The German psychologist Hugo Munsterberg (18631916) set the agenda for future generations of forensic psychologists in nominating evidence and the workings of the courtroom among the leading concerns of forensic psychology. This article will give a flavor of the contribution of psychological research to these two important areas of forensic psychology.

The Psychology of Evidence

Eyewitness testimony

The work of psychologists with respect to the accuracy of eyewitness testimony provides an excellent example of the interplay between psychological theory, applied psychological research and real-life decisions in the courtroom. This area of work is important, given that is now clear that the fallibility of human memory lies at the heart of a catalog of wrongful convictions based on the testimony of eyewitnesses. The work of psychologists, such as Whipple and Munsterberg, at the beginning of the twentieth century was concerned to establish the reliability of eyewitness evidence as an issue for law and psychology alike. Later research, from the 1970s onwards, has sought to offer an understanding of eyewitness testimony within more sophisticated psychological models of human memory.
One model of the functioning of human memory suggests that there are three stages to the process of memory: (1) the acquisition of information about the world (i.e. perception); (2) retention (i.e. storing information in memory); and (3) retrieval (i.e. remembering). If this three-stage model is matched against the process of witnessing an incident, then the stages can be defined as follows: (1) witnessing the incident (acquisition); (2) waiting to give evidence (retention); (3) giving evidence (retrieval). Further, within these three stages there are different kinds of variables to consider; these variables can be classified as social, i.e. to do with interactions between people; situational, i.e. concerning the setting in which the event occurred; individual, as with the characteristics of the individual witness; and interrogational, i.e. regarding the process of eliciting testimony. As can be seen from Table 1 ,a range of factors has been shown to be related to the accuracy of eyewitness memory. Clearly it is beyond the scope of this article to discuss these in detail – the topic of eyewitness memory has generated a large body of evidence, summarized in a number of texts. A discussion here of the key issues will give an understanding of the field.
Acquisition: nature of the witnessed incident The effect on memory of the nature of the witnessed incident has been considered in several studies. In a typical study, observers are shown one of two short filmed incidents; both films involve the same people, but in one film the content will be violent in nature, perhaps including scenes of physical assault. Analysis of the participants’ recall of the filmed events typically reveals that memory is significantly better for the nonviolent rather than the violent incident. An excellent example is provided by a study in which separate groups of observers saw different versions of a filmed bank robbery: in one scene shots were fired but no one was hit; in the other a small boy was hit and was seen bleeding from a head wound. Those witnesses who saw the violent scenes gave less accurate recollection of the incident.

Table 1 Variables in eyewitness memory research

Social Situational Individual Interrogation^
Attitudes Complexity of event Age Artists’ sketches
Conformity Duration of event Alcohol use Computer images
Prejudice Illumination Cognitive style Identity parades
Status of interrogator Time delay Personality Mugshots
Stereotyping Type of crime Race Instructions
Target factors Sex Photofit
Weapon Training

From this and other experimental evidence, can it be concluded that serious crimes lead to poorer memory performance, perhaps particularly so when a weapon is involved? In fact, it seems likely that it is the degree of violence that is important: less threatening events that engage the witness’s attention in a stimulating manner will produce satisfactory memories of the event. The research evidence suggests that witnessing extremely violent and personally threatening crimes may well be associated with poorer performance.
Retention: discussion between witnesses If there is more than one witness to an incident, should witnesses be allowed to discuss events before giving testimony? The research evidence suggests that group discussion can lead to a more complete description of an event, but at the price of introducing a source of error. While having the potential to prompt memory, discussion can bring about errors of commission – that is, testimony that includes errors related to items that were not seen in the original incident. Further, group dynamics can influence witnesses to change their testimony to agree with other members of the group. On balance, it is a moot point whether discussion between witnesses is desirable in the interests of producing accurate testimony.
Retrieval: leading questions A body of research, perhaps most notably by the American psychologist Elizabeth Loftus, has shown that even the most subtle changes in question wording can influence witness testimony. In one study, witnesses to a filmed automobile accident were asked to give estimates of the speed of the cars when ‘they … into each other’.
Different groups of witnesses were asked this question, with the verb being one of ‘contacted’, ‘hit’, ‘bumped’ or ‘smashed’. The witnesses’ estimates of the speed increased, following the order given above, from 31.8 miles (51.2 km) per hour to 40.8 miles (65.6 km) per hour. Further, when later requestioned, the witnesses who had been asked about the ‘smash’ were more likely to say in error that there was broken glass at the scene of the accident.
Witnesses can also incorporate misleading information into their testimony: in another study, observers were asked to judge the speed of the car ‘as it passed the barn’, although no barn had been seen. One week later, the observers were asked a set of questions about the film, including whether they had seen a barn; more than 17% of the ‘primed’ observers said that they had seen the fictitious barn.
It appears that misleading information is most likely to have an influence on testimony if the source of the information is seen to be of high status, or if the witness thinks the police have special knowledge about the incident, or if the misleading information relates to peripheral detail rather than central events.
Witness factors Witness factors, such as intelligence, personality, anxiety, age and sex, are active across all three stages of memory. Reviews of the research conclude that there is little evidence of such factors having any consistent effect on eyewitness memory. However, there are two exceptions to this general conclusion: the first is the research evidence on the relationship between witness confidence and accuracy; the second is the now substantial body of work on children as witnesses. The research evidence on the former suggests that it is prudent to be cautious about witness confidence: confidence in the accuracy of one’s memory is not an index of accuracy. With respect to children as witnesses, there are grounds to argue that, given sensitive questioning, even young children can be reliable witnesses, even to traumatic events.

