life while stating that the certainty of the evidence is low.
• State the limits, since these approaches manage the experience of memory loss and do not treat the neurodegeneration that causes it.
• Use a second application from a different setting, such as retrieval practice in schools or the cognitive interview in policing, to show the knowledge generalises.
• Evaluate on practical grounds, considering cost, staff training, the possibility of distress when a cue recalls a painful event, and whether benefits persist after the sessions stop.
• Define sensitivity by consequence rather than by topic, noting that the risk extends to the group the participant represents and not only to the individual in the study.
• Apply the specific consent problem, since capacity can fluctuate, so researchers use process consent with repeated checks and a consultee where capacity is absent.
• Give the concrete harms: distress when memory failure is exposed, loss of privacy for a highly identifiable patient, and stigma if results are reported without the range of individual variation.
• Argue the other side, because avoiding sensitive work protects nobody and leaves a group without evidence for better care, which is itself an ethical cost.
• Reference safeguards such as ethics committee review, the right to withdraw exercised on the participant's behalf, debriefing and careful wording of published conclusions.
• Separating ethics within the research from the ethical implications of its findings, and answering whichever the question names.
• Applying the code to real aggression research, for instance Bandura, Ross and Ross (1961), where nursery children watched an adult attack a Bobo doll and then reproduced the specific acts, raising protection from harm and the fact that children cannot themselves give informed consent.
• Treating confidentiality as a practical problem rather than a form to sign, since Brunner et al. (1993) studied one extended Dutch family carrying a mutation affecting the MAOA gene alongside impulsive violence, and a family that distinctive cannot realistically be anonymised.
• Arguing about the use of findings with a case: Raine et al. (1997) reported reduced glucose metabolism in prefrontal areas of defendants who had pleaded not guilty by reason of insanity, and the same result can support early treatment or a claim that violent people are born that way and should be detained, which is why Raine himself cautioned against the second reading.
• Covering animal research honestly, since procedures on rodents are licensed under the Animals (Scientific Procedures) Act 1986 and defended by a cost and benefit judgement, while the animal can never consent and the extrapolation to humans is itself disputed.
• Saying what each technique physically measures and what follows from it, so the blood oxygen level dependent signal is an indirect proxy for neural activity with poor temporal resolution, while a structural scan shows anatomy and can say nothing about function.
• Using the multiple comparisons problem concretely: Bennett and colleagues scanned a dead Atlantic salmon and obtained apparent activation when the thousands of simultaneous tests in a scan were left uncorrected, which is why correction and replication are demanded.
• Connecting cost to evidence quality, since expensive scanning produces small samples, low power and findings that may not generalise, and clinical or forensic samples such as the offenders in Raine et al. (1997) are opportunistic rather than randomly allocated.
• Questioning the task, because genuine aggression cannot be provoked inside a scanner, so researchers use proxies such as provocative images or a competitive reaction time task, and what the scan measures is the response to the proxy.
• Recognising the constraints of the other methods too, since positron emission tomography involves a radioactive tracer that limits repeat scanning and excludes some participants, and case evidence such as Phineas Gage in 1848 has no control group at all.
• Distinguishing the levels clearly, so biological reduction is about the level of explanation chosen and experimental reduction is about how the behaviour was operationalised for measurement.
• Supplying the evidence the reductionist account rests on, such as the raised salivary testosterone found in inmates convicted of violent crimes by Dabbs et al. (1987), the repeated association between low serotonin activity and impulsive aggression, and the mutation affecting the MAOA gene reported by Brunner et al. (1993).
• Stating the strength properly: isolating one variable produces falsifiable predictions, replicable methods and treatment targets, which is exactly why the biological approach is treated as scientific.
• Making the holistic rebuttal with evidence rather than assertion, since Caspi et al. (2002) found that the low activity variant of the MAOA gene predicted antisocial outcomes mainly in boys who had been maltreated, so gene and environment interact and neither level alone is sufficient.
• Noting that aggression is not one behaviour, because planned and reactive aggression have different correlates, so a single neural mechanism cannot be expected to explain every instance.
• Setting out the psychodynamic explanation accurately, with aggression as an expression of Thanatos, the destructive drive Freud proposed in his later work, redirected outwards to protect the self and displaced onto a substitute target by the ego's defences.
• Setting out the biological explanation with its evidence, including the reduced prefrontal glucose metabolism reported by Raine et al. (1997), the raised testosterone in violent offenders found by Dabbs et al. (1987), and twin concordance implying an inherited contribution.
• Organising the answer by criterion rather than by theory, taking cause, method of study, falsifiability, degree of determinism and practical application one at a time and judging both explanations on each.
• Identifying real similarities as well as differences, since both accounts are determinist, both treat aggression as arising from within, and neither gives much weight to learning or to the immediate situation.
• Using evidence that bears on the comparison, such as Bushman (2002), who found that participants who vented anger by hitting a punchbag were subsequently more aggressive rather than calmer, which counts against the idea that aggression is a pressure discharged by release.
• Naming the criteria before applying them, so objectivity, empirical method, control, replicability, falsifiability and paradigm each become a test the evidence can pass or fail.
• Using synaptic transmission as the strongest example, describing how an impulse causes vesicles to release neurotransmitter into the synaptic cleft, how it binds to receptors on the next neuron to produce an excitatory or inhibitory effect, and how it is then broken down by enzymes or taken back up, with each step measurable and repeatable.
• Turning that into a scientific argument rather than a description, since the account generates testable predictions, most obviously that blocking the reuptake of serotonin will alter mood, which is how antidepressant evidence is gathered.
• Presenting scanning as objective, standardised measurement while conceding that the blood oxygen signal is an indirect proxy, that Bennett and colleagues obtained apparent activation from a dead salmon when comparisons were left uncorrected, and that a large coordinated replication project found only about a third of published psychology findings replicated at their original strength.
• Explaining reverse inference as the specific flaw, because reading a mental state off an active region assumes that region does only one thing, which is untrue of the amygdala and of the prefrontal cortex.
• Name the bias precisely: alpha bias exaggerates differences between males and females, beta bias minimises them by generalising male samples, and androcentrism treats male behaviour as the norm against which female behaviour is judged deviant.
• Dabbs et al. (1987) sampled salivary testosterone from male prison inmates and found higher levels among those convicted of violent offences and among those with more disciplinary violations, which is a correlation in a self-selected sample, not a cause.
• Van Goozen et al. (1995) studied transgender people before and during hormone treatment and reported increased aggression proneness in those given testosterone and reduced aggression proneness in those given anti-androgens, which is stronger evidence because the hormone is manipulated rather than measured.
• Cross-cultural evidence works against a purely hormonal account: Mead described societies in which both sexes were gentle and others in which both were aggressive, and homicide rates differ enormously between populations whose hormones do not.
• Bjorkqvist, Lagerspetz and Kaukiainen (1992) found girls used more indirect aggression such as exclusion and rumour spreading, so a study that counts only punches builds the gender difference into its own measure.
• Credit for the direction of causation: on the challenge hypothesis testosterone rises after winning a contest, so hormone levels may follow dominance and aggression rather than produce them.
• Raine et al. (1997) compared forty-one murderers with matched controls during a continuous performance task and found lowered prefrontal and corpus callosum activity with abnormal limbic asymmetry; Raine himself concluded this does not show murderers are born, only that violence has a biological correlate.
• Say why a scan cannot settle the debate: it is correlational, the abnormality may be a consequence of a violent life including head injury or substance use, and the sample was people pleading insanity rather than violent offenders in general.
• Brendgen et al. (2005) found in young twins that physical aggression was substantially heritable while social or indirect aggression was better explained by shared environment, which shows the debate has different answers for different kinds of aggression.
• Caspi et al. (2002), in the Dunedin longitudinal cohort, found the low activity MAOA variant raised antisocial outcomes chiefly among men who had been maltreated as children, which is diathesis stress demonstrated in one dataset.
• Environmental evidence includes Bandura's children imitating an aggressive model they had only watched, and situational triggers such as alcohol, provocation and deindividuation.
• Conclude with interaction rather than a tally: epigenetic change means experience alters gene expression, so the two influences cannot be separated into percentages of a person.
• Set the techniques in order and say what each one added: post-mortem and case study gave inference from damage, electroencephalography gave timing, computerised axial tomography gave structure only, positron emission tomography gave function using a tracer, and functional magnetic resonance imaging gave function without radiation.
• Credit comes from linking a technique to a finding: positron emission tomography allowed Raine et al. (1997) to compare glucose use during a task, and functional magnetic resonance imaging allowed Gospic et al. to watch amygdala responses to provocation as they happened.
• Say explicitly how understanding changed. Localisation claims based on lesions have been replaced by network accounts, and language is no longer credited to Broca's area alone.
• Evaluate the tools on criteria: spatial resolution favours functional magnetic resonance imaging, temporal resolution favours electroencephalography, and positron emission tomography is invasive because it injects a radioactive tracer.
• Note the methodological correction as part of the development: early imaging used very small samples and flexible analysis, producing false positives, so larger samples and preregistration are now expected.
• Give a real mechanism of control, not a general worry: psychosurgery, medication, deep brain stimulation, or risk prediction from brain measures used in sentencing and parole decisions.
• Explain why consent is the pivot. A prisoner offered treatment linked to release, or a detained patient, cannot straightforwardly refuse, so the right to withdraw is compromised and consent is not fully informed or free.
• Use the statistics of prediction against neuroprediction: serious violence is rare, so a weak predictor flags far more people wrongly than correctly, and those people lose liberty for what a scan implies they might do.
• Tie the item to free will and determinism. If brain dysfunction causes violence, blame and punishment are undermined, which is the argument a defence makes when submitting a scan in mitigation.
• Give the other side for balance: medication that allows someone to leave hospital, or a programme that reduces assaults on staff, is also social control and is in the person's interest, so the judgement should weigh benefit against freedom.
• Each application needs three things for full credit: what it is used for, the evidence that it works, and what it does not touch. Lithium and serotonergic medication reduce impulsive aggression but do not address provocation, poverty or instrumental violence.
• Describe Novaco's three stages by name: cognitive preparation in which triggers are identified, skill acquisition in which relaxation and self talk are rehearsed, and application practice with role play under controlled provocation.
• Give the honest effectiveness picture. Gains appear on self-report and on institutional incident rates, evidence on reoffending is mixed, and brief programmes delivered to unmotivated prisoners have produced little measurable change.
• Gesch's trial is credited as evidence because it was randomised and placebo controlled, which rules out expectancy and makes a causal claim defensible, unlike correlational diet studies.
• State the social benefit concretely: fewer assaults in institutions, lower staff injury and lower cost, which is what the use of psychological knowledge within society means in the mark scheme.
• Distinguish confidentiality from anonymity. Anonymity means the researcher cannot identify the participant; confidentiality means the researcher can but undertakes not to disclose, which is what applies to offender and patient samples.
• Explain deductive disclosure: with a single family, a named institution or a rare condition, removing names is not enough because readers can work out who the participants are.
• Use the British Psychological Society code as the framework, naming valid consent, right to withdraw, protection from harm and data management, and note that prison and clinical samples raise particular questions about freely given consent.
• Argue both ways as Sieber and Stanley do: abandoning sensitive questions is itself a cost, since evidence about the causes of violence has therapeutic and legal value, so the judgement is a cost and benefit one taken by an ethics committee.
• Include the researcher's duty over how findings are reported, because genetic determinism in a headline can stigmatise a whole population even when the paper itself was cautious.
• Mention incidental findings in scanning research and the requirement to agree a protocol in advance for what participants will be told.
• Say what the animal actually underwent rather than that a study was unethical: surgery, food deprivation, inescapable shock and separation from a mother are the specifics a marker rewards.
• Set out the regulation by name, the Animals (Scientific Procedures) Act, Home Office project and personal licences, local ethics committee approval and the British Psychological Society guidelines for work with animals.
• Use the cost benefit argument properly: systematic desensitisation and token economies grew out of this work, so the defence is utilitarian, while the objection is that the animal cannot consent, cannot withdraw and gains nothing.
• Higher marks come from the philosophical position, Singer's charge of speciesism set against Regan's view that rights cannot be traded away, and Bentham's question about whether an animal can suffer rather than reason.
• Distinguish the two separate arguments, that animals differ qualitatively from people in language and cognition, and that the animal laboratory is so controlled that the behaviour is artificial even for that species.
• Name the constraint evidence, instinctive drift in Breland and Breland and biological preparedness in Garcia and Koelling, instead of asserting that animals are simply different from us.
• Give the counter argument, that basic conditioning processes clearly do transfer, since systematic desensitisation and token economies were built on animal work and do change human behaviour.
• Design points earn marks as well: standardised procedures, operationalised behavioural categories, control of extraneous variables, sample size, attrition in longitudinal work and the cost of running it.
• Define the position in a clause and then show the level of explanation, behaviour as stimulus and response units rather than as meanings, intentions, brain chemistry or social context.
• Argue the strength through application, since reducing behaviour to measurable units is precisely what made token economies and systematic desensitisation designable and testable.
• Argue the weakness through evidence that needs a higher level, latent learning in Tolman's maze work and the mediational processes in Bandura's account of observational learning.
• A holistic comparison earns credit, for instance that humanistic psychology treats the person as an indivisible whole and that an interactionist model of phobia keeps both the conditioning and the cognitive appraisal.
• Comparison marks need a stated basis, mechanism, evidence, degree of determinism or scope, with both explanations then judged against that same basis in the same paragraph.
• Show where the accounts agree, that behaviour is learned from the environment and neither requires an inherited predisposition, before showing where they separate.
• Hold one behaviour throughout, phobia, aggression or gender role, so the comparison stays concrete instead of collapsing into two separate descriptions.
• Bandura's reciprocal determinism is worth naming, because it is the one learning account allowing the person any influence over the environment that is shaping them.
• Give the criteria for a science explicitly, objectivity, empiricism, replicability, falsifiability, control and hypothesis testing, and then test the behaviourist method against each in turn.
• Evidence the claim rather than asserting it, for example that a Skinner box yields ratio level data and that conditioning findings have been replicated across many species and laboratories.
• Name the limits, that unobservable mediating processes are excluded by definition and that a reinforcement history supplied after the behaviour is not a falsifiable prediction.
• Credit is available for comparison with social learning theory, which keeps the experimental method while inferring attention, retention, reproduction and motivation from what the observer later does.
• Separate the two strands, that culture changes which behaviours are reinforced and which models are available, while gender concerns whose behaviour was studied and how the resulting difference is then applied.
• Use Becker and colleagues as evidence that modelled behaviour transmits into a culture that previously lacked it, rather than as a loose general comment about the media.
• Name the bias precisely, an imposed etic when American conditioning findings are treated as universal, and androcentrism, alpha bias or beta bias when a difference between men and women is exaggerated or ignored.
• Link it to the practical, since a gender comparison run on an opportunity sample in one location supports a tentative conclusion only, and stating it more strongly is socially sensitive.
• Define the debate in a clause and then apply it to the observation in the stem, saying which behavioural categories, such as hitting, helping or toy choice, could be acquired by modelling and reinforcement and which could plausibly reflect maturation, hormones or temperament.
• Support nurture with evidence rather than assertion: Bandura, Ross and Ross found imitation of specific novel acts, and Bandura (1965) showed that seeing a model punished suppressed spontaneous imitation while an offered incentive revealed that the children had learned the sequence anyway.
• Support nature properly: Garcia and Koelling on taste aversion and Seligman's idea of biological preparedness show some associations form in one trial and others barely form at all, and Menzies and Clarke (1993) found many water phobic children had no conditioning event to point to.
• Reach an interactionist conclusion using diathesis-stress or epigenetics, and note the methodological point that an observer records behaviour, never its origin, so an observation alone cannot settle where a gender or age difference comes from.
• Link the debate to environmental determinism and to reductionism, because a marker rewards the connection between the position taken and the wider debate it belongs to.
• Chart the sequence and say what each addition bought: Pavlov on reflex association, Watson's behaviourism refusing to talk about mental processes, Skinner on reinforcement schedules and shaping, then Bandura adding observation, vicarious reinforcement and the mediational processes of attention, retention, reproduction and motivation.
• Show that the research questions changed with society and technology, from a live and filmed adult model in the Bobo doll work, to television and eating attitudes in the Becker study of Fijian girls, to dehumanisation and desensitisation in violent video game research.
• Show therapy practice changing: aversion therapy was once used in an attempt to change sexual orientation and has been abandoned as ineffective and unethical, while systematic desensitisation, token economies and cognitive behavioural therapy are now used with agreed goals and recorded consent.
• Explain why understanding developed, naming better controls and measurement, the arrival of ethics committees and professional codes, cumulative replication, and changing social values about what counts as a disorder.
• Evaluate the story rather than narrating it, arguing that later work usually narrows the scope of earlier claims rather than refuting them, since conditioning effects themselves replicate reliably.
• Define social control as shaping behaviour towards standards set by an institution rather than by the person, then apply it: a token economy in a secure setting rewards compliance with the regime, which is not the same as recovery.
• Use evidence for the mechanism, quoting Ayllon and Azrin on the rise and collapse of ward performance with the contingency, and Hobbs and Holt (1976), whose token economy with institutionalised delinquent boys raised targeted behaviours relative to a comparison unit.
• Set out the power asymmetry explicitly: the therapist selects the target behaviour, holds the reinforcers, writes the progress record, and may be working with someone detained under mental health law who cannot simply withdraw.
• Give the counter-case for credit on both sides, since systematic desensitisation for a phobia is agreed with the client, uses a hierarchy the client builds, and has good outcome evidence, so control by consent is not the same as coercion.
• Tie the argument to free will and determinism, and to ethical implications, noting that behaviour that persists only while the rewards continue may serve the institution more than the person.
• Name the application and the mechanism together, for example a school merit system as positive reinforcement on a variable schedule, or a prison token economy where tokens are secondary reinforcers exchanged for primary ones such as visits or time out of cell.
• Support effectiveness with evidence, using Ayllon and Azrin on a psychiatric ward and Hobbs and Holt on institutionalised young offenders, and say what was actually measured, which was the frequency of specified target behaviours rather than long term change.
• Raise generalisation as the central weakness, since behaviour maintained by tokens tends to decline when the contingency ends, and reoffending figures after release show little effect from institutional programmes.
• Use the overjustification finding of Lepper, Greene and Nisbett to argue that rewarding something a child already enjoys can reduce intrinsic motivation, which matters for classroom reward charts.
