You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
Eyewitness testimony can send a person to prison or set them free, and yet human memory is one of the least reliable instruments the justice system relies on. It does not record events like a video; it reconstructs them, and the reconstruction is vulnerable to decay, distortion and the very questions used to retrieve it. If memory is fragile, how should police interview witnesses so as to recover as much accurate detail as possible without contaminating it? This is the third topic of the criminal option and the first drawn from the cognitive area. In Background we examine how evidence is gathered from witnesses and suspects, why standard interviewing fails, and how psychologists designed the cognitive interview to do better. In Key research we study Memon and Higham's (1999) review of the cognitive interview — its component techniques, the evidence for its effectiveness, and the methodological problems of evaluating it. In Application we design a strategy for police interviews. Throughout, this topic builds directly on the reconstructive-memory tradition (Loftus and Palmer) you met in Component 02.
| This lesson covers | OCR H567 Component 03, Section B (Criminal) topic | AO focus |
|---|---|---|
| Gathering evidence from witnesses and suspects; reliability of memory | Collection of evidence — Background (Cognitive) | AO1; AO3 evaluation |
| The cognitive interview and its four techniques; the Enhanced CI | Background — improving witness recall | AO1; AO2 explanation |
| Key research: Memon & Higham (1999) review of the cognitive interview | Collection of evidence — Key research | AO1 findings; AO3 evaluation |
| A strategy for police interviews | Collection of evidence — Application | AO2 application; AO3 judgement |
The specification is referenced descriptively throughout; consult the official OCR H567 specification document for the exact published wording. This lesson develops AO1 (knowledge of witness memory, the cognitive interview and Memon and Higham's review), AO2 (applying the techniques to a police-interview strategy) and AO3 (evaluating the cognitive interview and the problem of assessing it).
Memory is reconstructive. When a witness recalls a crime, they do not replay a stored recording; they rebuild the event from fragments, filling gaps with expectation, inference and knowledge of how such events usually go — a process shaped by schemas. This reconstruction is open to error at every stage: at encoding (the event may be brief, distant, poorly lit, or witnessed under the stress and narrowed attention that a weapon can cause — the "weapon focus" effect); at storage (memories decay and can be altered by later information); and at retrieval (the cues available, and the way questions are asked, shape what is recovered).
The classic demonstration is Loftus and Palmer's (1974) work from Component 02: changing a single word in a question ("smashed" versus "hit") changed witnesses' estimates of a car's speed and even led some to "remember" broken glass that was never there. This is the phenomenon of misleading (post-event) information and leading questions: the interview itself can implant or reshape memory. The lesson for evidence-gathering is stark — a badly conducted interview does not merely fail to retrieve memory; it can contaminate it, creating confident but false testimony.
Before psychologists intervened, the typical police interview had well-documented flaws. Officers tended to ask a rapid series of short, closed, direct questions ("What colour was the car?"), frequently interrupting the witness, following a fixed checklist rather than the witness's own memory, and sometimes using leading or suggestive phrasing. This style has three problems: interruptions break the witness's concentration and train of recall; closed questions retrieve only what the officer thinks to ask about, missing detail the witness could have volunteered; and suggestive questioning risks contamination. The result is testimony that is both incomplete and, at times, distorted.
The cognitive interview (CI), developed by Fisher and Geiselman in the 1980s, is a structured interview technique grounded in the psychology of memory. Its aim is to increase the amount of accurate information recalled without increasing errors, by exploiting what cognitive science knows about how retrieval works. It rests on two principles: that memory is made of multiple associated cues, so more retrieval routes recover more information; and that recall is better when the retrieval context matches the encoding context. From these principles come four core techniques.
| Technique | What the witness is asked to do | Cognitive rationale |
|---|---|---|
| Report everything | Report every detail, however trivial or seemingly irrelevant, without editing | Small details can cue further memories; witnesses wrongly self-censor "unimportant" information |
| Context reinstatement | Mentally recreate the environment and internal state at the time (sights, sounds, feelings, weather) | Matching retrieval context to encoding context provides cues (encoding specificity) |
| Change the order | Recall the events in a different sequence (e.g. reverse order) | Reduces reliance on schema-driven expectation of how events "should" flow, exposing detail |
| Change perspective | Describe the event from another person's point of view (e.g. another witness) | Provides new retrieval cues and disrupts schema-based reconstruction |
A later development, the Enhanced Cognitive Interview (ECI), added social and communication components — building rapport, letting the witness lead, minimising anxiety, adapting language, avoiding interruption and judgemental cues — recognising that the interaction matters as much as the memory techniques. The ECI is more effective but also more demanding to conduct.
A caution built into the technique. "Change perspective" is the most controversial component, because asking a witness to imagine what another person saw risks inviting them to invent — to report inference rather than memory. This tension, between recovering more detail and preserving accuracy, is exactly what a careful review of the CI must weigh, and it is central to Memon and Higham's analysis.