Confession evidence

For a person to confess to a crime is, surely, an admission of guilt beyond dispute. While people unconnected with a case do come forward and make false confessions, police investigation often shows beyond doubt that such admissions are indeed fake. Such confessions are sometimes given by people who are seeking attention or who are psychologically distressed in some way. On other occasions, suspects will knowingly give a false confession during police interrogation; they may do this to relieve themselves from the stress of interrogation or to protect another person. Such false confessions are given in the (sometimes mistaken) belief that the police investigation will inevitably show that their confession is untrue.
In the examples given thus far, the individuals, unless severely disturbed, are aware that their confession is false. Is it possible that a suspect could wrongly confess to a crime he or she did not commit?
The phenomenon of the ‘coerced-internalized false confession’ is seen when the suspect’s own private version of the truth actually changes to come into agreement with the interrogator’s version. How can such a transformation of memory take place? In order to build an understanding of the coerced-internalized false confession it is important to look at the characteristics of both the interrogation and the suspect.
The work of the forensic psychologist Gisli Gudjonsson has played an enormous role in helping to understand the processes that act to produce false confessions. Gudjonsson suggests that the suspect under interrogation is in fact taking part in a highly unusual ‘closed’ social interaction: interrogation occurs in a closed room; the suspect is closed off from his or her normal surroundings; and the suspect’s attention is closed in on the interrogator. Further, the interrogator holds the balance of power and therefore controls all aspects of the interaction. Trained in in-terrogational techniques based in the psychology of persuasion, the interrogator’s focus is on gaining a confession from the suspect.
Gud jonsson suggests that some suspects will cope with the stress of interrogation by actively dealing with the situation and resisting the pressure to confess. However, other suspects will become passive and take a helpless stance, seeking to avoid confrontation with the interrogator, so reducing their emotional stress. Suspects who react in this passive way may be overly suggestible with regard to the interrogator’s persuasive tactics.
The concept of suggestibility has a long history in psychology: in 1900, Alfred Binet wrote La Sugges-tibilite, and since then topics and journal articles have regularly appeared in the psychological literature. In the context of an interrogation, interrogative suggestibility is the extent to which people come to accept as true the messages conveyed by the interrogator during formal questioning. Thus, both publicly and privately, the suggestible suspect agrees with the account offered by the interrogator, even though it differs from their own memory for events.
Some individuals will be more susceptible than others: Gudjonsson and his colleagues have shown that interrogative suggestibility is related to poor recall of the original event (perhaps because of intoxication at that time), acquiescence during questioning, low self-esteem, low intelligence, high emotionality and anxiety, and a high need for social approval.
The picture that emerges from the research is that, through an interaction of situational cues and suspect characteristics, the likelihood of a false confession can be increased.

Psychology in the Courtroom

The psychological study of courtroom processes covers a great deal of ground. This aspect of forensic psychology spans both criminal and civil courts; incorporates diverse areas of law, including juvenile and family law, mental health law and discrimination law; and considers the problem of competency to stand trial. However, perhaps the area in which the study of the courtroom is best seen lies in the study of the jury.
The jury is a natural arena for psychological research: the jury must listen to and evaluate evidence; each juror must formulate an opinion; then the jurors must discuss and reach a collective decision. The cognitive and social processes involved are both intriguing and critical for justice to be done. The problem lies in conducting the research: for obvious reasons, it is not feasible that studies of jurors are conducted during trials, nor in some jurisdictions is it permissible to interview and seek to debrief jurors after a trial. Thus, a great deal of our knowledge of jury and juror behavior comes from, so-called, ‘mock jury’ studies in which controlled simulations of trials are conducted and ‘mock jurors’ can be closely observed while key variables are introduced into proceedings.