• Judge the value, weighing cheapness, ease of staff training and clear measurement against loss of autonomy, inconsistent staff delivery and the risk that only the measured behaviour improves.
• Define socially sensitive research using the Sieber and Stanley sense, covering consequences for participants, for the wider group they represent and for policy, then apply it to a named piece of learning theories research.
• Give a therapy example with the power problem made explicit: clients approached by their own therapist may agree because refusing feels risky to their care, and clinical case material is hard to anonymise without losing the detail that made it worth reporting.
• Use Watson and Rayner as the historic case, noting deliberately induced distress, an infant who could neither consent nor withdraw, and no attempt to remove the conditioned fear before he left.
• Argue both directions, since Sieber and Stanley point out that refusing to study a sensitive question also has costs, and evidence on media effects, phobia treatment and eating disorders has improved practice and protected people.
• Bring in the safeguards and their limits, naming professional codes, ethics committees, cost benefit judgements made before data collection, and the fact that a committee cannot control how findings are reported afterwards.
• Distinguish the three ethical questions: harm caused by the diagnostic label itself, the difficulty of obtaining valid consent from unwell participants, and the professional obligations of a practitioner.
• Give labelling a mechanism, that a diagnosis becomes a master status reorganising how others read the person's behaviour, and evidence it with Rosenhan's finding that normal conduct was recorded as symptomatic.
• Set out what valid consent requires, information, capacity and voluntariness, and explain why detention under mental health law compromises the third of these.
• Name the regulator correctly for practitioners and give at least two of its requirements, such as confidentiality with defined limits and a duty to work within one's competence.
• Supply the counterweight, that a diagnosis also unlocks treatment, support and explanation, so an answer treating labelling as purely harmful is unbalanced.
• Define reliability as consistency and validity as measuring what was intended, then show a real clinical case where raising one lowers the other.
• Work an example on each side: a structured interview or symptom inventory for quantitative reliability, an open interview with thematic analysis for qualitative validity.
• Name the practical constraints of clinical samples, that participants come from those already in services, that medication and comorbidity vary between them, and that attrition over a long study is not random.
• Explain the control problems in treatment research, the need for a placebo or waiting list comparison, the double blind procedure, and the ethical limit on withholding an effective treatment.
• Reach a judgement, for instance that mixed methods use the quantitative measure for comparison and the qualitative material to interpret it, rather than declaring one kind of data better.
• Define the idea in a clause and name which version you mean: a biological version explaining a disorder by neurotransmitters or genes, an environmental version explaining it by conditioning, and holism insisting the whole experience is more than the sum of those parts.
• Attach a named example with its finding rather than the label alone: Carlsson and Lindqvist (1963) found antipsychotic drugs raised dopamine metabolite levels, which was the evidence that they block dopamine receptors and the origin of the hypothesis.
• State the strength precisely, that isolating one variable produces an operationalised hypothesis, a controlled trial and a treatment that can be evaluated, which is why this strategy is the one that generates therapies.
• State the cost with evidence, not as an opinion: Tienari et al. (2004) found that high genetic risk raised schizophrenia spectrum outcomes in Finnish adoptees mainly where the adoptive family was disturbed, so a purely biological account under predicts who becomes ill.
• Make the point about diagnosis explicitly, that counting criteria means two people carrying the same label can share very few experiences, which weakens the construct validity of the category.
• Name the systems accurately: the Diagnostic and Statistical Manual is published by the American Psychiatric Association and covers mental disorder, while the International Classification of Diseases is published by the World Health Organization and covers all illness.
• Give a real difference with its consequence, for example that the American criteria require six months of continuous disturbance for schizophrenia against one month in the international system, so prevalence figures from studies using different manuals are not directly comparable.
• Use evidence for the comparison rather than assertion: Cheniaux et al. (2009) showed the two manuals disagreeing on the same patients, which is a criterion validity problem for both.
• Credit the similarities as well: both systems are categorical, both were revised to drop the paranoid and catatonic subtypes, and both still depend on a clinician judging self reported experience.
• When comparing explanations rather than manuals, compare on a named basis such as testability, level of explanation, supporting evidence and the treatment each predicts, so the analysis is visible in the structure.
• List the features of science and tie each one to a clinical method, for example objectivity to structured scanning measures and control to a randomised trial with a placebo arm.
• Use falsifiability properly with an example: the original dopamine account predicted rapid improvement because receptor blockade happens within hours, yet symptom change takes weeks, and that mismatch forced the revised version rather than being ignored.
• Show replication as evidence of scientific status by naming a finding that has been repeated, such as raised concordance for schizophrenia in identical twins across separate twin samples.
• Argue the opposing case with data, not assertion: the field trials for the fifth edition of the American manual reported weak agreement between clinicians for major depressive disorder, so the measurement is not as objective as a thermometer.
• Handle the trade off in laboratory work, that the control which makes an experiment scientific also strips away the conditions in which the behaviour normally occurs.
• Distinguish a genuine difference in prevalence from bias in the process of diagnosis, and say which of the two your example is evidence for.
• Name the evidence for bias rather than asserting it: Ford and Widiger (1989) found clinicians reading identical symptoms diagnosed histrionic personality disorder more often when the case was female and antisocial personality disorder more often when the case was male.
• Use the imposed etic: criteria written in one culture and applied unchanged elsewhere treat that culture norms as universal, which is why culturally specific presentations of distress are read as symptoms.
• Offer more than one explanation for the higher recorded rate of depression in women, including greater help seeking, a clinician readiness to see distress in women, and real differences in exposure to stressors, and say what evidence would separate them.
• Draw out the ethical implication, that diagnosis rates feed into detention rates and access to services, so bias in the clinic has consequences well beyond the clinic.
• Define the predisposition and the trigger separately and state the prediction the model makes, that neither on its own is sufficient, which is what makes it an interactionist position rather than a compromise.
• Give the logic of twin research and then its flaw: identical twins share almost all their genes, so higher concordance suggests heredity, but they are also treated more alike, which is the equal environments assumption critics such as Joseph attack.
• Use adoption evidence because it separates the two sources, and report Tienari et al. (2004) as an interaction rather than as a genetic finding.
• Bring a social explanation with its own data: Brown and Harris (1978) interviewed women in Camberwell and found severe life events produced depression far more often in those carrying vulnerability factors such as early loss of a mother, several young children at home and no confiding relationship.
• Use epigenetics as the modern resolution, that environmental experience alters gene expression, so the two influences are not separable sources to be weighed against one another.
• Give at least two dated changes and say what each actually changed: operational criteria told clinicians what to count and raised agreement, while removing the bereavement exclusion widened who now counts as depressed.
• Say why the change happened, whether by new evidence, by professional vote, by campaigning or by reliability data, because the point of this debate is that classification is shaped by more than data.
• Track one treatment across time with its evidence: chlorpromazine, introduced in the 1950s, allowed long stay wards to empty but produced movement disorders such as tardive dyskinesia, which is exactly why atypical drugs were later developed.
• Turn the history into evaluation: a system revised this often has questionable construct validity, and a behaviour classified as a disorder in one edition and not the next was never a fixed natural category.
• Note the direction of travel in explanation, from possession and moral failing through psychodynamic accounts to biological and cognitive models and now to biopsychosocial ones, and say what each shift did to treatment.
• Define the concept and then name a specific mechanism: legal detention, treatment without consent, a community treatment order, or shaping behaviour by reinforcement on a ward.
• Argue the protective side as well, that compulsory admission prevents suicide and protects others, that most people detained are discharged, and that refusing to intervene has victims too.
• Use a historical case where control was clearly abuse, such as the nineteenth century diagnosis invented to explain enslaved people escaping, or the use of a schizophrenia diagnosis against political dissidents in the Soviet Union.
• Connect the argument to determinism: if behaviour is caused by neurotransmitters or by conditioning then the person is not responsible, which makes treating them without consent far easier to justify, and that is why this debate is argued alongside free will.
• Use Rosenhan (1973) for the loss of agency after labelling, including discharge with the qualifier in remission, which never lets the label go.
• Say what the application is used for, not only what it is: family intervention is used to lower relapse in schizophrenia by reducing criticism and emotional over involvement in the home.
• Quote effectiveness evidence with its shape, for example that Vaughn and Leff found around half of patients returning to high expressed emotion homes relapsed against a small minority of those returning to low expressed emotion homes.
• Say what the treatment does not do: drugs manage symptoms rather than curing, relapse on discontinuation is common, and cognitive therapy is hard to deliver to someone acutely psychotic or lacking insight.
• Include the economic and social argument, that treating depression returns people to work and study, which is the case made when talking therapy services were expanded in the health service.
• Note who is left out, since access varies by region and by group, so effectiveness inside a trial is not the same as effectiveness across a population.
• Separate an ethical issue from an ethical implication: the issue is consent or deception inside the study, whereas the implication is what the published finding does to the group afterwards.
• Use the four areas identified by Sieber and Stanley as a structure, because it gives the answer a shape the examiner can follow and covers method and application together.
• Give a worked example with both sides: Rosenhan (1973) could not have been run with informed staff, and the attention it brought to reliability of diagnosis is a benefit that would not otherwise exist.
• Discuss a genetic example, since publishing heritability estimates invites screening and encourages the belief that carriers are destined to become ill, which is a determinist misreading with real consequences for insurance and employment.
• Argue that avoiding the sensitive question is itself a cost, because differences in diagnosis and detention rates between groups would be invisible if nobody counted them.
• Separate the ethics of the procedure from the ethics of the use, since marks come from showing both the harm and consent issues for participants and the social consequences of publishing a finding about juries or witnesses.
• Name a study and the specific breach, so a staged theft in a lecture offers no informed consent and no right to withdraw before the event, and debriefing reaches only the people who stay behind.
• Loftus and Palmer (1974) is the low cost laboratory alternative worth citing, where changing one verb in a question raised the estimated speed and led more participants to report broken glass that was never in the film, with consent and no risk of harm.
• Argue the cost and the benefit both ways, because the same body of findings produced the cognitive interview and the judicial warning given to juries about identification evidence.
• Use the vocabulary of socially sensitive research, that the risk lies in how findings are applied to real defendants and witnesses rather than in the testing session itself.
• Explain why the artificiality exists rather than simply noting it, that real deliberation cannot lawfully be observed and defendants cannot be randomly allocated to conditions.
• Name the specific losses, no deliberation, no consequence for a real person, written rather than live evidence, and samples that are younger and more educated than a real jury panel.
• Balance it with what control buys, since a mock jury lets the defendant's accent or appearance be varied while everything else is held constant, which no real trial would permit.
• Credit the partial solutions, real jurors working on simulated cases, filmed rather than written trials, and requiring deliberation before verdicts are recorded.
• Tie the point to generalisability, and say what further evidence would be needed before a laboratory finding could justify changing court procedure.
• Distinguish experimental reductionism, isolating one variable under control, from biological reductionism, explaining conduct at the level of brain, genes and hormones, and give a criminological example of each.
• Use Raine and colleagues with the method attached, a comparison of murderers pleading not guilty by reason of insanity with matched controls, reduced prefrontal activity and abnormal subcortical asymmetries, and no claim that the scan caused the killing.
• Give the holistic counterweight, that Caspi and colleagues found the low activity variant of the monoamine oxidase gene raised antisocial outcomes mainly in men who had been maltreated, so the lower level only explains anything alongside the higher one.
• Say what reductionism buys, measurement, testability and the possibility of a targeted treatment, because an answer that only attacks it has argued one side.
• Applied to weapon focus, credit the point that the laboratory version measures attention to an object while a real armed robbery adds fear, duration and consequence, so the reduced version may not be the same phenomenon at all.
• Compare on stated dimensions, the evidence each rests on, the level of explanation, how deterministic each is and what intervention each implies, rather than writing two descriptions with a sentence of comparison at the end.
• Name the pairings, biological with Raine and colleagues or with Brunner and colleagues on the monoamine oxidase deficient family, learning with Bandura and with Sutherland's differential association, social with group processes and with Farrington's longitudinal risk factors.
• Show the similarities as well as the differences, since all three are deterministic and nomothetic, seeking general laws about offenders rather than an account of one life.
• Use treatment as the decisive comparison, because a biological account points towards medication, a learning account towards token economies and modelling, and a social account towards changing groups, housing and opportunity.
• Reach a judgement the evidence supports, usually that a diathesis stress or interactionist position fits the adoption and gene environment findings better than any single theme on its own.
• Naming the criteria of science and tying each to a method: objectivity and standardised procedure in the laboratory, falsifiable hypotheses, replicability shown by repeated misinformation studies, and empirical brain data from imaging.
• Loftus and Palmer (1974) reported that the leading verb changed mean speed estimates, about 34.0 mph for hit against about 40.8 mph for smashed, and raised false reports of broken glass a week later, because only the verb was manipulated.
• Field experiments manipulate an independent variable in a natural setting, which increases ecological validity but reduces control over confounding variables, contrasting with natural experiments where the independent variable is merely measured.
• Raine et al. (1997) provides a biological and objective example: positron emission tomography of murderers pleading not guilty by reason of insanity showed reduced prefrontal activity, which is measurable and replicable but correlational, so it cannot show that the brain difference caused the killing.
• Credit for naming the bias precisely, accent and ethnicity as culture linked cues and gender as expectation about who commits which offence, rather than a vague claim that juries are prejudiced.
• Dixon et al. (2002): the same recorded interview with accent manipulated produced higher guilt ratings for the Birmingham accent, and a larger gap when the defendant was black and the charge was armed robbery.
• Pfeifer and Ogloff (1991): racial bias in guilt ratings appeared with identical evidence and shrank when jurors were given an explicit instruction on the burden of proof, which is the practical remedy examiners like to see.
• Stewart (1980) used real trials rather than mock juries, so the finding that more attractive defendants received lighter sentences has ecological validity but no control over the evidence each jury heard.
• Evaluation credit for the limits of mock jury research: student samples, no deliberation, written or taped vignettes and no real consequence for the defendant, so the findings may overstate or understate courtroom bias.
• Credit for framing the debate as an interaction, using diathesis stress to say that an inherited vulnerability is expressed only under particular rearing conditions.
• Mednick et al. (1984): a Danish adoption cohort in which conviction rates were highest when biological parents had convictions and lowest when neither set had, which separates genes from rearing better than a twin study can.
• Brunner et al. (1993): one extended family with a mutation in the gene for monoamine oxidase A and a history of impulsive aggression, showing a genetic route on a sample far too small and specific to generalise from.
• Bandura and Sutherland for the nurture side: imitation of a reinforced model, and the learning of criminal techniques, motives and justifications from those closest to the offender.
• Evaluation credit for Caspi et al. (2002) as evidence that neither pole is sufficient, and for the point that genetic accounts of offending are determinist and raise hard questions about responsibility and punishment.
• Credit for showing a sequence, from laboratory demonstration of misinformation to false memory implantation and then to courtroom and policy application, rather than three unconnected studies.
• Loftus and Palmer (1974) and Loftus and Pickrell (1995) with findings: the verb changed estimated speed and later reports of glass, and about one participant in four came to report a childhood event that never happened.
• The four cognitive interview components named and explained, with the Miami field study showing a large rise in correct information recalled by trained detectives.
• The ethical interview described as a shift from securing a confession to gathering an account, through planning, engage and explain, account, closure and evaluation, and linked to miscarriages of justice involving false confessions.
• Evaluation credit for noting that the cognitive interview raises correct detail but can raise incorrect detail slightly too, that it takes time and training forces do not always fund, and that few studies follow through to conviction accuracy.
• Credit for defining social control as the use of psychological technique to change behaviour in the interests of an institution, and for noting that consent given in the hope of earlier release is not freely given.
• Token economy explained in operant terms, with secondary reinforcers exchanged for primary reinforcers, target behaviours chosen by staff, and Hobbs and Holt (1976) as evidence of behaviour change within the institution.
• Anger management outlined through Novaco's three stages with Ireland (2004) as evidence, plus the honest limitation that the outcome measures were self report and officer report over a short follow up.
• Evaluation credit for identifying who holds the power: the forensic psychologist writes the parole report, the therapist defines progress, and the offender who refuses treatment may serve longer.
• Credit for saying what these programmes do not treat, since they leave the poverty, addiction and unstable housing an offender returns to untouched, so effects often fail to generalise after release.
• Credit for naming the applications rather than only the findings: video identification procedures, Turnbull warnings to juries, expert testimony on memory, and police training in the cognitive interview.
• Weapon focus evidenced through Loftus, Loftus and Messo (1987), whose eye tracking showed more and longer fixations on the weapon and poorer identification of the person holding it.
• Jury applications: a clear reasonable doubt instruction reduces measured bias, whereas telling jurors to disregard inadmissible material can backfire, so the recommendation is to keep the material out.
• Evaluation credit for the tension between laboratory evidence and courtroom practice, since courts may exclude expert testimony on memory as being within the ordinary experience of a jury.
• Credit for weighing the social benefit against the cost, since safeguards protect the innocent but may make genuine convictions harder, which is a judgement about values rather than about data.
• Credit for using Sieber and Stanley's framework by name and applying at least two of their concerns to a criminological example rather than listing the four abstractly.
• A worked example with its consequence, such as the extra chromosome claim producing stigma and courtroom defences that the later evidence did not support.
• The measurement point that arrest and conviction statistics record the responses of the justice system as well as behaviour, so socio-economic status, ethnicity and age effects are confounded with policing and sentencing practice.
• Evaluation credit for the cost and benefit argument, since refusing to study these variables leaves discrimination undetected, so the ethical answer is careful design, honest reporting and control over how findings are publicised.
• Credit for linking to practice through confidentiality, protection from harm, and the researcher's responsibility for the interpretation placed on findings by press and policy makers.
• Name the convention in full, the United Nations Convention on the Rights of the Child, and pin the tension to specific articles: the participation right to be heard and taken seriously against the protection right to have best interests treated as paramount.
• Credit comes for the machinery, not the sentiment: parental or guardian consent alongside the child's own assent, presumptive consent where assent is impossible, an observed right to withdraw, debriefing pitched at the child's level, and referral if what a child says suggests harm.
• Apply it to a named piece of child research. Ainsworth's Strange Situation deliberately induces mild distress, and the protocol curtails a separation episode early if the infant becomes too upset, which is protection built into the design rather than bolted on afterwards.
• Curtiss's research with Genie is the standard example of protection failing: the same people acted as both researchers and carers, testing ran alongside therapy, and when funding ended she was moved into care where she regressed.
• Say who decides. The cost and benefit judgement is made in advance by an ethics committee against the British Psychological Society code, and the benefit claimed has to be one that could not be gained from adult participants.
• For a meta-analysis, credit the gain first: pooling lifts the total sample far beyond any single study and lets a difference between countries be tested rather than asserted.
• Then name the comparability problems in the board's own terms: different samples, different age ranges and different coders, unequal numbers of studies per country, and the greater likelihood that significant findings reached publication at all.
• For observation, the marks sit in the method: operationalised categories such as proximity seeking, contact maintaining, separation anxiety, stranger anxiety and reunion behaviour, rated by trained observers behind a one-way mirror, with inter-rater reliability calculated so that the coding is not one person's opinion.
• Say why standardisation matters. Ainsworth's procedure runs to a fixed sequence of episodes of fixed length, which is what makes replication and cross-cultural comparison possible at all.
• A strong answer links the practical issue to validity: a controlled observation in an unfamiliar room has high control and low mundane realism, so part of what is recorded may be a reaction to the setting.
• Define reductionism in a clause and then argue: a complex relationship is operationalised as behaviour in a brief structured procedure, and the classification is driven largely by reunion behaviour.
• Give the payoff, because reductionism is not automatically a criticism. Reducing attachment to codeable behaviour is what made the measure reliable, replicable and usable in cross-cultural meta-analyses.
• Give the cost with evidence. Main and Weston found infants classified differently with mother and with father, so what is measured is one relationship rather than a fixed property of the child.
• Bring in the competing reduction. Kagan's temperament hypothesis argues that innate irritability and sociability, rather than caregiver sensitivity, drive the classification, which reduces attachment to biology instead of to interaction.
• Set holism against it: an interactionist account keeps caregiver sensitivity, temperament, family stress and cultural child rearing in one explanation, and says which of them the procedure cannot see.
• Compare on stated grounds rather than in parallel paragraphs: the question each theory answers, whether the cause offered is innate or interactional, the evidence base, and the predictions each makes about adolescence and adulthood.
• Show how they interlock. The sensitivity hypothesis supplies the mechanism that monotropy leaves vague, and the classification gives the continuity hypothesis something measurable to predict.
• Use animal evidence for the evolutionary theme accurately. Lorenz found greylag goslings followed the first moving thing they saw within a narrow window, and Harlow found rhesus infants clung to a cloth surrogate that provided contact but no food.
• Quote Bowlby's own study properly: among the forty four juvenile thieves, the affectionless group was far more likely than the other delinquents to have had prolonged early separation, but the data were retrospective and were gathered by Bowlby himself.
• Bring in the rival theme. Learning theory explains attachment through association and reinforcement and Kagan explains it through temperament, so the comparison has three positions rather than two.
• State the criteria before applying them, namely objectivity, control, standardisation, replicability, hypothesis testing and falsifiability, then show where the research meets each one.
• Explain the logic of the cross-cultural argument. If secure attachment is the majority pattern in every culture sampled, it is unlikely to be the product of any one child rearing practice, which is evidence for an evolved mechanism.
• Balance it with the imposed etic problem using the Japanese data: infants rarely separated from their mothers showed extreme distress, episodes were stopped early, and almost none were classified avoidant, so the same behaviour can mean different things in different settings.
• Use falsifiability with a real example. The claim of a critical period was testable and was undermined by later adoption studies showing recovery, so the account was revised to a sensitive period.
• Credit the limits of the scientific model itself: much child research is observational rather than experimental, causation cannot be read from a correlation between sensitivity and security, and psychology lacks the single agreed paradigm that a mature science is said to have.
• Distinguish cross-cultural findings from culture bias. The first is data about how children are raised; the second is the claim that a measure carries one culture's values into every sample it visits.
• Use the country comparisons precisely: the highest avoidant rates came from the German samples, the highest resistant rates from Japanese and Israeli samples, and variation between studies within a country exceeded the variation between countries.
• Define the terms a marker looks for, namely ethnocentrism, imposed etic, emic, and the individualist and collectivist distinction, each attached to a finding rather than left floating.
• For gender, argue both ways. Day care studies report outcomes separately for boys and girls, and the maternal deprivation tradition focused on mothers so heavily that fathers were written out of theory and of policy.
• Note that Schaffer and Emerson found most infants had formed an attachment to the father within months of forming their first attachment, which is direct evidence against a mother only account.
• Set out what each side predicts before judging. Nature predicts the same pattern everywhere, nurture predicts patterns that follow local practice, and the cross-cultural data contain both results at once.
• Use institutional evidence for the nurture side. Romanian adoptees followed by Rutter and colleagues showed that those adopted early largely caught up with home adopted children, while later adoption was followed by lasting difficulties and by disinhibited attachment.
• Credit an interactionist mechanism rather than a compromise sentence. Differential susceptibility holds that some infants are strongly affected by the quality of care for better and for worse, which explains why the same arrangement produces different outcomes.
• Say what the evidence cannot do. Cross-cultural comparison is quasi-experimental, adoption studies confound age at adoption with time spent in the institution, and neither separates heredity from environment cleanly.
• Tie the debate to determinism. A strong biological determinist reading makes genetics destiny, whereas environmental determinism makes early experience destiny; the recovery seen after early adoption is evidence against environmental determinism.
• Trace the sequence rather than the names alone: an early claim that separation from the main caregiver causes lasting damage, a critique that distinguished deprivation from privation, and later research that tested day care outcomes directly.
• Quote the critique accurately. Rutter argued that the original evidence lumped together separation, loss and never having formed an attachment, and that later difficulties often followed family discord rather than absence in itself.
• Use day care evidence on both sides. Belsky and Rovine reported more insecure avoidant classifications where infants had extensive non-maternal care during the first year, while large longitudinal projects found that high quality provision improved cognitive and social outcomes, particularly for disadvantaged children.
• Say what changed in practice. The Robertson films of children in hospital led to open visiting and to parents staying overnight, which is a direct line from research to policy.
• Note the methodological improvement over time: retrospective interviews and single cases gave way to longitudinal cohorts with measured quality of care and controls for family background.
• Define social control as the use of psychological findings to shape behaviour and policy, then show the two directions the day care evidence has been pushed, towards provision and away from it, depending on economic need.
• Use autism interventions concretely. Lovaas reported that a substantial minority of children given intensive early behavioural intervention reached typical educational functioning, which is why the approach spread, while critics including autistic adults argue that suppressing repetitive behaviour serves the observer and can cause distress.
• Contrast approaches. Picture exchange communication systems and structured teaching approaches aim at access and independence rather than at appearing typical, so they meet part of that objection.
• Bring in socially sensitive research: the question is not only whether a study was conducted ethically but who is affected by its conclusions, and mothers, autistic children and low income families are the groups affected here.
• Use the discredited vaccine claim as the standard example of harm through use, since a retracted paper changed parental behaviour and public health outcomes long after it was withdrawn.
• Naming an application with the study behind it, for instance that applied behaviour analysis uses reinforcement and discrete trial training to build imitation and language in autistic children, and that Lovaas (1987) found forty-seven per cent of his intensive group achieved normal intellectual and educational functioning against one child in the control group.
• Turning day care research into advice: the Effective Provision of Pre-School Education project led by Sylva and colleagues found that high quality pre-school raised cognitive and social outcomes and helped disadvantaged children most, while the NICHD Study of Early Child Care linked long hours of group care begun before the age of two to a small rise in later anti-social behaviour, which supports advice about quality, ratios and a consistent key worker rather than advice to avoid day care.
• Turning research on looked after children into advice: Rutter and the English and Romanian Adoptees team found children adopted before six months had largely caught up in cognition and attachment by age four, while those adopted later still showed disinhibited attachment and quasi-autistic features years on, which supports early permanent placement, few placement moves and attachment-based training for foster carers.
• Weighing the usefulness against the cost, so credit for saying that intensive behavioural programmes are expensive and demanding of parents, and that advice drawn from group averages may not fit an individual child.
• Tying the content to the debates by name, so real-world usefulness on one side and ethical implications and socially sensitive conclusions about working mothers on the other.
• A definition in a clause plus the point that matters, which is that participants can be protected perfectly and the group they belong to still be harmed by the conclusion drawn.
• The autism example handled accurately: the Wakefield paper of 1998 rested on twelve children, had no comparison group, was retracted by the journal in 2010 and the author was removed from the medical register, and its social cost was measured in vaccine hesitancy and in autistic people framed as damaged.
• The socio-economic example: Hart and Risley followed forty-two families and reported a very large gap in the number of words children heard by age three, a finding used to justify early intervention but criticised for its small unrepresentative sample and for framing poorer parents as deficient.
• The privation example: Rutter and colleagues followed Romanian adoptees for decades with consent from adoptive parents rather than the children, and case studies such as Curtiss on Genie show research access and the duty of care pulling against each other.
• Applying a framework rather than listing worries, so risk against benefit, anonymity, wording of conclusions, consultation with the community studied and the professional code of ethics, ending with the argument that not researching autism or adoption also carries a cost.
• Credit for naming the framework rather than the feeling: the Animals (Scientific Procedures) Act 1986 requires project, personal and establishment licences, the three Rs govern design, and Bateson's decision cube sets quality of research against animal suffering and certainty of benefit.
• Credit for the detail of Olds and Milner (1954), where rats with electrodes in the septal area self-stimulated at very high rates and some neglected food to do so, while other placements produced no such effect, so real suffering bought the mapping of a reward pathway.
• Credit for British Psychological Society principles applied to drug research: valid consent is hard to obtain from someone dependent or withdrawing, confidentiality matters more than usual when data concern illegal acts, and withholding treatment from a control group needs justifying, which is why Pengpid and colleagues (2012) gave their control group health education material rather than nothing.
• Credit for evaluating aversion therapy on ethical grounds, since pairing alcohol with an emetic or with shock distresses by design, consent is questionable where courts, employers or families apply pressure, and the same technique was once used in attempts to change sexual orientation, which shows how a treatment can become an instrument of social control.
• Credit for a cost and benefit conclusion rather than a verdict: animal studies allow controlled work with drugs that could not ethically be given to people, and the argument turns on whether the treatments that follow justify the harm.
• Credit for concrete species differences rather than a slogan: brain structure, receptor distribution and metabolism differ, and human drug taking sits inside expectation, social meaning and law that no cage contains.
• Credit for Alexander and colleagues finding that housing conditions changed morphine consumption dramatically in Rat Park, which shows an animal result can be an artefact of the procedure rather than a truth about addiction.
• Credit for explaining why the brain resists holistic study: positron emission tomography needs a radioactive tracer so repeated scanning is limited, functional imaging is indirect and slow relative to synaptic events, and the reward pathway cannot be isolated from the systems it works alongside.
• Credit for practical problems in human designs, including obtaining a representative sample of users, poly-drug use as a confounding variable, attrition in longitudinal work, and the impossibility of randomly allocating people to years of heroin use.
• Credit for following the practical issue through to a consequence: a result from male laboratory rats on a single dose supports a claim about mechanism, not a claim about why a particular person relapses.
• Credit for distinguishing levels of explanation, since the same behaviour can be described as receptor binding, as reinforced responding, or as a person living among particular peers, prices and pressures.
• Credit for what the reduction buys: isolating variables permits experimental control and causal inference, makes findings replicable, and yielded usable treatments such as nicotine replacement, methadone and naltrexone.
• Credit for the costs: isolating a variable removes the social context that field evidence says matters, so results from singly housed animals may not describe people, and treating addiction purely as brain chemistry sidelines poverty, trauma and availability.
• Credit for linking reductionism to determinism, because if dependency is receptor adaptation then the user's scope for choice shrinks, which changes how courts and health services respond to relapse.
• Credit for the compromise position, a biopsychosocial or diathesis stress account in which inherited vulnerability sets the risk while learning and environment decide whether it is realised.
• Credit for comparing on explicit criteria such as determinism, reductionism, quality of evidence and treatment implications, and for offering a similarity as well as a difference.
• Credit for accurate learning detail: cue reactivity and conditioned withdrawal, positive reinforcement from euphoria and negative reinforcement from relief, and vicarious reinforcement through peers and media, with Mundt and colleagues (2012) reporting both peer selection and peer influence, so friends resemble one another partly because drinkers choose drinkers.
• Credit for accurate biological detail: twin and adoption evidence for inherited vulnerability to alcohol dependence, nicotine acting at receptors in the ventral tegmental area and raising dopamine in the nucleus accumbens, heroin acting at opioid receptors, and tolerance as receptor adaptation.
• Credit for what each account cannot do, since learning theory struggles to explain why only a minority exposed to the same cues become dependent, while biology struggles with sharp changes in use across cohorts, cultures and prices.
• Credit for the treatment comparison, because learning accounts predict aversion therapy, cue exposure and contingency management, biological accounts predict substitution and blocking drugs such as methadone, nicotine replacement and naltrexone, and relapse rates suggest that neither alone is sufficient.
• Credit for naming criteria and testing the topic against them, rather than asserting that psychology either is or is not a science.
• Credit for Olds and Milner (1954) as the worked example of scientific method: electrode placement as the manipulated variable, rate of lever pressing as an objective measure, a standardised procedure, and findings others were able to repeat.
• Credit for the shared decision rule of inferential testing, where a result counts as significant only when the observed value reaches the critical value at an agreed level, usually p less than 0.05, which gives objectivity of a kind while also encouraging publication bias.
• Credit for the counter evidence, since Dixit and colleagues (2012) and Mundt and colleagues (2012) are correlational, so no causal claim follows, and self-reported drug use is open to social desirability and recall error.
• Credit for falsifiability, because an explanation elastic enough to accommodate any pattern of use is not being tested by that pattern, alongside the reply that predictions about receptor action are testable and have been tested.
• Credit for the trade-off between control and ecological validity, which is why a laboratory science of addiction can be rigorous and incomplete at the same time.
• Credit for using the cross-cultural study properly, since Dixit and colleagues (2012) compared urban and rural communities within one country, so it shows variation inside a nation as well as between nations, and it is a survey, which makes the links associations rather than causes.
• Credit for the methodological problem of an imposed etic, where a screening questionnaire developed in one culture is translated and used in another, so the same score may not mean the same drinking.
• Credit for reading cultural difference through nature and nurture, because differing rates can reflect availability, price, religion and law as well as inherited differences such as the alcohol-metabolising enzyme variant.
• Credit for building the gender argument rather than asserting it: alpha and beta bias, male-only animal samples, differences in who seeks treatment, and the extra stigma attached to drug use in pregnancy, which makes honest self-report less likely.
• Credit for evaluating ethnocentrism in campaigns and treatments, since an anti-drug campaign designed for one culture may fail in another that does not share its assumptions about family, authority or shame.
• Credit for specific evidence on the nature side: higher concordance for alcohol dependence in identical than in non-identical twins, adoption findings such as those of Goodwin and colleagues, and inherited differences in receptor sensitivity or in how quickly alcohol is metabolised.
• Credit for specific evidence on the nurture side: conditioned responses to drug cues, positive and negative reinforcement, social learning from parents and peers, and the effect of price, availability and law on how much is used.
• Credit for noting that concordance in identical twins is well short of complete, which is itself evidence that environment contributes, and for the shared environment confound that makes twin estimates uncertain.
• Credit for an interactionist conclusion stated with a mechanism, whether diathesis stress, gene and environment correlation, or epigenetic change following early stress, rather than the bare claim that it is a bit of both.
• Credit for using Robins on American servicemen returning from Vietnam, where most of those who had used heroin did not remain addicted at home, as evidence that changing the environment changes the outcome.
• Credit comes from naming a sequence and anchoring it: moral and criminal framing, then the reward pathway work of Olds and Milner (1954), then learning explanations built on classical conditioning, operant reinforcement and social learning, then the biopsychosocial position most services now assume.
• Saying what each piece of research actually found rather than what it is famous for, for example that the rats in Rat Park, housed socially in an enriched enclosure, drank far less morphine solution than isolated caged rats, which moved explanation away from the drug itself and towards the environment.
• Linking a change in explanation to a change in practice: reward pathway models justify substitution such as nicotine replacement or methadone, conditioning models justify cue exposure, plain packaging and removing paraphernalia, and social learning models justify credible peer models in campaigns.
• Evaluating the development instead of narrating it, including the finding from evaluations of Project DARE by Ennett and colleagues that the programme shifted knowledge and attitudes but had almost no effect on drug use, so better explanations have not automatically produced better campaigns.
• Separating control that a person chooses from control imposed on them, and naming who holds the power in each case, because the mark scheme rewards the distinction rather than the examples alone.
• Naming the psychology behind each method: token economies and contingency management come from operant conditioning, aversion therapy from classical conditioning, and substitute prescribing from the dopamine reward account of dependence.
• Giving evidence on both sides, such as the retention and reduced illicit use reported for methadone maintenance, and the finding of Paul and Lentz (1977) that a social learning and token programme outperformed ordinary hospital care, set against high relapse once a regime is withdrawn.
• Turning the example into a debate about free will and determinism and about ethical implications, including whether valid consent is possible when the alternative is prison, and whether policing falls unevenly on some communities.
• Naming the application and the explanation it grew out of, so the marker can see that substitute prescribing follows the biological account while contingency management follows operant conditioning.
• Giving effectiveness evidence with a direction and a comparison, for example that nicotine replacement raises quit rates against placebo, and that voucher programmes beat standard counselling on continuous abstinence.
• Saying what the treatment does not do, including that substitution manages dependence rather than curing it, that aversion therapy has poor long term follow up, and that none of these address the conditions people return to.
• Adding an issues point, usually reductionism, since explaining a habit purely at the level of receptors makes the person and their circumstances disappear from the treatment plan.
• Using the definition of research whose consequences reach beyond the participants, and applying it to this topic rather than reciting it.
• Separating issues at the level of the individual, which are consent, confidentiality, protection from harm and the right to withdraw, from consequences at the level of the group, which are labelling, policy and the withdrawal of services.
• Saying what a researcher actually does about it, such as anonymised coding of responses, ethical approval before recruitment, working to the British Psychological Society code of human research ethics, and offering routes to support at debrief.
• Reaching a balanced judgement, since refusing to study a vulnerable group means services are designed without evidence about them, so sensitivity is a reason for care rather than for avoidance.
• Precision is credited, so distinguish an ethical issue, which is the conflict between the rights of participants and the needs of the research, from an ethical guideline, which is the written response to that conflict.
• Marks come from tying a guideline to what a study actually did, for example that Milgram withheld the nature of the shocks so consent was not informed, and that his extended debriefing and reunion with the learner was a partial remedy.
• Cost benefit reasoning earns evaluation credit when the cost is named as harm to a particular participant or animal and the benefit as a specific piece of knowledge, not as science in general.
• For animal work, credit follows knowing that the legal control is licensing under the Animals (Scientific Procedures) Act and that the three principles are replacement, reduction and refinement.
• Socially sensitive research is credited where the consequences beyond the laboratory are named, such as findings on group differences being used to support policy the researchers never intended.
• Credit follows a stated operational definition, so define the measure as it would be recorded, for example imitative acts counted in five second intervals, rather than as the abstract idea.
• Naming the design and the problem it solves earns marks, such as repeated measures controlling participant variables while introducing order effects that counterbalancing then addresses.
• Extraneous and confounding variables are credited when separated: an extraneous variable adds noise, a confounding one varies systematically with the conditions and so offers a rival explanation of the results.
• A strong answer states the trade-off explicitly, that tight laboratory control buys a causal conclusion at the price of mundane realism, while a field setting buys realism and loses control.
• Marks are available for practical remedies, including a pilot study to test materials and timings, standardised instructions, single or double blind procedures and inter-rater checks on observation.
• Marks come from naming the level of explanation being used and the level being ignored, rather than from asserting that an account is reductionist and leaving it there.
• Distinguishing the types earns credit: biological reduction to genes and neurochemistry, environmental reduction to learned stimulus and response links, and experimental reduction to variables that can be manipulated in a laboratory.
• The strongest evaluation is that reduction is what makes hypotheses testable and has produced working treatments, so the drug evidence should be used as the argument for it rather than only against it.
• Balance is credited by giving holism its own difficulty, that whole person accounts such as humanistic ones are hard to falsify and rarely specify a clear intervention.
• An interactionist conclusion scores well when it is specific, for instance that a genetic vulnerability only becomes illness in the presence of a particular environmental stressor.
• A comparison earns credit when it runs on a stated dimension, such as how deterministic each account is, how far each relies on nature or on nurture, or what type of evidence supports it.
• Similarities count as well as differences, for example that social learning and situational explanations both make behaviour environmental and therefore both predict that changing the setting changes the behaviour.
• Naming the evidence base for each side is what separates comparison from assertion, so a laboratory experiment, a field study and a brain imaging study should be identified as such with their findings.
• Application is credited by drawing out what each explanation implies for intervention, since a biological account points to medication and a learning account points to modelling and reinforcement programmes.
• A conclusion that neither account is sufficient scores only if it says what a combined explanation would look like and what evidence supports the combination.
• Credit follows naming a specific feature of science and then testing one study against it, rather than listing features and asserting that the subject qualifies.
• Falsifiability is best illustrated with a claim that cannot fail, so an explanation which accounts equally well for a behaviour and for its opposite is shown to be unscientific by that fact.
• Replicability earns marks when tied to standardisation, since identical instructions and materials are what make an independent repetition possible and a failure to replicate meaningful.
• The strongest evaluation is that unobservable processes are not automatically unscientific, because inference from measurable behaviour to an internal model is what the cognitive area does and what physics does too.
• Marks are available for noting that qualitative and idiographic work rejects the natural science model deliberately rather than failing to reach it, seeking meaning rather than general laws.
• The terms must be used accurately, so state that alpha bias overstates a difference, beta bias ignores one, and androcentrism treats male behaviour as the standard against which the other is judged.
• Credit follows an example that names the sample, for example that Asch tested only American male students and Milgram only men, so conclusions about conformity and obedience were generalised beyond their evidence.
• The distinction between an emic account, built from within a culture, and an etic one, applied from outside, earns marks when used to explain why a translated questionnaire can still be biased.
• Evaluation is strengthened by noting the practical consequences, such as diagnostic criteria that do not fit culturally normal experiences contributing to unequal rates of diagnosis between ethnic groups.
• Cultural relativism is credited when balanced, since insisting that every behaviour is only judgeable within its own culture makes harmful practices hard to criticise at all.
• Marks follow correct use of concordance, which is the probability that the second twin shows the trait given that the first does, and not a percentage of behaviour caused by genes.
• The diathesis stress model earns credit when both halves are named, an inherited or biological vulnerability and a specific environmental trigger that has to occur for the disorder to appear.
• Explaining why identical twins are not fully concordant is a high value point, since it shows directly that inheritance cannot be the whole account.
• Gene environment interaction is best evidenced by a study in which the genetic effect appears only under a particular environment, which is what the maltreatment and monoamine oxidase finding shows.
• Credit is available for the methodological caution that identical twins are often treated more alike than fraternal twins, so the equal environments assumption inflates apparent heritability.
• Credit comes for naming a stage and dating it: Wundt and structuralism in 1879, Freud's case study method, Watson's behaviourism from 1913, the cognitive revolution from the mid 1950s, and cognitive neuroscience from the 1980s onwards.
• Markers reward a stated change in method rather than a change in topic, so say that introspection gave way to controlled observation of behaviour, then to mental processes inferred from reaction times, then to scanning that images the process directly.
• Say what each named study actually found: Watson and Rayner produced a conditioned fear response to a white rat that generalised to other furry stimuli, and Miller found the span could be extended by chunking rather than by adding items.
• Top band answers judge the change rather than list it, arguing whether psychology has had a single paradigm shift or whether several competing paradigms run alongside each other.
• Name the mechanism, not just the practice: a token economy is operant conditioning, using secondary reinforcers exchanged for primary reinforcers and shaping by successive approximations.
• Say who is being controlled and by whom, since marks come from identifying the institution, prison, hospital, school, advertiser or government, and the behaviour it wants changed.
• Evidence of effectiveness earns credit: institutional behaviour improves while the programme runs, but Ayllon and Azrin's reversal design showed behaviour dropping back when tokens were withdrawn, and gains rarely generalise beyond the ward.
• Evaluation should reach the free will and determinism debate, because social control assumes behaviour is environmentally determined and therefore modifiable by controlling the environment.
• Every application needs a named source study and a named change in practice: eyewitness research to the cognitive interview, attachment research to hospital and nursery arrangements, cognitive theory to cognitive behavioural therapy.
• Say what the evidence for effectiveness is, for example that Geiselman and colleagues found the cognitive interview yields more correct detail than a standard police interview, while reviews also report more incorrect detail.
• Economic implications are creditworthy: effective therapy returns people to work and better interviewing reduces wrongful convictions and the cost of appeals.
• Strong answers question the transfer from laboratory to society, noting that Loftus and Palmer used film clips and student participants without the emotional stakes of witnessing a real crime.
• Give the definition in a clause and then the four aspects identified by Sieber and Stanley, attaching a real example to at least one of them.
• Name the group that could be harmed, not just the participants, because the argument is that an entire category of people can be stigmatised by a finding collected under perfect ethical procedure.
• Balance is rewarded: refusing to study sensitive topics leaves prejudice untested, so argue for doing the research carefully rather than for banning it.
• Link to determinism and to the nature and nurture debate, since genetic or brain based findings about crime and intelligence are socially sensitive precisely because they are read as fixed.

• 💡Issues and debates have their own extended essay section on Paper 3, and they also belong inside evaluation paragraphs on Papers 1 and 2, so rehearse them as paragraphs rather than as lists.
• 💡The question usually names one debate and leaves the content to you, so prepare two social psychology examples for each debate on the list.
• 💡Extended answers carry marks for evaluation as well as knowledge, so plan the judgement before writing the first paragraph.
• 💡Memorise the four BPS principles (respect, competence, responsibility, integrity) as they frequently appear as distractors in multiple-choice questions.
• 💡Be prepared to identify specific ethical safeguards used in contemporary studies, such as Burger's two-stage clinical screening.
• 💡When asked to identify an implication of prejudice research, look for options discussing societal impact or stigmatisation rather than participant consent.
• 💡Be able to distinguish between structured, semi-structured, and unstructured interviews based on their definitions and practical constraints.
• 💡Remember that a pilot study is a small-scale trial run used to test and refine practical aspects of research design, such as question clarity.
• 💡Identify interviewer effects as a specific practical issue in interviews that is less prominent in anonymous questionnaires.
• 💡Extended questions often name the debate and leave the content open, so keep one clearly reductionist and one clearly holistic social theory ready.
• 💡Write a judgement rather than a balance sheet: say whether the reduction bought useful prediction or lost the behaviour.
• 💡Where the wording mentions drawing conclusions from data, discuss how an aggregate percentage hides the variation between individuals.
• 💡Comparison answers reward a running structure, so take one criterion at a time and cover both theories inside each paragraph.
• 💡Learn one finding that each theory cannot explain, because that is where the analysis marks sit.
• 💡The same technique is asked for across topics, so practise comparing by theme rather than memorising this pair on its own.
• 💡Paper 3 sets this debate in a methods context, so anchor it in a described procedure rather than in philosophy of science.
• 💡Have one named threat to validity and one named threat to reliability ready for questionnaires and for laboratory experiments.
• 💡Where a scenario is supplied, say which criterion of science it satisfies and which it fails, and why that matters for the conclusion.
• 💡These questions normally want a difference claim and a bias claim, so plan a paragraph for each before writing.
• 💡Learn two cross-cultural obedience figures precisely, because a vague statement that other countries obeyed differently earns almost nothing.
• 💡If a source describes a new study, apply the bias to that sample rather than reciting the criticisms of Milgram.
• 💡Name the variations by their content, such as the rundown office block and telephonic instructions, because the specification lists three of them by name.
• 💡Where the question names nature and nurture, state which evidence is genetic, which is environmental and which is neither.
• 💡Close with an interactionist conclusion supported by the evidence you have just used, not with a restatement of the debate.
• 💡The question often supplies two studies and asks what has changed, so practise the comparison as ethics, then method, then findings.
• 💡Keep a prejudice example alongside the obedience one, since the specification offers Tajfel and a contemporary study as the alternative route.
• 💡State explicitly whether the later work supported, refined or overturned the earlier claim; that sentence is where the judgement marks are.
• 💡This content sits in the social topic of the first paper and is often set as a short applied item about a workplace, a hospital or a school, so keep one obedience study and one prejudice reduction study ready to apply rather than describing both at length.
• 💡Longer essays here carry as many marks for application and evaluation as for knowledge, so plan two evaluation paragraphs before you start writing.
• 💡When the command word is assess or discuss the marks live in the judgement, so end each paragraph with a line on what it means for actually using the research to control behaviour.
• 💡This is asked both as a straight issues and debates question and as the application part of a scenario, so practise writing the same study twice, once as description and once as advice to a named organisation.
• 💡Discuss and assess questions want a verdict on usefulness, so say plainly which application has the stronger evidence and why.
• 💡Be ready to compare a social psychology application with one from another topic, because the papers allow questions that cut across the whole specification.
• 💡Socially sensitive research is a named issue in this specification, so it can be set as a whole question rather than only as a paragraph of evaluation inside one.
• 💡Always anchor the issue to a specific study from the social topic, because a general answer about sensitivity with no named research is capped in the lower bands.
• 💡When asked to assess, use an explicit costs and benefits structure and reach a decision instead of listing points in no order.
• 💡These are examined synoptically, so a question may let you choose your own cognitive material; prepare two studies you can steer into any of the arguments.
• 💡Extended answers are marked on a levels scheme whose top band needs sustained argument, so plan two developed points rather than five thin sentences.
• 💡The word discuss signals that description and evaluation are both wanted, roughly balanced, ending with a judgement.
• 💡Ethics questions on this case usually ask you to explain one ethical issue raised by research with brain damaged patients, so give the issue, the evidence from the case and the consequence for the participant.
• 💡If asked whether the research was justified, argue both sides and commit to a judgement, because the top band rewards a supported conclusion rather than a balanced shrug.
• 💡Method marks sit close by, so be ready to add that findings from a single patient generalise poorly while offering rare detail no experiment could ethically produce.
• 💡Method questions carry a large share of the marks and are usually set on an unfamiliar study, so practise naming a threat to validity and its control in the same sentence.
• 💡When evaluating a study, put the methodological point first and then say what it means for the conclusion, because the second half is what lifts the mark.
• 💡A question asking how memory could be measured wants an operationalised measure with a number attached, not the word tested.
• 💡Questions often ask you to assess how far a named model or study takes this approach, so plan a two sided answer that ends in a judgement.
• 💡Keep one concrete example of something the simplified account cannot explain, such as why a remembered story shifts to fit the culture of the person retelling it.
• 💡This argument pairs naturally with the debate about scientific status, and linking the two shows the synoptic thinking the top band rewards.
• 💡The command word compare asks for similarities and differences; answers giving only differences will be restricted in the level they can achieve on holistic mark schemes.
• 💡Use comparison connectives openly, whereas, in contrast, similarly, because a marker credits the comparison that is made rather than the one that is implied.
• 💡In extended answers keep a running judgement so the conclusion grows out of the argument instead of arriving from nowhere.
• 💡This is commonly set as an extended question on whether psychology should be regarded as a science, so choose a line of argument early and hold it through the answer.
• 💡Name the philosophers where they help, since Popper on falsification and Kuhn on paradigms are credited when used to make a point rather than merely dropped in.
• 💡Balance the answer with one less controlled method, such as the reconstructive studies of Bartlett, to show the discipline uses more than one approach.
• 💡Issues and debates are examined on the psychological skills paper, where extended essays ask you to assess or discuss a debate across topics, so prepare one memory example and one example from another topic rather than a list of definitions.
• 💡Questions frequently supply a short scenario about a test used in a different country or with one sex; earn the application marks by naming the feature of the scenario that creates the bias before evaluating it.
• 💡A discuss question wants both sides: give the evidence that culture shapes recall and the counter evidence that some memory features, such as the serial position curve, appear widely.
• 💡Scenario questions on this paper often describe someone whose memory has changed after an injury or after repeated questioning; name which side of the debate the scenario illustrates and then supply the other side.
• 💡For an extended answer, plan three paragraphs of nature evidence, nurture evidence and interaction, and finish with the implication for treatment or for court evidence rather than with a summary.
• 💡Learn the patient as Henry Molaison with the researchers who reported him, because a named researcher and a stated finding score where an anonymous anecdote does not.
• 💡This appears both as a memory question on the foundations paper and as an issues and debates essay on how understanding develops, so rehearse the same studies for both uses.
• 💡An essay asking you to assess how understanding has developed wants a chronology with a driver for each step; award yourself no credit for a date without the evidence that moved the field.
• 💡Where a question offers a case of brain injury in the stem, apply it to the model rather than retelling the case.
• 💡Questions on social control usually supply a real setting such as a police interview or a trial; use the setting given rather than substituting a generic example.
• 💡Pair one benefit with one cost in every paragraph, because an essay that only praises applications leaves the assessment marks unclaimed.
• 💡Remember this debate can be asked alongside any topic, so keep a memory example and a biological or social example ready to compare.
• 💡Application questions reward the link, so write one sentence of theory, one sentence of what is done, and one sentence of what the evidence shows.
• 💡Where the stem describes a service or a scheme, quote the detail from the stem when you explain how the cue works.
• 💡An evaluate question here expects effectiveness evidence, so learn one review level finding rather than only an anecdote.
• 💡Use the phrase socially sensitive deliberately and define it early, because examiners reward a candidate who separates sensitivity from a simple ethics breach.
• 💡Scenarios often name a group such as older adults, a clinical population or a minority sample; anchor every point to that named group.
• 💡For extended answers, balance the cost to participants against the benefit to the group and reach a stated judgement rather than trailing off.
• 💡These debates are examined in the skills paper as extended open response, so hold a bank of two studies for each debate that you can deploy without hunting for examples.
• 💡The marks here lean heavily towards analysis and evaluation rather than knowledge, so plan for roughly a third description and two thirds argument.
• 💡Cross-topic material is rewarded, so a biological explanation set against a psychodynamic or learning explanation earns the comparison credit the specification is asking for.
• 💡A short item usually asks for one ethical issue raised by a described study and how it was or should be dealt with, so always add the safeguard to the issue.
• 💡In the extended answer the credit is in weighing cost against benefit and reaching a judgement, not in the length of the list of guidelines.
• 💡Read whether the question says ethical issues or ethical implications, because the first is about running the study and the second is about living with its conclusions.
• 💡Practical issues usually arrive attached to a described study, so say what the issue does to reliability, validity or generalisability rather than naming it and moving on.
• 💡Comparison items should be answered on shared criteria such as resolution, invasiveness, cost and what may be concluded, taken one at a time across techniques.
• 💡Keep one scanning study you know well, so every practical point you make has a concrete example to land on.
• 💡This is nearly always a discuss or evaluate item, so structure it as what the level explains, what it leaves out, and what an interactionist account adds.
• 💡Answer about one behaviour throughout, moving up and down the levels on it, rather than listing approaches one after another.
• 💡The same material argues the other way in a question about psychology as a science, where reduction to measurable variables is a strength, so learn it once and use it twice.
• 💡Plan the criteria before writing, because a comparison essay is built from paragraphs that each hold one criterion and both explanations.
• 💡Bring a study to each side of each criterion, since an unsupported comparison reads as opinion.
• 💡Where the question names only one theme, choose the partner theme you have real evidence for rather than the one that sounds most different.
• 💡Answers score when the criteria are stated and then tested against named methods, so avoid a general discussion of what science is.
• 💡Give the essay range by setting a strong case such as synaptic transmission against a weaker one such as an unfalsifiable psychodynamic claim.
• 💡Finish with a judgement about which parts of psychology meet the criteria and which do not, because the conclusion is where the evaluation marks are confirmed.
• 💡Issues and debates are examined directly in Paper 3 and also earn synoptic credit inside biological essays on Paper 1, so keep one hormone study and one cultural study ready to insert.
• 💡Assess and discuss both demand a judgement, so finish by saying how far hormones explain the sex difference in aggression rather than listing evidence on each side.
• 💡If the stem names a person or a setting, apply the bias to them: say whose behaviour the theory was built from and who is now being judged by it.
• 💡Plan this as three paragraphs, one biological study, one environmental study and one interaction study, then a judgement, because that structure maps onto the levels in extended response mark schemes.
• 💡An analyse question wants you to break a behaviour down into genetic, neural and environmental contributions rather than argue a case.
• 💡When the stem describes a violent individual, apply diathesis stress to that person specifically, naming their predisposition and their trigger.
• 💡This content doubles as biological research methods, so the same material can be used for a methods question and for the issues and debates question, provided it is framed as how understanding developed.
• 💡Use one named study for each technique so development is evidenced rather than asserted.
• 💡Compare questions need a criterion stated before the comparison; spatial resolution, temporal resolution, invasiveness and cost are the four that earn marks.
• 💡Structure it as who gains control, who loses freedom, and what safeguard exists, because that produces the two sided argument extended response levels reward.
• 💡Pair this with free will and determinism, since the two are almost always examined together in a biological essay.
• 💡Scenario items often describe a policy such as screening or compulsory medication, so apply the ethics to that policy rather than writing in general terms.
• 💡Biological essays gain a mark band by ending on an applied paragraph, so keep one named intervention with a result and a limitation.
• 💡Explain questions want the mechanism spelled out, for example why raising serotonin activity should reduce impulsive rather than planned aggression.
• 💡If the stem sets a prison, a school or a workplace, choose the intervention that fits that setting and refer to details in the stem.
• 💡Keep one genetics example and one scanning example so a named case can be supplied quickly, since examiners reward the specific group placed at risk.
• 💡This attaches to any extended evaluation of a biological study, so it can rescue an essay that is running short on evaluation.
• 💡Assess questions want a cost and benefit structure ending in a judgement about whether the research should have been carried out or published.
• 💡These themes are examined as an extended synoptic essay and also as short application questions attached to a source, so prepare both the long argument and the quick applied point.
• 💡Plan with a grid, debates down the side and studies across the top, so a single study can be recycled into several different essays.
• 💡Read the command word first, because discuss and evaluate require both sides plus a judgement, while explain wants reasoning applied to the stimulus only.
• 💡Questions often describe a short unfamiliar animal study and ask you to explain two ethical issues it raises, so answer from the procedure in front of you rather than from a memorised list.
• 💡In the extended answer, run the argument as a weighing exercise and reach a stated conclusion, because an unresolved list of concerns is capped.
• 💡Keep one human parallel ready, such as Watson and Rayner, so you can compare how differently the two sets of rules protect a participant.
• 💡This is often set as an applied question with a novel animal study in the stem, so quote details from that stem when explaining why generalising to people is risky.
• 💡Balance is required, because an answer that only attacks animal research cannot reach the higher bands.
• 💡Learn one therapy derived from animal work so you always hold evidence that some transfer does succeed.
• 💡The debate is normally set against holism, so structure the answer around levels of explanation and finish with a judgement about which level answers the question asked.
• 💡A short question may ask you to explain how one named study shows reductionism, so keep Skinner's lever pressing rats as the clean worked example.
• 💡Use the word parsimony once, correctly, to show you know why simplicity is a scientific virtue as well as a limitation.
• 💡Comparison questions carry several marks per developed point, so write elaborated comparisons rather than a long list of one line differences.
• 💡A source may ask you to compare a learning explanation with a biological or cognitive one, so keep a non learning explanation of the same behaviour ready.
• 💡Use comparative connectives such as whereas and by contrast, because they show the marker exactly where each comparison sits.
• 💡The synoptic extended answer may ask you to assess whether the discipline is a science using evidence from any topic, so pair a behaviourist example with a less measurable one such as a case study.
• 💡Popper and falsifiability deserve a sentence each, but only alongside a psychological example that shows the idea working.
• 💡Define terms inside the argument, because a clear definition of objectivity or falsifiability where it is used earns more than a glossary paragraph at the start.
• 💡Culture and gender are usually attached to a source in the paper, so quote the sample described in that source when you argue that generalisation is limited.
• 💡Carry one cross cultural study for each topic, because a debate answer resting only on British and American evidence undermines its own argument.
• 💡When asked to discuss socially sensitive research, name who could be harmed by the finding and how it could be used before you judge whether the work was worth doing.
• 💡Issues and debates are examined inside the topic on Paper one, most often as an eight or twelve mark discussion, so plan one paragraph for nurture, one for nature and a conclusion that argues interaction rather than sitting on the fence.
• 💡When the stem describes an observation, quote its details back, because the application marks are awarded for using the scenario and not for a general essay about the debate.
• 💡Roughly half the marks in an extended answer are for evaluation, so keep the description of Bandura or Skinner to a couple of sentences and spend the rest on argument.
• 💡This debate is usually asked as an assess or discuss question worth eight or more marks, so structure it as then, next and now, with a named study at each stage.
• 💡Anchor at least one stage in current practice, because the specification example points at therapy, and answers that stop in the nineteen sixties look incomplete.
• 💡Where a question offers a scenario about media effects, use it as your modern stage rather than inventing a separate example.
• 💡Expect this as a scenario about a ward, a school or a prison, and use the setting's own details when you argue about who holds the power.
• 💡Pair every control point with a benefit point, then judge, because the marks at the top are for the weighing rather than for the list.
• 💡Keep one named therapy ready in enough detail to describe the reinforcers and the target behaviour, since a vague reference to behaviour therapy will not carry the knowledge marks.
• 💡Questions often ask you to apply learning theory to a novel setting, so practise converting a scenario into target behaviour, reinforcer, schedule and measurement.
• 💡Keep one effectiveness study and one limitation study ready, because a supported judgement needs evidence on both sides.
• 💡If the question is worth eight or more marks, finish with a line on what the application does not achieve, since that is the conclusion markers reward.
• 💡This is usually an extended discuss or evaluate question, so plan participant level issues, group level issues and the counter-argument that avoidance is also a choice.
• 💡Have one classic and one modern example ready, because an answer resting only on studies from a century ago looks dated to a marker.
• 💡Where the stem names a therapist or a clinic, build the answer around that relationship rather than switching to a laboratory example of your own.
• 💡The skills paper sets extended writing on issues and debates, so build a bank of clinical examples you can drop into whichever debate appears.
• 💡Plan essays as debate, evidence, counter evidence, conclusion, because a conclusion that takes a position and justifies it is what reaches the top band.
• 💡One study can serve several debates: Rosenhan works for ethics, for the validity of diagnosis and for social control, so learn it once and use it three ways.
• 💡Ethics questions usually ask you to apply a code to a scenario, so quote the principle and then say what the psychologist in the stem should actually do.
• 💡Keep one benefit and one cost of diagnostic labelling ready, because assess and evaluate both demand the two sided answer.
• 💡If asked about consent in clinical research, mention prospective or process consent, where permission is revisited as the study proceeds rather than signed once.
• 💡Questions here often give a short study description and ask you to identify and explain a practical or methodological issue, so practise reading a stem for its design first.
• 💡Use the vocabulary that earns method marks: operationalisation, standardisation, replicability, attrition and demand characteristics.
• 💡Where a question asks you to evaluate a named clinical study, run it through sampling, measurement, control and ethics in that order so nothing is left out.
• 💡Issues and debates are examined synoptically in the psychological skills paper, so keep one clinical example per debate ready rather than trying to recall everything at once.
• 💡When the command word is evaluate or assess, finish with a judgement in your own voice, for example that reduction is a productive research strategy but a poor description of a patient.
• 💡In an extended response, run one named study per paragraph; an essay with no research attached is capped however fluently it argues.
• 💡The command word compare asks for similarities and differences, so a one sided answer limits itself before it starts.
• 💡Build a revision grid with rows for publisher, duration required, subtypes, cultural coverage and reliability evidence; that grid answers most versions of this question.
• 💡Check the command word before planning, because explain wants the mechanism of one account whereas compare wants both held against a shared basis.
• 💡This is examined synoptically, so you may be handed an unfamiliar method section and asked how far it is scientific; practise applying the features to a study you have never seen.
• 💡Carry two clinical examples, one that supports scientific status and one that undermines it, so a balanced answer is available whichever way the question leans.
• 💡Decide your verdict before writing, because the top bands reward a sustained line of argument rather than a list of points.
• 💡Questions often present a novel scenario containing a rate difference and ask you to explain it, so practise generating two competing explanations for any difference you are shown.
• 💡Say which kind of bias you mean, bias in how the research was done or bias in how the diagnosis is made, because the evaluation differs for each.
• 💡Keep one cultural study and one gender study with named researchers, since an unsupported claim of bias earns little in an extended response.
• 💡Any essay here is improved by one genetic study and one social study, each named with what it actually found rather than what it is famous for.
• 💡Where the question sets biological against social psychology explicitly, organise by explanation and put the comparison inside each paragraph rather than saving it for the end.
• 💡A concluding paragraph that states which side the evidence favours and why is what separates the top band from a competent list.
• 💡Two or three changes explained properly beat a timeline of everything, because the marks are for the reasoning about why understanding shifts.
• 💡Link this deliberately to social control and to socially sensitive research, since the same examples serve all three and examiners reward synoptic connections.
• 💡If the question says assess, judge whether the changes represent genuine scientific advance or shifting social values, and support the judgement.
• 💡Expect a short scenario, such as a person refusing medication, followed by a question asking you to discuss the issues raised; answer with one named mechanism and both sides of it.
• 💡Keep one legal example and one research example ready, because an answer resting entirely on opinion cannot reach the higher bands.
• 💡Where the command word is evaluate, judge how far control is justified rather than simply describing that it happens.
• 💡One therapy for a mood disorder and one for a psychotic disorder will cover almost any version of this question, so learn two properly rather than five thinly.
• 💡Where the command word is assess, judge the quality of the evidence as well as its direction, naming the sample, the length of follow up and whether the trial was blind.
• 💡A short sentence on cost and availability adds the societal angle that a purely clinical answer misses.
• 💡Expect an unfamiliar study in the question stem with an instruction to explain why it is socially sensitive; apply the four areas to that study rather than reciting them.
• 💡Name the governing code, the British Psychological Society ethical guidelines, and make the point that a code regulates conduct during a study but not the later use of its results.
• 💡Keep one clinical example and one genetic example, since the two raise different consequences and a question may push towards either.
• 💡Debates are examined in extended items that carry most of the evaluation marks, so plan first and give each paragraph one debate and one study.
• 💡Questions usually name the debate and the topic, so build a grid of debate against study for criminological psychology and learn a few cells thoroughly rather than all of them thinly.
• 💡Assess and discuss both expect a conclusion saying which way the evidence falls and why, so leave time and space for it.
• 💡Ethics items often give an unfamiliar study and ask for two issues it raises, so practise reading a method for consent, harm, deception, confidentiality and debriefing in that order.
• 💡When the command is discuss, give the defence as well as the charge, because a one sided answer will not reach the higher bands.
• 💡Keep one field example and one laboratory example ready so you can compare what each buys and what each costs.
• 💡Practical issue items are often worded as explain one difficulty in researching this area, so have a two sentence answer ready giving a reason and its consequence.
• 💡This point earns evaluation marks in any jury essay, so attach it to a named study rather than saving it for a debates question.
• 💡If asked to suggest an improvement, propose something a researcher could lawfully do, such as filmed cases with full deliberation, rather than studying a live trial.
• 💡Reductionism questions name a topic and expect study detail, so learn one biological study and one eyewitness experiment well enough to describe the method.
• 💡Pair reductionism with holism inside the same paragraph, because the mark scheme rewards the comparison rather than two definitions.
• 💡An assess item wants a conclusion about usefulness, so finish by saying what an interactionist account adds that neither level manages alone.
• 💡The command word compare expects similarities and differences, so plan two of each before writing anything.
• 💡Extended items on this theme reward a worked example, so carry one offence type, such as violent offending, and explain it three ways.
• 💡Keep the debates vocabulary in reserve, nature and nurture, determinism and levels of explanation, because it is how the top band answers are written.
• 💡This is usually asked as an essay telling you to assess or evaluate how far psychology can be called a science, so plan two named studies you can use as evidence on both sides rather than a list of criteria.
• 💡Questions often supply a short scenario about a police or courtroom study and ask you to explain whether it is scientific. Take the scenario apart by criterion: control, objectivity, replicability, falsifiability.
• 💡When the command word is discuss, ensure you both describe the criteria and evaluate them, so never spend the whole answer merely defining terms such as empiricism.
• 💡Expect an extended essay asking you to evaluate the claim that culture and gender affect jury decision making, so pair every study with one methodological criticism.
• 💡Application questions give a named defendant with an accent, an ethnicity or a gender and ask you to explain likely juror responses, so justify the prediction with the named study rather than common sense.
• 💡If the command word is analyse, break the influence into stages: first impression, interpretation of ambiguous evidence, then sentencing recommendation.
• 💡The essay usually asks you to assess biological against social or learning explanations of criminal behaviour, so keep one adoption or twin study and one learning study ready with findings, not just titles.
• 💡Application items give a family history plus a difficult upbringing and ask you to explain the offending, and the credit lies in using both sides and naming the interaction.
• 💡When asked to compare explanations, use shared criteria such as quality of evidence, determinism and usefulness for prevention rather than describing each one in turn.
• 💡This statement is a favourite for the essay asking you to discuss how psychological understanding has developed over time, so structure the answer chronologically and finish with current practice.
• 💡Questions may ask you to compare the cognitive interview with a standard police interview, so use the same headings for both, such as question type, control of the witness and risk of contamination.
• 💡If asked to outline the enhanced cognitive interview, add the social elements, for example rapport building, witness compatible questioning and transfer of control, on top of the original four.
• 💡Essays here ask you to evaluate a treatment for offending as an instrument of social control, so describe the programme briefly and spend most of the answer on power, consent and effectiveness.
• 💡Where a scenario names a prison programme, apply the ethical issues by name: informed consent, right to withdraw, and the psychologist's dual role as therapist and assessor.
• 💡Use the word effectiveness carefully and always say which outcome measure a study used, because markers reward that distinction more than a general claim that the therapy works.
• 💡Extended essays ask you to assess how useful psychological knowledge has been in the justice system, so structure the answer by application and judge each one in turn.
• 💡Where a scenario describes a robbery involving a weapon, explain the identification risk and then recommend the safeguard, because the application marks sit in the recommendation.
• 💡Avoid listing studies, since the credit is in linking each finding to the change it produced or should produce.
• 💡Essays ask you to discuss issues of socially sensitive research using criminological examples, so prepare two examples where you can state both the harm and the benefit.
• 💡In application items, name the group at risk, name the likely misuse, then give one design or reporting safeguard.
• 💡If asked to explain why a named study is socially sensitive, anchor the answer in what could be done with the finding rather than in how the participants felt.
• 💡Issues and debates are examined in extended essays that name the debate and leave the content to you, so prepare child psychology as your worked example for at least three of them.
• 💡Credit comes from using content as evidence inside an argument, so avoid retelling a study in full and cite only the part that carries the point.
• 💡A question may ask you to set child psychology against another topic area, so know one debate where the two differ sharply.
• 💡Issues and debates questions in the skills paper usually hand you a novel piece of research in the stem and ask you to discuss its ethics, so practise reading a scenario and naming the right that the procedure threatens.
• 💡A longer evaluate question needs both sides plus a judgement, so bank one sentence that comes down on whether the research could have been done any other way.
• 💡Quote the convention sparingly. One accurate article and its meaning earns more than a list of numbered articles with nothing attached to them.
• 💡Method questions in the skills paper often ask for one strength and one weakness of a named technique in the context of child research, so finish each sentence with what it means for the conclusion.
• 💡If a stem gives numbers of studies or of participants, use them. Quoting the unevenness of a sample is concrete evaluation and reads as knowledge rather than as a template.
• 💡Keep one worked example of an operationalised category ready, because describing how a behaviour was turned into something countable is the quickest route to method marks.
• 💡Debate questions usually name the debate and a piece of content, so plan in two columns, what is reduced and what is lost, then write a judgement about whether the reduction was worth making.
• 💡Use one example across levels of explanation: learning theory reduces attachment to feeding, Bowlby reduces it to an evolved system, Ainsworth reduces it to observable behaviour.
• 💡Examiners reward a named alternative. Harlow's infant monkeys preferred a cloth surrogate that gave no milk, and Schaffer and Emerson found many infants were most attached to someone who did not feed them, which sinks the cupboard love reduction.
• 💡Comparison questions in this option often supply a short scenario about a particular child, so anchor each theory to something in the stem before you weigh them.
• 💡Keep comparative connectives visible, because phrases such as whereas, by contrast and both accounts predict are what tell a marker you are comparing rather than listing.
• 💡One sharp methodological evaluation beats three vague ones. Retrospective data, small samples and generalising from geese and monkeys to human infants are the three most creditable lines here.
• 💡Science questions are usually worded as an assess or evaluate task on whether a named area is scientific, so decide your line early and defend it rather than sitting on the fence.
• 💡The same cross-cultural evidence serves the science debate, the culture debate and nature and nurture, so learn it once and label which debate you are using it for.
• 💡Name the researchers. Attributing the pooled cross-cultural analysis to van IJzendoorn and Kroonenberg and the Japanese sample to Takahashi is the difference between knowledge and impressions.
• 💡Culture questions are often set as an extended discuss or evaluate task, so structure by finding, then bias, then implication, and keep a conclusion about whether the measure travels.
• 💡Bring one non-Western child rearing example and use it to show that the procedure, rather than the children, may be what is unusual.
• 💡If the question names gender, do not force a male and female comparison where the evidence is thin; say openly that the option treats gender indirectly, then use the day care outcomes.
• 💡Nature and nurture questions in the skills paper often supply a scenario about one family, so state the two predictions and then decide which the scenario fits.
• 💡Keep one piece of evidence for each of the three positions, innate, environmental and interactionist, so a longer answer has structure without padding.
• 💡Write the debate out in words rather than shorthand, and make the judgement explicit in the final sentence.
• 💡Questions on how understanding has developed reward chronology plus a reason for each change, so pair every step with the evidence that forced it.
• 💡Dates help but a finding beats a date, so if you cannot recall a year, describe what the study showed and who carried it out.
• 💡This content doubles as evaluation of Bowlby elsewhere in the option, so learn it once and deploy it in both places.
• 💡Social control questions usually arrive as a discuss or assess task attached to a scenario about advice given to parents, so identify who benefits and who loses before you evaluate.
• 💡Hold one example from day care and one from autism, because the specification names both and an answer with only one looks thin.
• 💡Link social control to free will and determinism explicitly, since an intervention justified by deterministic claims about early experience is the argument examiners expect to see made.
• 💡Issues and debates are examined through extended writing, and Pearson often hangs them on an applied stem about a parent, a nursery or a social worker, so answer the person in the stem rather than writing a general essay.
• 💡Prepare one worked example from each bracketed area so that no scenario leaves you with nothing, and learn a number or a named finding for each.
• 💡Credit in the top band comes from a conclusion that weighs usefulness against limitation, so plan two paragraphs for and one against before you start writing.
• 💡This is a named issues and debates strand, so it appears both as a standalone essay and as a line you are expected to weave into the evaluation of any child psychology study; keep one sentence of social sensitivity ready for each study you learn.
• 💡Watch for the command words assess and discuss, and for stems phrased as to what extent, all of which demand both sides and a judgement.
• 💡Learn one example from each bracketed area and rehearse the risk against benefit conclusion so that it is written rather than improvised.
• 💡The extended open response in the third paper carries twenty marks, so plan a line of argument before writing and budget the time to finish the conclusion.
• 💡Assess and discuss ask for a judgement, while analyse and explain ask for the mechanism, so let the command word decide the shape of the answer.
• 💡Keep a grid of the debates against three pieces of health psychology evidence for each, because a question can pair any debate with this topic.
• 💡Ethics questions on this topic usually divide, so be ready to write separately about animals and about humans with one named study attached to each.
• 💡If asked to evaluate the use of animals in drug research, structure by the three Rs, evidence it with one study, and close with the benefit that could not have been obtained any other way.
• 💡Where a scenario describes a therapy, the marks come from naming the principle at stake, such as protection from harm or valid consent, and applying it to that scenario rather than to research in general.
• 💡Practical issues appear in the methods section of the paper as well as in the debates essay, so be ready to apply them to a study described in a source, such as a scanning study of drug action.
• 💡Answers score higher when each issue is followed to a consequence for the conclusion, so build the paragraph as issue, example, effect on the claim.
• 💡Keep one animal example and one human example ready, because questions often ask you to compare the practical strengths of the two approaches.
• 💡Questions on reductionism in this topic expect a named animal experiment, so have Olds and Milner ready with what was manipulated and what was measured.
• 💡In an extended response, organise by level of explanation rather than study by study, then judge which level answers the question set.
• 💡Pair every reductionist claim with one piece of evidence that context matters, so both sides are present without a separate theory paragraph.
• 💡When the command word is compare, three criteria with one point per side on each is quicker to write and scores better than prose that drifts between accounts.
• 💡Scenario questions name a person and a drug, so pick the explanation the details point to and quote those details back as you apply it.
• 💡Hold one learning and one biological explanation for each of alcohol, heroin and nicotine, because the specification asks for them drug by drug.
• 💡Carry one study that shows the scientific method working well and one that shows its limits, and the essay has a structure whatever the wording.
• 💡Where a question names objectivity or replicability in particular, answer on that criterion instead of reciting the whole list.
• 💡This debate is often paired with reductionism in extended responses, so a paragraph linking control of variables to level of explanation can serve both.
• 💡A question on culture expects the cross-cultural research from the methods part of this topic, so link the debate to that study rather than to one from another application.
• 💡When gender is asked about, say plainly that this topic does not prescribe gender content and then argue from evidence that does exist, because the reasoning is what is rewarded.
• 💡Watch the tariff: short questions want one developed point supported by evidence, while an extended response wants both sides and a conclusion.
• 💡The cleanest structure is nature evidence, then nurture evidence, then interaction, with a named study in each of the first two parts.
• 💡If the question names one drug, keep the evidence to that drug where you can, since alcohol carries the strongest genetic evidence while nicotine and heroin carry strong learning evidence.
• 💡This debate sits next to the comparison of explanations, so the same studies can be reused provided you change what you are arguing with them.
• 💡This is asked both as a health psychology question on the applications paper and as an issues and debates essay on the skills paper, so prepare one worked line of development that can be stretched into an essay or compressed into a paragraph.
• 💡When the command word is discuss or assess, give roughly half the answer to the limits of the development, such as the weak evidence that campaigns change behaviour and the survival of moral framing inside current policy.
• 💡A year earns credit only when it is attached to a finding, so write what the study showed first and keep the date short.
• 💡Questions often supply a short scenario about someone ordered into treatment, so quote the scenario back; application marks are lost by answering with a generic essay.
• 💡Pair every example of control with the ethical principle it strains, normally valid consent or the right to withdraw, and name the code of conduct you are relying on.
• 💡An extended answer wants two contrasting examples and a conclusion that states when control is justified, rather than a list of things society does.
• 💡Application questions hand you a named person with a habit, so put the therapy into their case step by step before evaluating it in general terms.
• 💡Effectiveness marks need something to compare against, so always state what the treatment was better or worse than, whether placebo, waiting list or counselling alone.
• 💡Keep one methodological criticism ready, such as high dropout from treatment trials leaving a sample of the people who were doing well anyway.
• 💡Questions usually print a short method, for example a researcher interviewing users at a needle exchange, so take the issues from the method described rather than from a memorised list.
• 💡Name the guidance you are applying and one specific safeguard, because saying that participants were protected earns far less than saying how.
• 💡In an extended answer, use one real example that had to manage this in practice, such as follow up research on returning Vietnam veterans, which depended on men admitting illegal heroin use.
• 💡The longest essay on the skills paper is on issues and debates, so rehearse timed plans that list four studies and the debate each is being used for before any prose is written.
• 💡Where the question names one debate, keep every paragraph on it; candidates lose marks by touring all the debates instead of developing the one asked about.
• 💡Keep two examples per debate from different areas of psychology, because a question can require application to an unfamiliar context and one example rarely stretches.
• 💡Ethics questions frequently arrive as application to an unfamiliar scenario, so practise naming which guideline is threatened and what the researcher should do differently.
• 💡A question that says animal research expects the legislation and the three principles; a question that says human research expects the professional code, and the two lists should not be swapped.
• 💡When asked to evaluate ethical decisions, argue both directions, because the highest bands need the case for the research as well as the case against it.
• 💡Design questions often ask you to outline a study and then justify one decision, so write the justification as a because clause linking the choice to a threat it removes.
• 💡Where an improvement is requested, give one change and then say what it would do to reliability or validity, since an unexplained improvement rarely scores beyond one mark.
• 💡These ideas are tested on the methods section of the skills paper as well as inside topic questions, so prepare them as transferable tools rather than as a separate topic.
• 💡Questions usually anchor the debate to content, asking you to assess reductionism in one explanation, so choose two explanations from different areas and know their evidence in advance.
• 💡Write the level of explanation into the paragraph opening, because an answer that says which level it is arguing at reads as analysis and one that does not reads as description.
• 💡The word assess signals a judgement, so end with a stated position rather than a summary of the arguments already made.
• 💡The command word compare expects both similarities and differences, and an answer that gives only one of the two caps itself.
• 💡Prepare one behaviour, usually aggression, obedience or addiction, with an explanation from each area attached, so any comparison question has ready material.
• 💡Use connectives that do comparative work, such as whereas, by contrast and similarly, because they force each point to mention both sides.
• 💡This is a standard extended question on the skills paper, so prepare a structure of three features, each with a study that meets it and a study that does not.
• 💡Where the question names an area of psychology, keep the examples inside that area, since a general essay about science will not reach the higher bands.
• 💡Finish with a judgement about degree, because the defensible conclusion is that the subject is scientific in some of its methods and not in all of them.
• 💡Questions often supply an unfamiliar study and ask you to explain one cultural or gender issue in it, so practise naming the bias, the evidence for it in the extract, and its effect on the conclusion.
• 💡Keep one gender example and one culture example ready from different areas of psychology, because a single example rarely covers both halves of a question.
• 💡When asked to evaluate, include the counter argument that awareness of bias has changed practice, for example through more representative samples and locally developed measures.
• 💡Twin and adoption designs appear in methods questions as well as in this debate, so be ready to say why each design was chosen and what it cannot control.
• 💡Prepare one clinical example and one non-clinical example, since questions can anchor the debate in aggression or intelligence rather than in disorder.
• 💡The verb discuss expects both positions developed before a conclusion, so allocate roughly equal space and save the interactionist judgement for the end.
• 💡This is usually an extended answer, so plan three dated stages plus one evaluative line about progress before writing, because the analysis marks come from linking each stage to the method that made it possible.
• 💡The stem expects evidence from the approaches studied across the whole qualification, so draw on learning, cognitive, biological and social psychology rather than importing outside history.
• 💡A short version asks you to outline one way psychology has changed; give one named contrast with a study on each side instead of a survey.
• 💡The board likes an application scenario, such as a prison governor planning a reward scheme, so name the approach, describe the procedure using the details in the scenario, then evaluate it.
• 💡Assess and evaluate questions want a judgement, so finish with a stated criterion such as informed consent or benefit to the person being controlled rather than a summary.
• 💡Keep one named ethical principle ready, because markers reward specific reference to consent, right to withdraw and protection from harm rather than the bare word unethical.
• 💡Extended questions usually supply a context, so make the application fit the context in the stem rather than reciting a prepared list.
• 💡Keep three applications from three different approaches ready, because the stem often rules out the obvious one or asks for more than one area of psychology.
• 💡When told to evaluate an application, separate effectiveness from ethics, since markers credit both strands and thin answers give only one.
• 💡Extended answers here are often tied to a named study, so practise taking any study you know and arguing which of the four aspects makes it sensitive.
• 💡Questions frequently pair this with the researcher's responsibility, so include a line on how findings should be reported and to whom.
• 💡In an evaluate question do not simply list problems; reach a judgement on whether the knowledge gained justified the risk to the group studied.
• 💡Anchor synoptic debates in specific named studies across different content areas (Social, Cognitive, Biological, and Learning approaches) to access the top level of AO3 marks.
• 💡Structure debate essays using balanced evaluation chains: state the debate stance, cite supporting psychological research, provide a counter-perspective, and conclude with the practical implications for society.
• 💡Avoid descriptive listing of studies; instead, continuously refer back to the specific terms of the question, assessing how the evidence demonstrates reductionism, scientific validity, or ethical compromise.

• Writing a general essay on determinism or reductionism with no social psychology study in it, which strands the answer on knowledge alone.
• Reciting the list of debates as definitions; the board asks for them applied to named content, so an unapplied definition earns very little.
• Filing an ordinary evaluation point under the wrong debate, such as calling a small unrepresentative sample an example of reductionism.
• Stating flatly that Milgram removed the right to withdraw; the verbal prods made leaving difficult rather than impossible, and the sharper point is that the procedure undermined a right the participants still formally held.
• Confusing ethical issues (how participants are treated during the study) with ethical implications (how the findings impact society afterwards).
• Assuming that deception automatically makes a study invalid, rather than recognising it as an ethical breach that is sometimes deemed necessary for methodological validity.
• Defining social desirability without identifying its effect on the data, failing to recognise that it systematically reduces validity.
• Calling a questionnaire unreliable when the weakness described is about validity, or confusing the two concepts.
• Claiming anonymity completely removes social desirability; it only reduces it, as sensitive subjects such as prejudice still draw guarded answers.
• Assuming social psychology cannot be reductionist, when theories inside it reduce obedience to a state or to a set of numbers.
• Using the term as a general insult meaning too simple, which loses the point about levels of explanation entirely.
• Describing a study's sampling weakness and labelling it as the wrong debate.
• Describing both theories in turn and never comparing them, which forfeits the marks for analysis.
• Claiming contact reduced hostility at Robbers Cave; joint activities without a shared goal made things worse, and only the superordinate tasks worked.
• Swapping the researchers, so the minimal group studies are credited to Sherif or the camp study to Tajfel.
• Treating quantitative data as automatically scientific and qualitative data as automatically not.
• Confusing ecological validity with population validity when arguing about whether findings generalise.
• Calling questionnaires objective because they produce numbers, when those numbers come from subjective self-ratings.
• Asserting that women are more or less obedient than men; the replications point in different directions and the best supported conclusion is no consistent difference.
• Treating nationality as culture, so that a Dutch and an American sample stand in for individualist against collectivist societies.
• Describing Sherif's boys as typical children, when they were deliberately selected as white, Protestant and of similar background and age.
• Equating dispositional with nature and situational with nurture, when personality, culture and upbringing are all learned influences.
• Quoting the authoritarianism scale as proof of cause when Elms and Milgram measured it after the obedience had happened.
• Claiming the variations prove situational determinism outright, when a minority refused in every single condition.
• Writing that Burger found people are less obedient today, when the difference was not significant and he argued for continuity.
• Saying Burger replicated Milgram exactly, which throws away the whole point of the comparison.
• Listing dates and study names without saying what changed, why it changed, and what that means for the original conclusion.
• Writing that Milgram's participants were cruel or that they knew the shocks were faked; their visible distress and the debriefing evidence are the point, and the study is used as evidence for situational rather than dispositional control.
• Treating uniform and status as the same variable, when Bickman held the request constant and varied only the clothing, so the conclusion concerns visible symbols of legitimacy.
• Claiming that contact by itself reduces prejudice, when in the Robbers Cave camp shared meals and films made the hostility worse until the groups actually needed each other.
• Listing applications with no evidence attached, when the evaluation marks ask how effective it was and how effectiveness was measured, not whether it sounds sensible.
• Confusing the jigsaw classroom with ordinary group work, when its defining feature is that no pupil can complete the task without the others, which is exactly what removes competition.
• Assuming that a result from a short field study in one country transfers unchanged to any culture, age group or institution.
• Reducing social sensitivity to the usual list of consent, protection and confidentiality, when those apply to every study and sensitivity is about the wider group and the use made of the data.
• Arguing flatly that socially sensitive research should never be conducted, which removes the evidence base for anti discrimination work and earns little credit as an argument.
• Treating a difference between two countries as proof of cultural bias without first checking whether the samples, tasks and procedures were even comparable.
• Writing a general essay about reductionism or about science with no cognitive content, which cannot be credited on a cognitive question.
• Revising these as separate topics rather than as tools for evaluating the studies already learned, so the material never reaches the answer it belongs in.
• Listing several arguments briefly instead of developing one or two, which sacrifices the marks awarded for a sustained line of reasoning.
• Saying he had no memory at all, when short-term memory, language and procedural learning were largely intact and older childhood memories were partly preserved.
• Claiming the research was unethical because he was deceived, when the real issues are capacity to consent and the impossibility of withdrawing.
• Describing the surgery as an experiment, when it was a clinical treatment for severe epilepsy and the amnesia was an unintended consequence.
• Stating that confidentiality was breached during his lifetime, when he was known publicly only by his initials until after he died.
• Using the phrase lacks ecological validity as a complete evaluation, without saying which feature of the task is unlike real remembering or how that would change the outcome.
• Confusing reliability with validity, so a study is praised as valid simply because the procedure is standardised and repeatable.
• Assuming a field study must be more valid, when losing control of what witnesses discussed or how long they watched weakens the conclusion in a different way.
• Reading forgetting curves as evidence that material has been lost, when cued recall studies show much of it is stored yet inaccessible.
• Defining this as oversimplifying, which loses the point that it is a deliberate research strategy with real scientific advantages.
• Confusing it with determinism, which is a separate argument about what causes behaviour rather than about levels of explanation.
• Saying the multi-store model breaks memory into parts and stopping there, with no statement of what is gained or lost by doing so.
• Treating holism as obviously correct, when holistic accounts are harder to test and often describe behaviour rather than explain it.
• Describing two models fully and adding a single sentence of comparison at the end, which is marked as description.
• Saying one model replaced the other, when the working memory account elaborates the short-term stage and leaves the rest of the structure standing.
• Attributing the episodic buffer to the original working memory model, when Baddeley added it years later.
• Using evidence loosely, such as claiming KF supports working memory without saying which subsystem was spared and which was damaged.
• Listing features of science without applying them to any cognitive study, which cannot gain the application marks.
• Treating quantitative data as automatically scientific, when a number produced by a badly operationalised measure is no more objective than a description.
• Saying laboratory experiments prove cause and effect, when they support a causal inference that later work can overturn.
• Confusing reliability with objectivity, so a procedure is called objective merely because it gives consistent results.
• Reporting the Spanish digit span result as proof that Spanish speakers have poorer working memory, when the explanation offered is the number of syllables in the number words rather than capacity.
• Treating culture bias as a synonym for ethnocentrism and writing nothing about how the bias enters the method, the materials or the sampling.
• Claiming women simply have better memories than men from a single study, with no mention of the type of material or the size of the difference.
• Describing Bartlett as a memory experiment with controls, when serial and repeated reproduction lacked standardised instructions and recall was scored subjectively.
• Writing that the patient lost all his memory, when semantic knowledge from before the surgery, his vocabulary and his procedural skills survived.
• Treating nature and nurture as a choice to be settled, instead of arguing for an interaction and giving evidence for it.
• Using Loftus and Palmer as proof that memory is wholly unreliable, ignoring that the participants watched a film in a laboratory with no consequences for being wrong.
• Calling a single case study an experiment and claiming it establishes cause when nothing was manipulated.
• Describing the working memory model as a replacement for the whole multi store model, when it re specifies the short term store and leaves the long term store largely intact.
• Listing the components of working memory with no evidence attached, so the answer reads as description where evaluation marks were available.
• Saying rehearsal explains everything in long term memory, ignoring flashbulb and one trial learning that needs no rehearsal.
• Confusing the episodic buffer with episodic long term memory, which is a different idea entirely.
• Writing that eyewitness testimony is always wrong, when studies also find recall of central details of a real event can be resilient over time.
• Describing the cognitive interview as one technique instead of a set of components, and giving no evidence about what it improves.
• Treating social control as a synonym for anything applied, so the answer never says who is controlling whom.
• Reporting the fabricated childhood event study as showing that most people accept false memories, when only a minority did and the suggestion came from a trusted family member.
• Claiming reminiscence work improves memory in general, when the measured gains are mostly in wellbeing and communication rather than in recall scores.
• Describing the application with no psychology attached, so the answer reads as a leaflet rather than as an explanation grounded in cue dependent forgetting.
• Assuming any application is automatically a strength of the research, with no discussion of evidence quality or of who is excluded from it.
• Confusing reminiscence work with reality orientation, which is a different intervention with different aims.
• Listing general ethical guidelines with no link to why this group is particularly exposed.
• Saying the research should simply not be done, without weighing the cost of leaving a vulnerable group unstudied.
• Treating consent as a single event, when the point of the example is that it cannot be remembered and must be revisited.
• Confusing socially sensitive research with unethical research, when a study can be scrupulously conducted and still be socially sensitive.
• Writing a general essay about free will and determinism that contains no biological study at all.
• Describing studies at length and never converting the description into an argument, which leaves the answer scoring on knowledge alone.
• Assuming reductionism, determinism or science are automatically criticisms, when each is a trade with gains as well as costs.
• Confusing the ethics of how a study was run with the ethics of what its findings are used for, and answering the wrong one.
• Listing the guidelines with no study attached, which answers a question about ethics in general rather than about this research.
• Saying consent was obtained without saying who gave it, when for young children it is parental consent plus the child's own agreement to take part.
• Using socially sensitive to mean upsetting, when it means the findings carry consequences for the people or group the research is about.
• Claiming that a debrief removes harm, when it can inform and reassure but cannot undo distress that has already been caused.
• Concluding that a scan proves a brain area causes aggression, when a group difference alongside a behaviour is still correlational evidence.
• Describing functional magnetic resonance imaging as a direct record of neurons firing rather than of blood oxygenation.
• Offering the word expensive as a complete evaluation instead of tracing cost through to sample size, power and generalisability.
• Using a structural technique as evidence about activity during a task, which it cannot provide.
• Writing that reductionism is bad without saying what is lost by working at that level.
• Confusing reductionism with determinism, when one is about the level of explanation and the other is about whether behaviour could have been otherwise.
• Presenting a genetic finding as making violence inevitable, when the same research shows the effect depends on experience.
• Treating an operationalised laboratory measure of aggression as if it were the real behaviour outside the laboratory.
• Producing two accurate descriptions with no comparison, which caps the answer on knowledge marks alone.
• Claiming the psychodynamic explanation has no evidence, when it rests on case material such as Little Hans (1909); the sustainable criticism is that case interpretation cannot be falsified.
• Calling the biological explanation scientific and stopping there, without saying what its methods can and cannot establish.
• Ignoring an obvious third theme when the question allows it, such as the learning account demonstrated by Bandura, Ross and Ross (1961).
• Arguing that psychology is a science because it uses laboratories and equipment, when the criteria are falsifiable prediction, control and replicability.
• Treating scan images as photographs of thought rather than as statistical maps built from an indirect signal.
• Forgetting that human participants react to being studied, so demand characteristics limit objectivity in a way that does not apply in chemistry.
• Equating objectivity with putting a number on something, when a biased measure yields precise numbers that are still wrong.
• Writing that testosterone causes male aggression. Human evidence is overwhelmingly correlational, and meta-analysis puts the relationship at about a correlation of 0.14, which is weak.
• Muddling alpha and beta bias, or using gender bias simply as a synonym for sexist without saying whether differences are being exaggerated or ignored.
• Treating culture as a synonym for country or ethnicity and then offering no cross-cultural study at all.
• Forgetting female aggression exists, when premenstrual hormonal claims have been argued in court, which is exactly where this issue becomes socially sensitive.
• Treating the debate as a choice and arguing for one side, when the mark scheme rewards an interactionist judgement supported by a gene and environment study.
• Reporting Raine as proof that brain damage causes murder, ignoring that scanning shows association and that the control group was matched on age and sex, not on lifestyle.
• Citing twin studies as clean evidence for nature while ignoring the equal environments assumption, which is that identical twins are treated more alike as well as being more alike genetically.
• Saying aggression is in the amygdala. The evidence points to a network in which weak prefrontal regulation fails to restrain limbic responding.
• Listing machines without saying what each changed about psychological knowledge, which answers a methods question rather than the debate the statement sets.
• Claiming a scan shows the brain lighting up with thought. Functional magnetic resonance imaging measures blood oxygenation with a lag of several seconds and produces a statistical map, not a picture of a mental state.
• Assuming newer always means better, when electroencephalography still has far better timing and a scanner requires an artificial task performed while lying still.
• Writing about scanning with no historical starting point, so there is no development to describe.
• Turning the answer into science fiction about mind control instead of citing documented practice such as leucotomy, sedation in care settings and risk assessment.
• Assuming all social control is unethical, when consented and beneficial treatment is control too and the examiner wants a weighed judgement.
• Naming lobotomy with no detail about who it was performed on or what it cost them, which reads as a slogan rather than evidence.
• Confusing social control with social sensitivity; the first is about influencing behaviour, the second about consequences for a group.
• Listing therapies with no outcome evidence. The credit is for what the research showed, not for what a treatment is intended to do.
• Claiming anger management stops violent reoffending, when it suits reactive aggression and the reconviction evidence is inconsistent.
• Ignoring the costs of application, such as labelling people as biologically violent, or treating diet and medication as substitutes for tackling social causes.
• Describing a treatment without saying which explanation of aggression it comes from, so the link between knowledge and application is never made.
• Reciting general ethical guidelines. The statement is about social sensitivity, so the answer must identify harm to a group, not only risk to a participant.
• Saying data were anonymous when they were merely confidential, which misdescribes what the researcher actually did.
• Concluding that socially sensitive research should simply not be conducted, with no account of the benefits that would be lost.
• Treating confidentiality as absolute, when a disclosure of ongoing harm obliges a researcher to break it, and participants should be told that in advance.
• Treating a debate as a settled verdict, for example writing that behaviourism is reductionist and stopping there, when the marks come from showing what that reduction buys and what it costs.
• Importing studies from other topics without ever linking them back to learning, so a long description of a biological study of aggression earns nothing.
• Discussing the debate in the abstract with no study attached, which holds the answer in the lowest band however fluent the writing is.
• Applying human safeguards such as informed consent and the right to withdraw to animals and treating the breach as self evident, when the framework for animals is licensing, justification and harm minimisation instead.
• Claiming animal research is banned or unregulated; it is legally licensed and inspected, and saying otherwise throws away the accuracy marks.
• Asserting that the findings were worthless, which removes the benefit side of the argument and leaves no balanced judgement to reach.
• Writing only that humans are more complex than animals, which is assertion; the mark comes from naming the process, language, expectancy or preparedness, that fails to transfer.
• Confusing generalisation from animals with generalisation from a small or unrepresentative human sample, when the question normally specifies which of the two is at issue.
• Treating instinctive drift and preparedness as one idea, when drift is learned behaviour decaying back towards instinct and preparedness is a bias in what can be learned at all.
• Using reductionism as a term of abuse, when examiners reward the recognition that reduction is what makes a theory testable, so an answer carrying no strength is one sided.
• Confusing environmental reductionism with biological reductionism and then writing about neurotransmitters in what is meant to be a learning answer.
• Saying social learning theory is not reductionist at all, when it still reduces gender and personality to observed models, simply with cognitive steps inserted.
• Describing each theory in turn and never comparing them, which is the commonest way a well informed answer loses every comparison mark available.
• Saying classical conditioning explains voluntary behaviour or that operant conditioning explains reflexes, when the split between elicited and emitted behaviour is what the whole comparison rests on.
• Treating vicarious reinforcement as identical to direct reinforcement, when the point of Bandura's work is that no reward ever reached the observer.
• Listing the features of an experiment without ever saying what makes something scientific, so the debate itself is never engaged.
• Confusing reliability with falsifiability, when a finding can be repeated dependably and still make no prediction that could ever fail.
• Assuming that laboratory means scientific, since a tightly controlled study of an artificial task in a cage can still tell us very little about behaviour outside it.
• Writing that a study is culturally biased because the sample was American, without ever saying what the bias does to the conclusion being drawn.
• Confusing alpha bias, which exaggerates the differences between men and women, with beta bias, which minimises or ignores them.
• Assuming modelling works identically everywhere, when the reinforcement that follows an imitated act is exactly the thing that varies between cultures.
• Writing the debate in the abstract and never touching the observation described in the question, which loses every application mark on offer.
• Claiming the Bobo doll studies prove aggression is learned, when what was measured was imitation of specific acts minutes after exposure, in a room with the same toys and no victim.
• Treating nature as meaning genes only and forgetting hormones, brain maturation and preparedness, which are the parts of the biological case that actually bite on a behavioural observation.
• Assuming an observed gender difference must be innate, when adults reinforce the same behaviour differently in boys and girls and children are more likely to copy a same-sex model.
• Producing a list of dates and names with no argument about what actually changed in the explanation or the practice.
• Saying behaviourism was proved wrong, when operant principles still underpin behaviour change programmes in schools, wards and prisons.
• Describing one study in detail and calling that development over time, when the credit is for the comparison between at least two points in the timeline.
• Confusing development of understanding with development of the person, and drifting into child development instead of the history of the explanation.
• Asserting that all therapy is social control without distinguishing agreed treatment of distress from behaviour management imposed for institutional convenience.
• Describing how a token economy works for most of the answer and never addressing power, consent or whose goals are being served.
• Ignoring the benefits, so the answer reads as one sided and cannot reach the top band, which needs a supported judgement.
• Confusing social control with social influence or with obedience research from a different topic.
• Listing applications without explaining the conditioning principle behind them, which leaves the answer descriptive and unrewarded on the knowledge criterion.
• Claiming token economies cure mental disorder or rehabilitate offenders, when the evidence is about behaviour inside the institution.
• Muddling negative reinforcement with punishment when describing how a school or prison system works.
• Ignoring who decides which behaviours are worth rewarding, which is the evaluative point that usually earns the higher band.
• Treating socially sensitive research as a synonym for unethical research, when a well conducted study can still be socially sensitive because of what its results are used for.
• Reciting ethical guidelines such as consent and debriefing with no reference to the social consequences of the findings themselves.
• Forgetting the group level harm, for example the effect on a culture or on people with a diagnosis, and writing only about the individuals tested.
• Concluding that sensitive research should not be done, which ignores the cost of the knowledge not gained and rarely scores as a supported judgement.
• Defining the debate correctly and then never applying it, which leaves the answer as textbook recall.
• Treating reductionism as automatically bad; the credited argument is that reducing to one level buys testability and treatment at the cost of completeness.
• Confusing determinism with fatalism, or with mere predictability of behaviour.
• Naming a study as evidence for a debate without ever saying what it found.
• Reciting the research ethics code when the question asks about practitioner standards; a research code and professional regulation are different documents with different enforcement.
• Asserting that people with a mental disorder cannot consent, when the legal presumption is capacity unless the contrary is shown, and capacity is decision specific.
• Calling Rosenhan unethical without saying why, that staff were deceived and could not consent, and without noting the counterargument that the findings were unobtainable otherwise.
• Claiming a practitioner must never break confidence, which ignores the defined limits where there is risk of serious harm.
• Defining reliability and validity and stopping there, with no clinical example attached to either.
• Saying qualitative data is unreliable and quantitative data is valid, which reverses the usual argument and loses the trade off entirely.
• Confusing inter rater reliability, agreement between two clinicians, with test retest reliability, agreement across time using the same measure.
• Treating a double blind procedure as an ethical safeguard when it is a control for expectancy in both participant and researcher.
• Using the term as an insult, so the whole answer argues one way and never explains why researchers deliberately choose to isolate variables.
• Confusing it with determinism, and writing about whether behaviour is caused rather than about the level at which it is explained.
• Describing the dopamine hypothesis at length and never returning it to the debate, so the material scores as knowledge and not as analysis.
• Claiming a holistic account is simply better, without admitting that it is harder to test and rarely yields a treatment that can be trialled.
• Writing two separate descriptions and leaving the marker to do the comparing, which never opens the analysis marks.
• Saying the two manuals are basically the same, with no duration criterion and no concordance evidence to support or refute it.
• Blurring reliability and validity: agreement between two clinicians is reliability, whereas whether the label picks out a real condition is validity, and the same study can look good on one and poor on the other.
• Assuming the American manual is used only in America; British services code with the international system while research routinely applies the American criteria, which is part of why comparison matters.
• Listing the features of science and never applying one to any clinical content, which leaves an answer about philosophy rather than psychology.
• Reading a brain scan as proof of cause, when a difference found in people who are already ill and already medicated could be a consequence of the illness or the drug.
• Writing that psychology is a science because it uses experiments, with no definition of science against which to judge that claim.
• Treating drug therapy as automatically scientific while ignoring that the diagnosis it is prescribed against relies on self report.
• Treating culture as the patient country of birth, so the answer never considers the clinician culture, the language of the interview, or the norms already written into the criteria.
• Assuming every group difference must be bias; some differences are real, and the higher bands reward saying which is which with evidence.
• Using the terms alpha bias and beta bias as labels with no example of what is being exaggerated or what is being minimised.
• Writing about gender differences in the disorder while the question asked about gender differences in diagnosis, or the reverse.
• Reading a concordance figure as the probability that an individual twin will become ill, when it is a rate across pairs in a sample.
• Concluding that it is a mixture of both with no mechanism of interaction, which is the sentence examiners see most and credit least.
• Treating family studies as evidence for genetics when relatives share environments as well as genes, so the design cannot separate them.
• Describing a social explanation with no research attached, so the biological half of the essay carries all the evidence and the answer looks one sided.
• Listing edition numbers and dates with no account of what changed or why, which reads as history rather than as psychology.
• Assuming every revision is progress; broadening criteria also broadens the population offered treatment, which is the socially sensitive reading of the same change.
• Writing that the manual keeps getting more accurate without stating any criterion by which accuracy could be judged.
• Describing changes to therapy with no mention of the evidence that prompted them, so effectiveness never enters the answer.
• Arguing only one side, so the answer reads as a complaint about psychiatry rather than as an assessment of it.
• Confusing this with social influence, and writing about conformity experiments that have nothing to do with legal or clinical power.
• Treating the token economy as a cure, when it changes behaviour on the ward, does not treat the disorder, and often fails to transfer once the person leaves.
• Asserting that drugs are a chemical restraint with no acknowledgement that antipsychotics measurably reduce positive symptoms for many people.
• Describing how a therapy works and never saying whether it works, when the question is about the use of the knowledge in society.
• Claiming a therapy cures depression, when trials report improvement, remission rates and reduced relapse rather than cure.
• Assessing drug therapy without side effects, dropout or relapse, which examiners mark down as one sided evaluation.
• Listing applications with no evidence attached, so the answer never rises above description.
• Treating this as a synonym for unethical research, when the defining feature is the consequence for a wider group rather than harm to the participant.
• Listing ethical guidelines for consent and debriefing instead of discussing what happens to the findings after publication.
• Concluding automatically that the research should not have been done, when the higher bands reward a cost and benefit judgement that weighs both.
• Forgetting that participants with a current disorder are a vulnerable group, so the usual assumption that consent is freely given needs examining.
• Learning the debates as definitions and never applying them to criminology, which produces an answer that would fit any topic on the paper equally badly.
• Confusing determinism with reductionism, when one is a claim about causes being fixed and the other a claim about the level at which behaviour is explained.
• Calling a study unethical because participants were deceived, without saying what harm followed or how the cost was weighed against what the research achieved.
• Listing ethical guidelines in the abstract without tying one to a criminological study, which answers a question about ethics in general instead of this topic.
• Saying a field experiment is unethical merely because consent was missing, without noting that presumptive consent and full debriefing are the usual defences and explaining why they fail in a public place.
• Assuming a laboratory study is automatically ethical, when the London Dungeon participants felt real fear and mock jury work can expose people to genuinely distressing case material.
• Dismissing all mock jury research as worthless, which ignores review evidence that student and community samples usually return similar verdicts.
• Writing that researchers use mock juries because real juries are inconvenient, which misses the legal prohibition that actually forces the design.
• Confusing ecological validity with population validity, so a criticism of the artificial setting is offered as a criticism of the sample.
• Treating reductionism as an insult rather than a level of explanation, so the answer never states what the narrow approach gains.
• Reporting the imaging work as proof that brain damage causes murder, when the design compared existing groups and cannot establish cause.
• Confusing weapon focus with leading questions, when one concerns attention at the time of encoding and the other concerns information supplied after the event.
• Describing each explanation in its own paragraph and never comparing them, which caps the answer however accurate the content is.
• Claiming the Bobo doll work shows social learning causes crime, when the behaviour measured was aggression towards an inflatable toy in a laboratory room.
• Saying biology proves criminality is inherited, which ignores the environmental contribution shown in the same adoption studies and the fact that no gene has been shown to code for offending.
• Calling a study a field experiment simply because it happened outside a laboratory. A field experiment still needs a manipulated independent variable; if the variable is naturally occurring and just measured, it is a natural experiment.
• Writing that Loftus and Palmer proved memory is unreliable. The sample was students, the stimulus was a filmed crash with no consequences, and demand characteristics could explain part of the effect.
• Treating brain imaging as automatically scientific and therefore correct. The scan is objective, but the diagnostic category, the sample size and the leap to causation are all open to challenge.
• Using the word bias without saying who is harmed. A marker wants the direction of the effect, for example that the accented or the black defendant is rated guiltier on the same evidence.
• Treating mock jury studies as if they were real juries. Real juries deliberate, which can amplify a majority prejudice or correct an individual one, and in England and Wales real deliberations cannot legally be researched.
• Writing that all jurors are racist or sexist. The studies report average shifts in ratings under particular conditions, and several show the effect disappearing once the judge gives a clear instruction.
• Saying that a concordance rate below one hundred per cent proves environment matters and stopping there. The sharper point is that twins share environments as well as genes, so the two are confounded.
• Describing monoamine oxidase A as a crime gene. It codes for an enzyme that breaks down neurotransmitters, and the low activity variant raised risk only alongside maltreatment in the studies that found an effect.
• Using the Bobo doll studies as if they were studies of crime. They measured imitative aggression in nursery children in a laboratory, which is a long way from conviction for a violent offence.
• Listing the cognitive interview components without saying why each works. Context reinstatement supplies retrieval cues, while reverse order and change of perspective disrupt schema driven reconstruction.
• Claiming Loftus proved memories are always false. Her work shows memory is reconstructive and suggestible under particular conditions, and witness memory for central detail is often accurate.
• Treating the ethical interview as merely being polite to a suspect. It is a defined framework adopted after documented false confessions, not a change of tone.
• Assuming a fall in institutional incidents means a fall in reoffending. Token economies change behaviour where the tokens are, and reconviction data is the measure that matters.
• Describing informed consent as satisfied because the offender signed a form. The examiner wants the point that the power imbalance and the link to parole make refusal costly.
• Confusing anger management with general counselling. It is a structured cognitive behavioural programme with rehearsal and role play, aimed only at offending driven by anger arousal.
• Describing eyewitness research without saying what society did with it. The statement is about use, so the answer needs a procedure, a guideline or a courtroom practice.
• Quoting exoneration statistics as though every wrongful conviction were caused by memory failure. Mistaken identification is the commonest single factor, not the only one.
• Answering that juries should be abolished. Examiners want specific evidenced adjustments such as instruction wording, identification procedure or expert evidence.
• Treating socially sensitive as meaning only that participants might be upset. The larger risk falls on the wider group, whose members were never in the study at all.
• Concluding that socially sensitive research should not be done. Examiners reward the balanced position that avoidance carries its own social cost.
• Confusing socially sensitive research with unethical research. A study can follow every guideline and still be socially sensitive in how its results are used.
• Defining the debate and then never returning to child psychology, which leaves an answer that could have been written about any topic area.
• Treating nature and nurture as a contest to be won, when the attachment evidence is used to argue that an inherited tendency still needs an environment to develop in.
• Confusing reductionism with determinism, since one is about the level at which behaviour is explained and the other about whether it could have been otherwise.
• Treating protection as the only ethical duty, so the answer argues that children should simply not be studied. Excluding them is itself a breach of participation rights and leaves policy on day care and schooling built on adult data.
• Listing generic ethical guidelines learned for the foundations paper with no mention of the convention and no mention of childhood at all, which cannot reach the top band on a child psychology question.
• Saying children cannot give consent and stopping there, instead of distinguishing legal consent, given by a parent, from assent, given by the child.
• Confusing socially sensitive research with unethical procedure: a study can be run impeccably and still cause harm through the way its findings are used.
• Writing that a meta-analysis is just a review of studies and stopping there. The examiner wants the specific difficulty of combining results that were never designed to be combined.
• Claiming observer bias is removed by using more than one observer. Two observers trained together can share a bias, so agreement shows reliability rather than accuracy.
• Treating pooled national figures that rest on very few studies as solid averages for that country.
• Confusing inter-rater reliability with validity, and offering a correlation between observers as proof that the categories measure attachment.
• Writing that reductionism is bad and holism is good. Marks come from weighing what the simplification buys against what it hides, not from choosing a side in advance.
• Treating the procedure as a test of the child. It classifies a relationship, and the same child can come out differently with a different caregiver.
• Describing the episodes at length and never reaching the debate, which turns an evaluate question into a description answer.
• Muddling the disorganised category with the insecure resistant one, or crediting the extra category to Ainsworth herself.
• Treating Ainsworth as a straightforward test of Bowlby, when her work grew out of her own Uganda and Baltimore observations and addresses individual differences he barely specified.
• Saying Bowlby claimed children attach only to the mother. Monotropy claims a hierarchy with one main figure, and he accepted a network of secondary attachments.
• Describing the two theories in turn with no comparative language, which caps a compare or analyse answer well below the top band.
• Using the adult love quiz of Hazan and Shaver as proof of continuity without noting that it was a self-report questionnaire printed in a newspaper and relied on memory of childhood.
• Listing features of science without attaching a single one to child psychology, so the answer could have been written for any topic on the specification.
• Assuming that a large pooled sample makes a study scientific. Reach is not control, and a meta-analysis inherits every weakness of the studies inside it.
• Confusing universal with identical: the majority pattern is shared, the proportions of the insecure patterns are not, and that difference is the nurture half of the argument.
• Calling a cross-cultural comparison a laboratory experiment. Culture cannot be allocated to participants, so the design is quasi-experimental and confounding variables remain.
• Saying a study is culturally biased simply because it was carried out in one country. Bias is a claim about the instrument and the interpretation, and the reasoning has to be spelled out.
• Reading the within-country variation finding backwards and concluding that culture makes no difference to how children are raised.
• Treating collectivist and individualist as explanations in themselves rather than as labels that still need a mechanism, such as how often an infant meets strangers.
• Claiming the gender content is about differences between male and female researchers, when the examinable material is gendered outcomes and the assumption that care is a mother's job.
• Arguing that universality proves attachment is innate, ignoring that every sampled culture also contains mothers, infants and caregiving of some kind.
• Using the Romanian studies as evidence about maternal deprivation, when institutional care involves privation, poor nutrition and lack of stimulation together.
• Writing that the debate is now settled as an interaction and then offering no mechanism, which reads as a way of avoiding the question.
• Reversing the direction of the German and Japanese findings, which turns a good evaluation into a factual error.
• Describing the original theory as simply wrong. It was influential, testable and partly supported, and the marks lie in showing how it was refined.
• Treating quantity of day care and quality of day care as one variable, when the later evidence separates them and finds different effects.
• Assuming correlational day care findings show causation, when families who use more care also differ in income, stress and employment.
• Forgetting that the original hypothesis concerned prolonged absence of care, not an afternoon in a nursery.
• Treating social control as automatically sinister. An intervention that gets a child communicating or a family supported is also control, and the judgement has to weigh benefit against autonomy.
• Writing about autism without saying what any named intervention actually involves, so the evaluation has nothing to attach itself to.
• Claiming that behavioural intervention cures autism. It targets particular skills and behaviours, it does not remove the condition, and outcomes vary widely.
• Confusing the ethics of running a study with the ethics of applying its findings, which is the distinction this part of the specification is built on.
• Listing the three applications and never naming a study or a finding, which turns the answer into an opinion piece about whether day care is good for children.
• Treating the Lovaas result as proof that applied behaviour analysis works, when children were not randomly allocated, the outcome measures were school placement and intelligence scores, and autistic self-advocates object to goals built around appearing typical.
• Claiming day care damages children, when the evidence is correlational, the effect on behaviour is small and quality of care is tangled up with family income.
• Confusing deprivation, where an attachment formed and was broken, with privation, where none formed, and so using Bowlby on the forty-four juvenile thieves where the Romanian adoptee work is needed.
• Writing about ordinary ethics, consent, deception and the right to withdraw, as though that were the whole of social sensitivity, and missing the consequences for the wider group.
• Concluding flatly that socially sensitive research is unethical and should be banned, which throws away the counter-argument and caps the answer in the lower band.
• Misreporting the vaccine study, for instance saying it proved a link or inventing a sample size, which costs credit because the example is the evidence.
• Treating the word gap finding as an established cause of later attainment rather than a correlation from a small sample.
• Writing a general essay on debates illustrated from social or cognitive psychology when the question names health psychology, which leaves the application marks unreachable.
• Listing every debate briefly instead of developing two or three, because breadth without development sits in the lower bands for extended responses.
• Treating the debates as pure evaluation and forgetting the knowledge marks, so the essay never explains what the debate actually claims.
• Confusing an issue with a debate, for example presenting a research ethics rule as though it settled the argument about whether animal research is justified.
• Saying animal research is unethical because animals cannot consent and stopping there, when the marks are in the regulatory framework and in the benefit set against the cost.
• Mixing up the codes: the British Psychological Society code of human research ethics covers human participants, while animal work is covered by legislation and by separate guidance on research with animals.
• Describing aversion therapy as simply a drug that stops people drinking, without the conditioning, where the emetic or the disulfiram reaction is an unconditioned stimulus paired with alcohol until alcohol itself triggers nausea.
• Assuming aversion therapy is effective. Reported abstinence figures come largely from uncontrolled follow-ups of self-selected private patients, relapse is common once the pairing is not maintained, and the therapy addresses the drinking rather than the reasons for it.
• Treating generalisation as all or nothing. Findings about receptor action transfer between species reasonably well, while findings about motivation and social context transfer worst, and a good answer says which is which.
• Confusing practical issues with ethical ones. Cost, access, equipment and attrition are practical; consent and harm are ethical, and questions normally ask for one or the other.
• Writing that brain scans show addiction, when they show associations between activity or binding and reported use, so the direction of the effect usually remains unknown.
• Leaving the issue as a statement that research is difficult, without saying what it does to validity, reliability or generalisability.
• Using reductionist as an insult. The marks go to candidates who say what the reduction achieved before saying what it left out.
• Confusing biological reductionism with biological determinism. One is a claim about the level of explanation, the other about whether behaviour is caused rather than chosen.
• Naming no mechanism, so the answer says dopamine causes addiction without saying which drug acts on which system in which way.
• Writing as though only biology reduces, and forgetting that a laboratory experiment is itself a reduction of a complex situation to two variables.
• Writing two descriptions and calling it a comparison. Without whereas, in contrast or similarly doing real work, the comparison marks are not accessible.
• Getting the conditioning label wrong, usually by calling the removal of withdrawal symptoms a punishment instead of negative reinforcement.
• Saying a gene causes alcoholism. Twin concordance falls well short of complete, so inherited vulnerability is a risk and not a destiny.
• Using one study as evidence for both explanations without saying which part of the finding supports which account.
• Listing features of science with no health psychology content attached, which loses every application mark on offer.
• Treating equipment as proof of science, when a scan is objective measurement only if what it measures is clearly defined and cautiously interpreted.
• Calling a correlational survey an experiment because it reports numbers and a significance test.
• Confusing reliability with validity, so replication is offered as proof that the measure captures addiction rather than as evidence of consistency.
• Saying research is culturally biased without naming what is biased, whether the sample, the measure, or the definition of problem use.
• Treating culture and country as the same thing and ignoring variation within a nation, which is exactly what the urban and rural comparison demonstrates.
• Claiming gender is irrelevant here because the specification does not require it, when the specification invites the argument and a developed case about differences in drug taking or in treatment earns credit.
• Assuming higher recorded rates among men reflect behaviour alone, when under-reporting and under-diagnosis among women are equally plausible.
• Ending with the claim that it is both, with no account of how the two combine, which examiners read as avoiding the question.
• Reading a heritability estimate as the proportion of one person's addiction that is genetic, when it describes variation in a population under particular conditions.
• Assuming a family pattern proves genetics, when families share income, stress, drinking habits and modelling as well as genes.
• Forgetting that the cross-cultural evidence belongs to this debate, since differences within and between countries bear directly on how much environment contributes.
• Writing a history of drug laws with no psychology in it, so the answer never names an explanation, a study or a mechanism and cannot be credited as psychology.
• Treating the development as a straight line in which each new explanation replaced the last, when biological, learning and cognitive accounts are all still in use and most treatment services deliberately combine them.
• Asserting that anti-drug campaigns work because they are well intentioned, when the published evaluations of large media campaigns are mostly disappointing and one large national campaign was associated with no reduction in use at all.
• Using social control to mean anything a government does, so the answer drifts into politics and never names a psychological technique or the theory it rests on.
• Assuming control is automatically wrong, when the examiner wants the benefits as well, including fewer acquisitive offences, safer supply and access to treatment people would not otherwise reach.
• Confusing aversion therapy with punishment delivered after the event, when the pairing has to be close enough in time for an association to form at all.
• Describing a therapy in detail and never mentioning evidence, which leaves the answer scoring on knowledge alone with nothing for evaluation.
• Claiming that a therapy cures addiction, when trials report retention, abstinence at a follow up point or reduced use, and a proportion of people leave treatment early.
• Confusing aversion therapy, which pairs the substance with something unpleasant, with contingency management, which rewards evidence of abstinence.
• Treating socially sensitive as a synonym for unethical, when a study can follow every guideline and still be sensitive because of what is done with its results.
• Listing generic ethics such as deception and debriefing without connecting them to drug research, where the live problems are the confidentiality of illegal acts and capacity to consent.
• Concluding simply that the research should not be done, which throws away the counter argument the examiner is waiting for.
• Assuming self report data are accurate, so the answer never raises social desirability or the under reporting it produces.
• Revising the debates as abstract definitions, which produces an essay full of terms like holism and nomothetic and empty of psychology, since no study is ever named.
• Using the same study for every debate, so Milgram is stretched to cover ethics, determinism, culture and gender without the details changing to fit each argument.
• Treating the debates as an extra topic to learn at the end of the course rather than as the arguments already running through each topic area.
• Describing a study at length and then adding one sentence of debate, which inverts the balance the extended question rewards.
• Writing that Milgram broke the law, when the professional code has no legal force for human participants while animal licensing genuinely is statutory.
• Claiming his participants could not withdraw, when they could and some did; the accurate criticism is that the verbal prods made withdrawal psychologically difficult.
• Arguing that animals cannot consent so all animal research is unethical, which skips the cost benefit judgement the specification actually wants weighed.
• Treating debriefing as a cure that cancels harm already caused, rather than as a partial remedy that does not undo distress.
• Listing guidelines as a memorised set with no study attached, which earns knowledge marks and no application or evaluation marks.
• Writing that a study is not valid because it was done in a laboratory, without saying which aspect of validity is threatened and by what feature of the setting.
• Confusing reliability with validity, so consistent scores on a poor measure are treated as proof that the measure is a good one.
• Describing a sample as random when it was opportunity or volunteer, since newspaper advertisements and passing students are not random selection.
• Offering improvements that are not practical, such as testing an entire national population, when the examiner expects a change a researcher could genuinely make.
• Ignoring the ethical and financial limits that shape what design was possible, which is part of the practical picture being assessed.
• Treating reductionist as a term of abuse, when the examiner rewards a weighed judgement in which the simplification is shown to buy something.
• Confusing reductionism with determinism, since an explanation can be simple without claiming the behaviour was unavoidable.
• Saying that a study is holistic because it used several variables, when holism means the whole cannot be understood by its parts at all.
• Naming dopamine as the cause of schizophrenia with no reference to the evidence, when the drug and amphetamine findings are what make the claim arguable.
• Writing two separate descriptions one after the other and calling it a comparison, with no sentence that puts them against each other.
• Comparing whole areas of psychology in the abstract instead of comparing their explanations of the one behaviour the question specifies.
• Claiming the biological explanation is proven by brain scans, when the murderer scanning study was quasi-experimental and cannot show that the brain differences came first.
• Losing the studies inside the argument, so the essay compares ideas fluently but names no researcher and earns little knowledge credit.
• Arguing that the subject is a science because it uses experiments, without saying what makes an experiment scientific.
• Treating objectivity as achieved whenever numbers are produced, ignoring that researcher bias enters through the choice of measure and the interpretation of scores.
• Confusing falsification with being proved wrong, when the requirement is that a theory could in principle be refuted by an observation.
• Stating that there is no paradigm without explaining that the areas disagree about what should be studied, which is the point Kuhn was making.
• Using ethnocentrism and cultural bias as interchangeable words, when one is a particular form of the other.
• Confusing alpha and beta bias, most often by describing research that treats male findings as applying to everyone as alpha bias when it is beta bias.
• Treating any study with a single country sample as worthless, rather than stating which conclusion overreaches and to whom the findings can still be applied.
• Assuming a cross-cultural comparison is unbiased because it sampled several nations, when the measure itself may have been designed in one of them.
• Writing about culture as though it means nationality, ignoring subcultural and generational variation within a country.
• Writing the debate as a contest to be won, when the specification asks for the role of both and the mark scheme rewards interaction.
• Reading a heritability estimate as the proportion of an individual’s behaviour caused by genes, when it describes variation within a population under particular conditions.
• Treating adoption studies as a clean separation of the two influences, forgetting the prenatal environment shared before the child was placed.
• Naming epigenetics with no mechanism, when the creditworthy point is that experience can alter gene expression without altering the sequence itself.
• Listing the five approaches with no chronology, so nothing in the answer is about development; the content is about change over time and needs dates and a direction of travel.
• Judging early work only against modern ethics and stopping there. Milgram's obedience work in 1963 is part of why the British Psychological Society guidance was strengthened, so it belongs inside the story of change.
• Claiming psychology became a science at one moment, when behaviourism, cognitive psychology and neuroscience each redefined what counted as evidence and the earlier approaches did not disappear.
• Treating social control as automatically sinister. Credit is for weighing benefits such as safer wards and lower rates of self harm against costs such as loss of autonomy.
• Describing a token economy as a cure for schizophrenia. It manages behaviour and self care; it does nothing for hallucinations or delusions, which is what the drug treatment targets.
• Using a single example. A question worth several marks expects at least two sources of control drawn from different approaches.
• Describing the study and stopping. The marks are for the application, so the second half of each paragraph must say who now does something differently.
• Confusing this with social control. An application intended to benefit the individual, such as therapy or better interviewing, is not the same as behaviour shaped in an institution's interest, and the two are examined separately.
• Writing that research proves a policy works, when the underlying evidence is often correlational or laboratory based and needs hedging.
• Treating socially sensitive and unethical as the same word. A study can meet every guideline on consent and debriefing and still be socially sensitive because of how its results are used.
• Assuming an ethics committee settles the issue, when a committee can judge procedure in advance but cannot control media reporting or policy use years later.
• Using Milgram as the only example. That study is better used for deception and protection from harm, while this content needs findings with consequences for a social group.
• Assuming determinism means free will does not exist in any psychological perspective: soft determinism accommodates conscious choice within a bounded framework, whereas humanistic psychology rejects determinism entirely.
• Believing that holism is inherently superior to reductionism: holistic explanations often lack empirical testability and cannot easily be subjected to controlled scientific trials, making biological reductionism indispensable for pharmacological breakthroughs.
• Equating ethical issues solely to participant harm: social control and ethical debates extend beyond lab consent to the long-term societal consequences of publishing racially or culturally biased research (socially sensitive research).

1. 1Step 1: Map each core topic (Social, Cognitive, Biological, Learning) onto the key debates using a comprehensive grid matrix.
2. 2Step 2: Master scientific philosophy definitions (Popperian falsifiability, Kuhnian paradigms, empirical validation, replicability) and match them to psychological research methods.
3. 3Step 3: Practice paired evaluations contrasting reductionist and holistic treatments for specific disorders (e.g., depression or schizophrenia).
4. 4Step 4: Draft timed, synoptic 16-mark and 20-mark essays focusing on synoptic links, explicit debate terminology, and evaluative conclusions.

• 📋Short-answer application questions (4-6 marks): Applying a specific issue (e.g. social control or ethnocentrism) to a novel novel stem or scenario.
• 📋Structured synoptic evaluative questions (8 marks): Assessing a specific debate (e.g. reductionism vs holism) using research from two psychological approaches.
• 📋Extended response synoptic essays (16-20 marks): Broad evaluation of overarching philosophical debates (e.g. 'Evaluate whether psychology can be considered a science') requiring integrated evidence across the entire qualification.