So far we have concentrated on witnesses, but the collection of evidence also includes interviewing suspects, and here the psychology takes a darker turn. Interrogation techniques designed to secure confessions can, under the wrong conditions, produce false confessions from innocent people — one of the leading causes of wrongful conviction. Coercive, accusatorial styles that assume guilt, minimise the seriousness of confessing, maximise the apparent strength of evidence, and prolong questioning under stress can lead a suspect to confess to escape an intolerable situation, or even to come to believe they committed the act. Vulnerable suspects — the young, those with learning disabilities, the highly suggestible or compliant — are at particular risk. This is why modern practice in England and Wales moved decisively away from confession-seeking interrogation toward an information-gathering model. The PEACE framework (Preparation and planning, Engage and explain, Account, Closure, Evaluate) embodies this shift: interviews aim to obtain a full, accurate account rather than to extract a confession, using open questions and rapport rather than pressure. The cognitive-interview techniques for witnesses and the PEACE ethos for suspects share the same underlying insight: accurate evidence is best obtained by facilitating memory and account-giving, not by pressuring the interviewee, because pressure contaminates.
A further real-world complication is that many witnesses are vulnerable — children, older adults with memory difficulties, people with learning disabilities or trauma. Children's memory can be accurate but is more suggestible: leading questions and repeated questioning can more easily distort it, and young children may try to please an adult interviewer by supplying the answer they think is wanted. This has led to special measures and structured protocols for interviewing children (for example, phased approaches that begin with rapport and free narrative before any specific questions, mirroring the cognitive-interview logic). The general principle Memon and Higham's review reinforces — favour free recall, avoid leading questions, prioritise accuracy — applies with even greater force to vulnerable witnesses, where the risk of contamination is highest and the consequences (in, say, an abuse investigation) are gravest. A sophisticated understanding of "collection of evidence" therefore spans witnesses and suspects, adults and children, and recognises that the same cognitive vulnerabilities that make the cognitive interview necessary also make careless interviewing dangerous.
Full citation: Memon, A. & Higham, P. A. (1999) A review of the cognitive interview. Psychology, Crime & Law, 5(1–2), 177–196.
This key research is a review article, not a single new experiment. Memon and Higham surveyed and critically evaluated the existing body of research on the cognitive interview in order to answer several questions: Which components of the CI are responsible for its effects? How effective is the CI, and how should its effectiveness even be measured? What methodological problems bedevil the research, and what should future work do? A review of this kind is a legitimate and important form of scholarship: it synthesises many studies, exposes inconsistencies, and sharpens the questions — and OCR includes it partly to show that evaluating a body of evidence is itself a psychological skill.
1. Which components work, and do they interact? The CI is a package of techniques, so a central difficulty is disentangling which components drive the gains. Memon and Higham noted that studies vary in which components they include and how they combine them, making it hard to isolate effects. Some components (report everything, context reinstatement) have stronger support; others (change perspective, change order) are more questionable and may, in some conditions, encourage errors or confabulation. The components may also interact, so the effect of the whole is not simply the sum of the parts.
2. Comparison problems — what is the CI being compared with? The apparent superiority of the CI depends heavily on the control interview it is measured against. If the comparison is a poor "standard" interview (rushed, interrupting, closed questions), the CI looks dramatically better; if the comparison is a well-conducted "structured" interview that already uses rapport and open questions, the advantage shrinks. Much of the CI's headline benefit may reflect the deficiencies of the comparison rather than special properties of the CI's memory techniques. This is a crucial evaluative point: the size of an effect is only meaningful relative to a fair baseline.
3. How is "effectiveness" measured? The CI reliably increases the quantity of correct information recalled — but it can also increase the quantity of incorrect information and confabulated details. So whether the CI is "better" depends on how you weigh more correct detail against more errors. Memon and Higham stressed the importance of the accuracy rate (the proportion of recalled details that are correct), not just the raw amount of correct information, because for the justice system a witness who reports more but is less reliable per statement may be no more useful, and potentially more dangerous.
4. Training and administration. The CI (especially the ECI) is complex and demands skilled interviewers. Studies differ enormously in how much training interviewers received and how faithfully they applied the technique, which introduces huge variability and makes results hard to compare or generalise to real police work, where training may be brief.
Memon and Higham concluded that the cognitive interview is a valuable tool that generally increases the amount of correct information recalled compared with a standard interview, but that the research base is beset by methodological problems: inconsistent operationalisation of the CI, unfair or variable comparison interviews, insufficient attention to accuracy rate as opposed to raw quantity, and wide variation in interviewer training. They argued that future research needed clearer definitions, fairer comparisons, and closer attention to why and when the CI works, rather than simply demonstrating that it produces more information. The overall verdict is cautiously positive but methodologically demanding — a model of how to evaluate an applied technique.
| Strength | Limitation |
|---|---|
| Synthesises a large, scattered literature and exposes real methodological weaknesses others had overlooked | As a review it produces no new data of its own; its conclusions depend on the quality of the studies reviewed |
| Sharpens the crucial distinction between quantity of recall and accuracy rate, of direct value to the justice system | Reviews can be selective; the questions asked and studies emphasised reflect the reviewers' framing |
| Practically important: guides both interview practice and how future research should be designed | Rapid subsequent development of the CI/ECI means some specifics date quickly |
| Improves scientific rigour by insisting on fair comparisons and clear operationalisation | Does not settle which components are essential — leaves the "active ingredient" question open |
Broader evaluation. The review's enduring value is methodological. It teaches that an applied intervention cannot be judged by headline gains alone: the comparison must be fair, the outcome measure must capture accuracy and not just volume, and the technique must be defined and delivered consistently. These lessons connect to the psychology as a science debate (objectivity, replicability, operationalisation) and to validity and reliability as recurring themes of the option. Its limitation is inherent to reviews — it inherits the flaws of the literature it surveys and reflects the authors' choices about what to include and emphasise.
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.