Extraevidential influences

While evidence presented in the courtroom should inform juror decision-making, there is concern that jurors are subject to extraevidential influences. There are three likely influences: pretrial publicity; witness confidence; and juror sentiments and prejudices.
Pretrial publicity In typical studies, mock jurors are shown newspaper cuttings containing details about, say, a defendant’s criminal record or their retracted confession. When compared to mock jurors who read newspaper stories that did not contain the biasing details, the ‘exposed jurors’ are more likely to return a guilty verdict after a mock trial. Of course, judges can instruct jurors to ignore pretrial publicity, but this is probably easier to say than to do. In practice, the selection of jurors not exposed to the pretrial influence is the optimum strategy; again, this is probably easier said than done.
Witness confidence Is the confident witness more likely to be believed than the uncertain witness? The weight of evidence from experimental studies suggests that mock jurors are more likely to return a guilty verdict following testimony from confident witnesses. Conversely, jurors place less credibility on the testimony of the less confident witness. Given, as noted previously, that witness confidence is not always a reliable index of accuracy, a belief in a positive relationship between confidence and accuracy is of concern.
Interpersonal perception Are jury decisions and views of the evidence influenced by emotion, sympathy and prejudice? A body of research has pointed to several sources of influence. For example, the perceived attractiveness of the defendant can lead to a favorable outcome, but an attractive defendant with no justification for their offense may be seen in a more severe light. Similarly, defendants of high socioeconomic status are seen as less blameworthy for their offence; while a gender bias suggests that females are less likely to be found guilty for reasons of insanity. Finally, children are seen to be less reliable witnesses in court than adults (which may not be the case).
Some researchers have attempted to estimate the degree of influence these interpersonal factors have in real cases. The evidence actually gathered from real jurors strongly suggests that jurors’ decisions are primarily influenced by the evidence, and less so by the perceived characteristics of victims and defendants.


Selection There are formal selection criteria, such as age and eligibility to vote, for jury service. The issue here is whether, given these basic formal criteria, all members of the public are equally suited for jury service. The case might be made, for example, that individuals prejudiced against certain groups would not make impartial jurors in given trials. It would be prudent therefore to have the capacity to ‘screen out’ unsuitable jury candidates.
In practice, screening procedures are typically concerned with questioning prospective jurors with regard to their exposure to pretrial publicity of the case in hand; forming a judgment with respect to their attitudes towards the particular offense being tried; or discovering if they know anyone involved in the case. Other areas of concern might be the attitudes and prejudices of the potential juror, as well as his or her age, sex and education. The selection of jurors is often an emotive issue, taking a great deal of the court’s time, and raising the ethical issue of whether the legal process should use scientific jury selection, informed by psychological assessment. The proponents of scientific selection argue for an empirically based method of selecting jurors; opponents hold that selection is unsafe and contrary to the principles on which the jury system stands.
Competence Jurors must assimilate a range of evidence presented during the trial, and understand the judge’s presentation of the issues, before they reach a verdict. Several studies have looked at juror comprehension of the evidence in complex cases and their understanding of judges’ instructions, typically finding that jurors can struggle to make sense of proceedings. The implication of these findings is both concern if jurors do struggle for comprehension, and the need to clarify information presented in court.
Reaching a verdict The real process of jury deliberation during a trial is, of course, impossible to research. The nearest that can be achieved is either to observe mock jury deliberations or (where allowed) to interview jurors after a trial. The extant research hints at the group dynamics of structure, process and decision-making by 12 people thrown together in a highly charged situation. From appointing a leader, polling views, debating the evidence, recounting the judge’s instructions and seeking to apply the law to the evidence, the jurors will be in a state of continual negotiation. There may be disagreements and conflict, and some jurors will become anxious or angry. It may be that issues become personal, even insulting; some individuals may refuse to revise their opinion; some jurors may even refuse to participate, sitting apart from the others; while subgroups may form. Yet, further, the deliberation may go beyond the jury’s brief: jurors may blame the defendant for interfering in their everyday lives; and while sentencing is not their concern, jurors may reach a verdict on the basis of what would happen should they say ‘guilty’.
The process of jury deliberation is clearly important. Research suggests almost one-third of jurors do change their views; however, the best predictor of the eventual verdict is the balance of opinion at the onset of the deliberations.
As is often the case in this arena, it is impossible to know how often juries make wrong decisions. It is the case that innocent men and women have been sentenced because of a wrong jury decision; and it is likely that guilty people have been set free. A comparison of the judge’s views with the jury verdict may offer an insight into the frequency with which jurors reach a verdict that is, at least, at odds with an expert opinion. American research has compared the convictions and acquittals of judges and juries, finding that in more than 3500 trials, the judge and jury agreed in 78% of cases. When they disagreed, the jury acquitted in 19% of cases when the judge would have convicted, leaving only 3% of cases in which the judge would have acquitted but the jury found the defendant guilty. English research has reported similar findings: most of the time judges and juries agreed, and in instances of disagreement the jury was more lenient.
There are several possible reasons for juror leniency: a defendant with no history of offending may be given a greater benefit of doubt; the jury may feel sympathy towards the defendant; or, in some celebrated cases, the jury disagreed with the law. Of course, to follow the point made above, the jury may reach an incorrect decision because of a failure to comprehend the high standard of proof needed to reach a decision, or because of a lack of understanding of the evidence or the judge’s instructions.

