H567/03 Section B Option 2 - Memon and Higham (1999) and police interviews

Retrieval principles and cognitive interview techniques

Encoding specificity proposes that retrieval is helped when cues overlap with information present when the memory was encoded. Context reinstatement invites a witness to mentally recreate the environment and feelings at the time. It may supply retrieval cues without physically returning to a distressing scene. Multiple routes to retrieval motivate using more than one way of recalling an event.

The original cognitive interview included four techniques:

These are techniques, not four mandatory stages in a rigid order. The enhanced cognitive interview also emphasises rapport, witness control, open questions, sufficient pauses and minimal interruption. Better communication may allow a witness to use the retrieval techniques more effectively.

The review and effectiveness of components

Memon and Higham organised their 1999 review around four themes: effectiveness of individual techniques, appropriate comparison interviews, measures of memory performance, and interviewer training. They discussed existing experiments and field research to identify theoretical and practical problems. There is no single review-wide independent variable, dependent variable, sample or numerical result.

Studies attempting to isolate techniques did not consistently show that each technique alone improved recall. Context reinstatement appeared especially promising; changing order or perspective alone had not clearly demonstrated effectiveness. Evidence was sparse, however, so the review did not establish that only one technique can ever work. Components may interact: rapport and pauses can help context reinstatement, while the full combination may differ from isolated instructions.

Imagery also needs care. Imagining a detail can make it harder later to distinguish an experienced event from an imagined one. Interviewer suggestions increase that danger. The review called for research on component combinations and comparison with free recall, rather than treating an increased amount reported as a complete explanation of the mechanism.

The contribution to collection of evidence is therefore both practical and methodological: use retrieval support cautiously, and test what actually produces better evidence. The 1999 conclusions describe the evidence available then; they are not a claim that later research stopped or that the technique is ineffective in every setting.

Choosing a useful comparison interview

Early studies compared the cognitive interview with a 'standard' police interview. This can answer a practical question about the whole procedure, but standard interviews vary greatly. They can differ in interruptions, open questions, time, rapport, interviewer training and motivation. A cognitive-interview advantage over a poor comparison does not isolate a special mnemonic technique.

The guided memory interview uses contextual reinstatement and imagery. Comparing it with the full cognitive interview can help investigate whether additional techniques contribute beyond shared context cues. Studies of recognition and recall need to be distinguished: a technique that improves reporting details need not improve identification from a line-up.

A structured interview offers rapport, an uninterrupted narrative, time to answer and good non-leading questioning, while omitting the cognitive interview's special retrieval techniques. This is a stronger control for testing the added contribution of those techniques. Equalising training, time and communication reduces alternative explanations. Remaining differences in interviewer ability and adherence still need measurement.

For a worked design, randomly allocate comparable witnesses to enhanced cognitive or structured interviews after the same staged event. Train both sets of interviewers in rapport and good questioning, provide comparable time, and check actual adherence from recordings. Differences are then more plausibly connected to the retrieval techniques, though staged-event witnesses may not represent distressed witnesses to real crimes.