The Strength of the Research

As will be clear from the above, a great deal of research findings have accumulated that are relevant to the psychology of evidence. However, a great deal of this research, perhaps particularly in the area of eyewitness memory, has been carried out in laboratory settings. The conduct of laboratory-based research has obvious advantages for the researcher: the laboratory allows the investigator to control what the observer sees; it allows the investigator to select observers, say by age or sex; and it allows the investigator to control how the process of recall and recognition is structured. Indeed, the issue of control is central to laboratory research. By controlling as many variables as possible, the investigator can comment, with a high degree of confidence, on the effects of the particular variables of direct interest in the experimental setting. Thus, for example, in laboratory settings observers generally recall violent incidents less accurately than nonviolent incidents. The key issue that then arises centers on the concept of ‘generalizability’; that is, to what degree does the laboratory-based research inform our understanding of real-life cases? A review of the issue has been highly critical of experimental studies, pointing to the limitations of the laboratory (such as the use of staged crimes and filmed sequences, and observer awareness of the focus of the research) as a means by which to understand the complexities of the task facing a real eyewitness. The authors suggest that the limitations of laboratory research are such as to minimize their realism and hence their applicability to real life. Indeed, this point can be followed by questioning the professional wisdom of psychologists acting as expert witnesses and presenting such findings in real cases.
To support the points made above, a field study looked at a real-life case in which 13 of 21 witnesses to a shooting incident were reinterviewed 4-5 months after the event. The performance of the witnesses varied in a number of ways from that which would have been predicted from the laboratory research, particularly with respect to the effects of violence on memory.
Should the findings from laboratory-based studies be disregarded in favor of field studies? The problem is that field studies are open to criticism on the grounds that are the very strengths of laboratory research; that is, the sacrifice of control in the field setting. Thus, in the field study noted above, it is not known to what extent witnesses conferred with each other, how accurate the missing witnesses would be, and whether media coverage had any influence on what witnesses had to say when reinterviewed.
Is there a solution to this apparent impasse? While it is unlikely that the two positions can be fully reconciled, there are two steps that can be made.
First, with regard to the extant research, it is evident that not all the factors studied have an equal strength with respect to their relationship with the accuracy of eyewitness memory. Three categories of research findings have been suggested: (1) reliable and strong factors that show consistent effects on eyewitness memory; (2) reliable and moderate factors that show effects in some studies but not in others; and (3) weak or noninfluential factors that have little or no effect on witness accuracy. In real-life cases the psychologist acting as an expert witness may wish to temper his or her expert testimony in respect of strong rather than weak factors.
Second, looking to the design of future research, it is evident that the optimum approach is to gather data using a variety of research methodologies, including laboratory studies, case studies, field studies and archival studies. Studies that use a range of methodologies and produce consistent findings will considerably reduce the, quite proper, reservations about the applicability of research.


The historical emphasis in much of the psychological research has been to illuminate the fallibility of human performance, so alerting the criminal justice system to the potential for error. More recently, psychologists have turned to a more constructive task in considering ways to aid the gathering of evidence. In particular, psychologists are studying ways of enhancing the quality of witness testimony through more sophisticated interviewing procedures. The ‘cognitive interview’, which uses memory-retrieval mnemonics to enhance recall, is one aspect of the developing field of investigative interviewing. For the future, forensic psychologists seem set to give increasing attention to the development and refinement of constructive approaches in applying psychology to law.

Next post:

Previous post: