A key part of this process is stereotyping. This involves noticing something about the person and, based on this, attributing them to a particular group of people, then assuming the person has the characteristics of that group. An example might be noticing that the defendant has several tattoos and stereotyping him or her as criminal because it is believed that all criminals have several tattoos.
Lippmann (1922) defines stereotypes as follows:
‘For the most part we do not see and then define; we define first and then see. In the great blooming buzzing confusion of the outer world we pick out what our culture has already defined for us and we tend to receive that which we have picked out in the form stereotyped for us by our culture.’
Stereotyping has been shown to influence the recall and interpretation of information. Duncan (1976) showed white American participants a film without sound of two men talking intensely, then one of the men shoved the other. When it was the black man pushing a white man, 75% of the participants perceived the push as violent behaviour, but for the other way around only 17% saw it as violent behaviour.
In another experiment, Snyder and Uranowitz (1978) had participants read the details of the life of ‘Betty K’, with one of two endings to the story. The story concluded that she was either married or a lesbian. When participants were asked to recall the number of boyfriends she had had at school (a fact not given in the story), the answers were based on the stereotypes. For example, with the lesbian ending, the number of boyfriends recalled was lower. Also participants remembered more of the information consistent with the stereotype.
The practical application here is that people may perceive differences and similarities among group members even when they do not exist. It is important to note that these snap judgements and stereotypes are not particularly accurate. However, the juror or lawyer’s use of these processes of person perception is the same as impressions made by people everyday.
Key aspects of stereotyping
The text that follows outlines some of the key aspects of stereotyping and snap judgements.
Within society, physical attractiveness is prized highly in both sexes, and others treat such people positively. They receive less harsh sentences for the same crime and they do not fit the stereotype of a criminal.
In an experiment, Sigall and Ostrove (1975) asked 120 participants to recommend sentences for burglary or fraud either with or without seeing a photo of the defendant. The photos were either physically attractive or unattractive. The physically attractive photo received shorter sentences.
Saladin et al. (1988) showed participants eight photos of men and asked them to judge how capable they considered them to be of committing each of two crimes: murder and armed robbery. Overall, the attractive men were considered less likely to have committed either crime than the unattractive ones. This attractiveness effect is strongest with serious but non-fatal crimes such as burglary, and when females are being judged (Quigley et aL, 1995). An exception to this rule occurs if people are considered to be using their attractiveness for illegal material gain, such as in performing confidence tricks or in fraud.
This effect applies not only in the psychology laboratory but also in the actual courtroom. Downs & Lyons (1991) analysed the fines and bail payments of 1,500 defendants accused of minor crimes and asked police officers, who were not told the purpose of their assessments, to rate them on attractiveness. The results showed that, even when the seriousness of the crime was controlled for, attractiveness was negatively correlated with the amount of bail or the fine imposed by the judge. Attractiveness, however, appeared to be irrelevant when imposing fines or setting bail payments for more serious crimes (McKelvie & Coley, 1993). Stewart (1980) sent observers into courtrooms to rate defendants on various traits, including physical attractiveness. The results were similar to those found in simulated jury studies: when seriousness of the crime and race were controlled for, there was a significant negative correlation between attractiveness and length of sentence. In other words, attractive people received shorter sentences than unattractive people accused of a similar crime. In this study only one rating of attractiveness was made. Stewart later extended his research by asking the observers sitting in the courtroom to rate defendants not only on physical attractiveness but also on cleanliness, neatness and quality of dress. Using a seven-point scale, defendants were rated as attractive—unattractive, dirty—clean, sloppy—neat and well dressed—poorly dressed. Over a two-year period, 60 defendants were observed and it was found that those who were perceived as attractive, clean, neat and well dressed were treated with greater leniency than were those rated lower on these four dimensions. However, although attractiveness affected sentencing, it did not affect whether or not the defendant was convicted or acquitted.
As a variation, Kerr (1978) has shown that the physical attractiveness of the victim can influence the verdict reached by the jury. In the experiment, mock jurors read about the theft of a car from a physically attractive or unattractive woman (with a photograph attached to the transcript). The physically attractive victim was more often supported with a guilty verdict, regardless of how much care she had taken to protect the vehicle (like using a steering lock or not).
Although attractiveness is the main variable that has been shown to influence judges and juries, other features do have an effect. Berry & McArthur (1985) found, in a mock trial study, that defendants with ‘baby’ faces were more likely than those with mature ones to be found guilty of negligence in not warning customers of the possible dangers of the products they were selling. Those with mature features were more likely to be considered guilty of deliberately deceiving their customers by telling lies about their products.
However, not all studies have found a relationship between defendant features and jury decisions. Bull & McAlpine (1998), in a review of such research, point out that it is quite possible that many studies finding no effect of appearance on judgements may have been rejected for publication.
Again stereotypes are involved, but, in this case, there is evidence of racial bias. In a mock trial experiment, it was found that white students rated black defendants guiltier than white defendants, particularly when the victim was white (Pfeifer and Ogloff, 1991). Research on actual trials in USA found that black defendants received longer sentences than white defendants for the same offence (Stewart, 1980).
The OJ Simpson trial (1994/5) was viewed as a race-related trial by many. Stably and Walker (1997), who worked for the legal defence team, noted the psychological processes at work among viewers of the trial. They highlighted group identity and the ‘availability heuristic’. The availability heuristic is where we tend to recall only certain information when making a decision. For example, white Americans saw the obvious guilt of Simpson because of the physical evidence against him, while the African-Americans saw police misconduct.
Skolnick and Shaw (1997) used 213 students in a mock murder trial and varied the defendant’s race and celebrity status. The former influenced the verdict, but not the latter.
Gordon et al. (1988) showed that the type of crime could complicate the effect of race. They varied both the race of the defendant (black or white) and the type of crime supposedly committed (embezzlement or burglary). The white embezzler received a significantly longer sentence than the black embezzler but this situation was reversed for the burglary. It was suggested that this was because racial stereotypes for particular crimes were operating, with participants holding the black burglar and the white embezzler more responsible for their crimes.
Mahoney and Dixon (1997) report their research with the ‘Brummie’ (Birmingham) accent. Generally this accent is perceived of as low status. The researchers played a two-minute recording of a police interview with a suspect to 119 white non-Birmingham students. There were a number of versions of the recording played to different groups, including ‘Brummie’ accent versus ‘non-Brummie’, and white versus black ‘Brummie’. The students had to assess whether they felt the speaker was guilty of a particular crime. The ‘Brummie’ accent was perceived as more guilty than the ‘non-Brummie’, and the black ‘Brummie’ was perceived as most guilty, particularly for ‘blue-collar’ crime (like theft).
Seggie (1983) found similar results in Australia. Using recordings of suspects arguing their innocence with either ‘standard’ (‘received pronunciation’: RP) or ‘broad’ Australian accents, or an ‘Asian’ accent, participants were asked to assess the level of guilt on particular crimes. The ‘standard’ accent was seen as most guilty on ‘white-collar’ crimes (like fraud), but the ‘broad’ accent was most guilty for ‘blue-collar’ crimes.
The following variables have all been found to influence the juror’s decisions: ‘powerful speech’ (that is, speaking clearly and without hesitation) by the defendant, and their demeanour in court (for example, sitting up straight and paying attention to the evidence) (Hans and Vidmar, 1986). For example, during her trial in 1997, Louise Woodward (see page 61) showed no emotions, which the USA jury interpreted as a display of indifference. She was found guilty of ‘second degree murder’ (Harrower 1998).
III. THE WITNESS
A. Instructions to the Jury to Disregard Testimony
Almost every lay person who watches enough television is aware of the possibility that a barrister may ask a witness a prejudicial or accusatory question or may mention facts that are deemed inadmissible by the judge. The judge then instructs the jury to ignore the remarks that have been made. What has intrigued many lay people and psychologists alike is the issue of whether or not it is possible to erase the effects of such remarks from the opinions of the individual jurors. This issue is of great controversy in jurisdictions that have so-called "rape shield" statutes, where evidence describing a rape victim’s previous sexual history is deemed inadmissible. If a witness reveals that a rape victim has had many previous sexual relationships, is it sufficient for the judge to instruct the jury to disregard this fact when determining the truth of the victim’s claim that she had not consented to sex?
In a series of mock trial experiments, it emerged that judges’ instructions had little effect on jurors. The mock juries were given the details of a homicide case. When the prosecution relied on weak evidence, none of the juries found the defendant guilty. However, when a recording of an incriminating telephone conversation was played to the juries, and the judge subsequently instructed the juries that the tape was not legal evidence and so should be disregarded, the conviction rate rose to 33%. Similar experiments have shown that jurors tend to be unaware of the influence of revealed inadmissible evidence, but that in some mock trials, jurors can consciously rebel against a judge’s instructions and give the inadmissible evidence undue emphasis in their decisions.
Psychological research has discovered possible ways of diminishing the effects of inadmissible evidence. It appears that initial juror impressions and opinions are hard to erase. Thus, the first strategy is to attempt to ‘inoculate’ the jury. If, prior to the trial, the jury is informed about the various types of inadmissible evidence, then the effects of such evidence arising during the trial appear to diminish. In one experiment, when mock jurors were shown videotapes of trials and were subsequently reminded by the experimenter of the importance of the presumption of innocence, the number of guilty verdicts returned (59%) hardly differed from the number returned by jurors who were given no such admonition (63%). However, when the admonition was given before the tape was viewed, only 37% of jurors returned a guilty verdict. Similar research has shown that jurors are more likely to find an accused rapist guilty if they are told to disregard information about the victim’s sexual history before they actually hear the information rather than after.
A second strategy to combat the effects of inadmissible evidence is the use of videotaped evidence. As well as allowing inadmissible testimony to be edited out, videotapes allow for, perhaps, speedier trials. Although research indicates that there is little, if any, difference in the impact of videotaped evidence and live evidence, the procedure is not popular. Critics point to the rights of due process, the prevention of jurors from observing the live reactions of the defendant to the witnesses and the costs involved. In summary, judges’ instructions to juries to disregard inadmissible evidence is frequently ineffective. However, psychological research indicates that the matter can in some measure be redressed by inoculating juries appropriately.
B. Instructions to the Retiring Jury
Within the trial itself, there are a number of influences on jury decisions. One is the inclusion of non-evidence. This is evidence that the judge decides is legally inadmissible. In other words, it is presented to the jury, then the decision is made that it cannot be included. Wolf and Montgomery (1977) found that this type of evidence, which should be ignored, is noticed most by the jury.
A key influence is the judge’s instructions during the case. Generally research has shown that many juries are confused by the legal technicalities. Elwork et al (1981) showed mock jurors a videotaped trial including the judge’s instructions. Over one-third of the jurors made the incorrect legal verdict. When the judge’s instructions were clarified, the incorrect verdicts dropped to around 10%.
Instructions given before the case began had a better effect on the juries’ ability to integrate facts and the law. Overall, however, written instructions are better than verbal ones.
C. Judicial Inconsistency
Judicial discretion allows judges leeway in handing down sentences. Furthermore, in District Courts and the Special Criminal Court in Ireland, the judge has the final word on the outcome of the trial, as there is no jury. Therefore, there is some concern over the consistency of judges’ interpretations. For example, in the United States, fifty federal judges were given the details of a case of theft and possession of stolen goods and were asked to sentence the offender. Their sentences ranged from probation to 7½ years in prison. Although it is fair to assume a certain degree of variability among the interpretations and attitudes of different judges, it is difficult to generalise such research findings to different jurisdictions and different crimes. Moreover, there is considerable philosophical and criminological debate over what precisely constitutes an appropriate sentence in any given case. Thus, the establishment of appropriate sentencing norms must take place before the issue of judicial inconsistency in sentencing can adequately be addressed by research psychologists.
A related issue—concerning legal reasoning—concerns not just sentencing inconsistencies but also the verdict a judge may be required to return, as described by Irish psychologist, Paul O’Mahony. Unless a defendant pleads guilty, the truth of a case may never be established. Thus, legal reasoning will operate with its highest standard of proof—‘beyond a reasonable doubt’. However, when this standard of proof is utilised by a judge, a question arises as to the judge’s subjective assessment of the facts of the case. What may appear to be established beyond reasonable doubt to the judge may well be reasonably doubted by other impartial people. This is important in the light of the issue of judicial inconsistency, in that, even with the strictest standard of proof available to the legal system, different judges may well come to different conclusions.
V. THE JURY
A. Juror Selection
The issue of jury selection illustrates that psychological research only slowly alters judicial practice. In the United States, jury candidates for potential death penalty trials are screened so that opponents of capital punishment are systematically eliminated from the jury. However, a wealth of research exists to indicate that supporters of capital punishment are more likely to return a guilty verdict. Supporters of the death penalty are more likely to be authoritarian, to feel that the legal system is ‘too soft on the criminal’, that the police don’t make mistakes, and are more concerned with crime control than with the right of the accused to be presumed innocent until proven guilty. In short, juries composed of supporters of capital punishment are not only capable of returning a guilty verdict, but are actually predisposed to do so. However, in 1986 the US Supreme Court overturned a lower court decision that such juries represent a biased group. Some researchers have argued for ‘scientific jury selection’ where prosecution-biased jurors are identified by means of questionnaires and surveys. However, some have questioned the ethics of allowing questions to ‘trap’ unsuitable jurors. Furthermore, it is as yet unclear whether scientific jury selection is practical, useful or even possible. Whereas proponents of capital punishment are biased in favour of guilty verdicts, opponents of capital punishment may refuse on moral grounds to return a verdict of guilty—even if they feel that the accused did actually commit the crime. However, capital punishment is relevant to only a minority of trials in those countries not to have abolished it. Thus, in trials where capital punishment in not applicable, who lays down the criteria for determining an ‘unsuitable’ juror? In such trials, it appears that the general attitudes and characteristics of jurors are poor predictors of their verdicts. In general, therefore, the current debate among psychologists about scientific jury selection has a long way to go before it can be usefully applied to the courts system.
The voir dire (American term, known as peremptory challenges in Britain). This is the process of producing an impartial jury. Blunk and Sales (1977) have provided a useful overview. The voir dire examination (literally to see and to say) enables the trial judge to discharge his obligation to dismiss veniremen (those summoned for jury service) found unfit for trial duty. Counsel are allowed a set number of peremptory challenges when the judge refuses to act. The voir dire examination by counsel is proper so long as it tries to discover the veniremen’s state of mind directly or indirectly related to the trial in question.
Counsel have available to them a number of strategies for the voir dire.
1. They can accept the first 12 veniremen without question or ceremony. A variant of this is accepting them and, histrionically, declining with a grand gesture the opportunity to question and explicitly stressing faith in the jury system in general and in these 12 jurors in particular. This approach is taken in the belief that the jury’s first impression of counsel is critical in increasing their receptivity toward his arguments.
2. The voir dire can be used to the full in the belief both that in depth questioning will reveal prejudices and biases and that veniremen’s first impressions of counsel are important and will be well revealed by a lengthy examination. There are obvious difficulties: it is hard to elicit prejudices; verbal reports of beliefs may lack predictive validity; counsel may investigate the wrong prejudices or veniremen may simply lie; perhaps most important, veniremen may resent questioning and so develop an unfavorable attitude to counsel. Hence:
3. An indoctrinational strategy has developed around 2, above. It involves hypothetical questions to analyze potential areas of prejudice so as to ingratiate counsel with veniremen and make them aware of, and test their reactions to, certain aspects of the case (counsel’s theory of the case, acceptable defences, etc.). The questions posed are intended to influence the jury; counsel has little interest in the answers. This method is much deprecated judicially.
Kassin and Wrightsman (1983) constructed a juror bias scale to measure pre-trial bias amongst jurors. It is unlikely, however that a judge would agree to the questionnaire.
Items from the juror bias scale (after Kassin and Wrightsman 1983)
1 If a suspect runs from the police then he probably committed the crime.
2 In most cases where the accused presents a strong defence, it is only because of a good lawyer.
3 Out of every 100 people brought to trial; at least 75 are guilty of the crime with which they are charged.
4 Defence lawyers don’t really care about guilt or innocence; they are just in business to make money.
5 Generally the police make an arrest only when they are sure about who committed the crime.
6 Circumstantial evidence is too weak to use in court.
7 Many accident claims filed against insurance companies are phoney.
8 The defendant is often a victim of his own bad reputation.
9 If the grand jury recommends that a person be brought to trial, then he probably committed the crime.
1 A defendant should be found guilty if 11 out of 12 jurors vote guilty.
2 Too often jurors hesitate to convict someone who is guilty out of pure sympathy.
3 The death penalty is cruel and inhumane.
4 For serious crimes like murder, a defendant should be found guilty so long as there is a 90 per cent chance that he committed the crime.
5 Extenuating circumstances should not be considered — if a person commits a crime, then that person should be punished.
6 Too many innocent people are wrongfully imprisoned.
7 If a majority of the evidence — but not all of it — suggests that the defendant committed the crime, the jury should vote not guilty.
8 If the defendant committed a victimless crime like gambling or possession of marijuana he should never be convicted.
B. Other Issues
Broadly speaking, the field of Social Psychology concerns the study of the individual in a social context. Within social psychology, the specific field of group dynamics theory can be applied to many different types of small groups, including juries. However, given the broad context of such research, the principles arising from research in the area are largely global in nature and, so, difficult to apply to specific situations. A complete review of these principles would be outside the remit of this article. Nonetheless, a few universals are worth referring to briefly. For example, when a group appoints a spokesperson, it is more often the first member to speak, the person sitting at the head of the table, or a man. Each of these effects occur when juries appoint a foreman. Secondly, juries—like other groups—are prone to group polarisation, where deliberations usually strengthen individual members’ initial impressions. Thirdly, the larger the group, the more discussion of issues that takes place prior to reaching a decision. Some American courts have experimented with five- and six-person juries, and subsequent research on the matter suggests that the smaller juries are more likely to hold secret ballots and less likely to examine the evidence in detail. These principles of group dynamics suggest little in the way of improving the effectiveness of juries, except perhaps that a twelve-member jury is more likely to give the verdict more rigorous deliberation than the six-person version. This is particularly interesting given that the US Supreme Court has repeatedly argued that no difference exists between the two sizes, with regard to decision-making.
Psychological research has a potentially enormous contribution to make to the effectiveness of court trials. However, the various strands of research in the area are not without criticism. Furthermore, it appears that some members of the justice system—particularly the police—have a poor general understanding of the field of psychology and the expertise of psychologists.
One of the major sticking points concerns the predictive validity of the various research methods employed. The use of mock trials in particular has been the subject of some harsh criticism. Critics argue that there is a substantial difference between a university researcher’s students discussing some hypothetical court case and the deliberations of a genuine jury. However, there are two main points to be remembered here. First of all, it is impractical—not to mention illegal—to observe juries in the midst of their deliberations. An attempt by Kalven and Zeisel (1966) to record the actual deliberations of jurors (with the consent of trial judge and counsel, but the jurors were not informed) was followed by many states making it a criminal offence to record jury proceedings or in any way to interfere with the jurors pre-verdict. In the UK, it is usually held that revelations about juries can amount to contempt of court. However, in any science, direct observation of events in the real world reduces the ability with which the researcher can clearly identify the causes and effects of the observed occurrences. In any field of scientific inquiry, an environment of ‘controlled laboratory conditions’ is essential. The essence of a scientific experiment is the testing of hypotheses under controlled conditions, and then the application of inferred principles onto events in the real world. Human beings find it difficult to see themselves as the potential subject of some scientific experiment. However, psychology—as the scientific study of human beings—is not free from the obligations of the scientific method and forensic psychology is no exception. Thus, although generalisations from the laboratory to the real world must be made with care, videotaped or mock trials, together with mock juries, are important tools for learning about the trial system.
Methods of studying juries. Baldwin and McConville (1980) point out that most methods are indirect and, hence, inevitably imperfect.
1. Autobiographical accounts by jurors are unavoidably partial and idiosyncratic, conveying a wide spectrum of views, from the adulatory to the appalled.
2. The publicly expressed views of legal practitioners. These are much less varied, usually being positive, but inevitably do not allow generalization.
3. The systematic collection by researchers of jurors’ opinions. These suggest a general lack of understanding of legal concepts.
4. Simulated jury panels. The great advantage of this approach is that their deliberations can be recorded and analysed and systematic variations introduced. In the usual version, a mock jury is presented with a reconstruction of a real trial, perhaps a tape of actors reading from a trial transcript. A further development is the shadow jury, which is put into the courtroom alongside, as it were, the real one. Their subsequent deliberations are then recorded (e.g., Diamond and Zeisel 1974). While the shadow jury is much more exposed to courtroom reality than simulated juries, neither has to determine the fate of the defendant in question. However, the generalisation from the laboratory to the real world need not always be deficient. For example, in the research on eyewitness testimony discussed above, college students served as mock witnesses. College students are generally of above average intelligence and are accustomed to having their performance examined. If substantially poor memory performance is obtained with college students, then a convincing case can be made that the performance of actual eyewitnesses will be at least as poor, if not poorer. In any case, it certainly appears that eyewitness testimony is problematical, but it would take a large number of virtually identical court trials to establish this problem clearly or to suggest solutions. In real life, no two court trials are identical. In summary, no psychologist would claim that the use of mock trials, videotaped trials, or mock jurors is going to establish without doubt the outcome of every possible trial in the future. However, the use of such experimental methods provides us with a wealth of knowledge which can reasonably be applied to the interpretation of genuine court trials.
5. The performance of juries is pitted against the views of other participants in trials. This method was pioneered by Kalven and Zeisel (1966). Baldwin and McConville (1980) consider it to be “the most illuminating;” nevertheless, the judges or counsel studied “recollect in tranquillity,” whereas the jurors made their decisions in the context of the actual trial.
Research on jury selection. In the USA prior to 1968, the “key man” system was often used. This meant that key members of the community recommended people for jury service. In 1968, The Jury Selection and Service Act required voting lists to be used as sources of jury pools so that the representativeness of juries improved. But the young, racial minorities and the poor are still significantly underrepresented (Hans and Vidmar 1982). This is linked to administrative discretion, usually vested in court clerks, and there is some evidence of informal discrimination at the initial selection stage, fewer women than men being called.
Courtroom challenges to jurors are extremely rare in Britain (one in the early 80’s was thought to be the first for 150 years) but are very much more frequent in the USA, particularly in trials for murder (see below) and in those with political overtones. In the 1972 trial of the Berrigan brothers on conspiracy charges, a group of social scientists conducted community surveys to determine the characteristics of people favourable or opposed to the defendants and then constructed ideal juror profiles so that defence attorneys had much information on which to base their challenges. The same techniques have been used in other celebrated cases. In describing this approach. Hans and Vidmar (1982) raise doubts both as to the effectiveness and the ethicality of such work.
In the USA, about 5 per cent of jurors are removed for cause, that is by the judge, on the basis of voir dire responses. Removals by peremptory challenges (by counsel) are more frequent. In the federal courts, in felony cases, each side is given five challenges, in misdemeanours, two, Rights to challenge are used frequently, particularly by the defence — about three times more often than by the prosecution.
Most simulation studies show no effect of the sex of jurors but a few suggest women to be more defence oriented, except in the case of rape, in which they lean more toward the prosecution in their decisions (Hans and Vidmar 1982). Demographic and attitudinal data show a very limited ability to predict verdicts (Penrod 1980). Greater specificity of attitude questions (e.g., towards rape in a rape case) improves predictions. In principle, lawyers should try to assess case-specific attitudes rather than either general attitudes or demographic variables.
In fact, the evidence on attorney effectiveness in weeding out biased persons is somewhat contradictory (Hans and Vidmar 1982). They cite a study by Zeisel and Diamond (1978) as the only one providing direct information on this issue over a number of cases. The latter arranged for the challenged and rejected jurors in 12 cases to attend the trials concerned and to vote at the end of each concerning their verdict. In addition, Zeisel and Diamond interviewed the empanelled jurors at the end of the trial to get their votes on the first ballot (it is often the case that several rounds of voting are needed to reach a sufficient majority). This enabled them to reconstruct the first ballot vote of the original jurors in the absence of challenges. In two of the 12 trials the difference between the two first ballot votes was large enough to have produced a different verdict in the absence of challenges (i.e., the attorney’s challenges were effective in these cases). There were great variations in attorney’s abilities to eliminate biased jurors. Zeisel and Diamond concluded that, under the usual conditions of limited information about juror attitudes, most attorneys are only marginally effective most of the time in weeding out biased jurors.
A rather different approach to the effectiveness issue was taken by Jones (1987). Noting that jurors frequently distort their replies to questions during the voir dire she asked whether more accurate information was elicited by the judge or by attorneys. The attitudes of potential jurors were assessed both by questionnaire and then verbally in court. She found that attorneys were more effective than judges in eliciting candid self-disclosure from potential jurors and that participants changed their answers almost twice as often when questioned by judges as by attorneys. Jones concluded that the presence of the judge during challenges evokes a considerable pressure towards conformity among jurors to a set of perceived judicial standards. This is reduced by attorney-conducted voir dire, allowing individual biases to emerge and those holding them to be rejected.
Hans and Vidmar’s (1982) overall conclusion seems sensible: jury composition is probably not very important for the normal run of criminal cases but may be for the minority of political or socially controversial trials in which juror attitudes are likely to be polarized and the right to challenge might be crucial. However, in view of the retention of the death penalty for murder in the USA, juror attitudes to the capital sentence have been of great interest, both, to legal practitioners and to researchers.
A special case: death qualification. This is the process by which the courts identify and exclude from capital juries those persons whose views on the death penalty are considered incompatible with the duties of capital jurors (Fitzgerald and Ellsworth 1984). It followed from the case of Witherspoon versus Illinois heard in the US Supreme Court in 1968. The Court decided that those unwilling to impose the death penalty, on principle, regardless of the evidence, should be excluded from the jury concerned. Subsequently, in Hovey versus Superior Court (heard in the California Supreme Court in 1980), a contrasting group was identified — automatic death penalty (ADP) persons, who would always vote for the death sentence, irrespective of the facts of the case.
Fitzgerald and Ellsworth (1984) suggested that death qualified juries (those from which persons who are anti-capital punishment have been excluded under Witherspoon) might be biased against capital defendants. They compared the demographic characteristics and attitudes to capital punishment of a random sample of over 800 eligible jurors in Alamedo County, CA. Just over 17 per cent were “excludable”, including significantly greater proportions of blacks than whites and females than males. Death qualified respondents were consistently prone to favour the point of view of the prosecution, to mistrust criminal defendants and their counsel, to take a punitive attitude to offenders, and to be more concerned with crime control than with due process.
Two further studies by this group of researchers add to the case against death qualification. Both used video-simulated murder trials and included death qualified as well as excludable participants. The first (Cowan, Thompson, and Ellsworth 1984) found that death qualified persons were significantly more likely to vote guilty, both on an initial ballot and after an hour’s deliberation in 12 person juries, when these consisted solely of death qualified participants, than when they were mixed (two to four excludables took part). Mixed juries were more critical of witnesses and were better able to remember their evidence than those which were homogeneous, so that diversity may improve the vigour, the thoroughness and the accuracy of jury deliberations. And Thompson, Cowan, Ellsworth, and Harrington (1984) found death qualified jurors to be more favourable to the prosecution and to express less regret concerning erroneous convictions, but more regret about wrong acquittals, compared with excludables. Horowitz and Seguin (1986) also found death qualified jurors to give the most severe sentences. In a study of actual jurors, Moran and Comfort (1986) reported that those favourable to capital punishment tended to be white, male, married, politically conservative and authoritarian. They reported reaching their verdicts more quickly and participated more in jury discussions than their demographic and personality counterparts.
In a mock murder trial experiment, Ellsworth (1993) used jury-eligible California residents and showed them part of a video of a police officer’s and the defendant’s testimony in an assault trial. Ellsworth found that supporters of the death penalty were more likely to vote guilty immediately before jury deliberations had taken place. Ellsworth argues that attitudes come in a ‘bundle’. Death penalty supporters tended to show more trust in police practices and to be more sceptical of the defendant’s case, compared to those opposed to the death penalty.
Summary of the results from Ellsworth (1993) Mean evaluation of evidence by supporters and non-supporters of the death penalty
Accuracy of witnesses
Scale ranged from 1 to 6; higher number indicates more favourable to evaluation
* = significant difference at p0.05; ** = p0.01.
The evidence against the system of death qualification seems clear and unequivocal, but Turkington (1986) points to the decision of the US Supreme Court of April 1986, in the case of Lockhart versus McCree. By a majority of five to four the Court decided that excluding jurors opposed to the death penalty does not violate a criminal defendant’s right to a fair trial. The Court made it clear that its decision was reached without regard to social science research, which suggests that death qualified jurors are more prone to convict precisely when the evidence is ambiguous. The Court based its decision instead on legal analysis, but the majority did note “several serious flaws in the social science evidence accepted by the lower courts,” asserting that many studies were irrelevant and that all were simulated. In contrast, the four-person minority termed the scientific evidence “unanimous and overwhelming” (a rare state of affairs in the social sciences!). Had the Court ruled in favour of McCree (i.e., if one justice had voted for instead of against) 1,714 death row inmates, convicted by death qualified juries, could have requested new trials (Turkington 1986).
Zeisel (1971) suggested the greater the size of the jury, the greater the risk of a hung jury. A hung jury is one in which a verdict is not possible. There is some evidence that a requirement of unanimity in the decision increases the likelihood that the jury will be hung. Similarly, it is possible that complex cases are more likely to result in failure to reach a decision. In research to test these ideas (Arce et al., 1998), participants eligible for Spanish jury service were selected at random from the electoral register.
Gender was equalised in the juries. Participants viewed the re-enactment of a real-life rape trial including the testimony of eyewitness and forensic experts, opening and closing defence and prosecution arguments, and the judge's definition of the legal terms involved and the decision rule. The participants were randomly assigned to one of a number of juries that were then studied through the use of questionnaires. Among the findings were the following:
.Hung juries deliberated longer.
.Hung jury members employed more assertions in their communications with each other.
.Hung juries tended to manifest more simultaneous interruptions.
.Hung juries report perceptions of intransigence, lack of dialogue and irrelevant deliberations.
.Hung juries tended to use less of the trial evidence.
It is impossible to summarise the studies into the effects of jury size on verdicts and other matters. Fortunately, there is available a meta-analysis which summarises the findings of 17 studies (Saks and Marti, 1997). The following trends were found in the research studies:
.Deliberation time is longer for larger juries.
.Hung verdicts are commoner for larger juries. This is a finding that is really only true for studies that used a mock jury. In studies involving real juries, hung verdicts were rare - they occurred in only about 1 % of instances.
.Guilty verdicts are not more common in large juries.
.For civil cases, smaller juries tend to award more to the injured party.
These trends were found for studies that have a unanimous decision rule (that is, all of the jurors must agree for a guilty verdict). There is evidence to suggest that such juries tend to be evidence driven. For example, they make more references to the evidence, establish more connections between the evidence and legal issues, examine the evidence in detail, and their deliberations are more exhaustive and detailed. Juries operating with a majority decision rule are more driven to reach a verdict. So majority verdict juries are more likely to begin their deliberations with a vote. For example, Arce et al. (1998) compared six- and twelve-person juries that used unanimous decision criteria. The smaller juries made rather fewer references to the evidence and seemed to make fewer pro-defendant arguments.The performance of juries. This has been one of the main preoccupations of researchers into the jury system. Baldwin and McConville (1980) note that some reviewers consider juries, both real and simulated, to be very competent, deciding their verdicts solely on the evidence. But they themselves are much more cautious:
1. The considerable evidence that mock juries are markedly influenced by defendant attributes, and the possibility that this is as true of real life juries, was reviewed above.
2. “Juries are as likely to understand the meanings of the words in the judge’s charge as if they were written in Chinese, Sanskrit or Choctaw” (Frank 1949). Arens, Granfield, and Sussman (1965) have also pointed to the difficulties juries have in understanding the more technical content of instructions. (A comprehensive overview by Ley, 1977, shows the problems patients have with doctor’s communications.) Severance, Greene, and Loftus (1984) have reviewed the considerable evidence concerning juror difficulty in understanding judge’s instructions. They developed a set of standard instructions to improve those revealed by their research as problematic, using legal as well as psychological criteria. The revised set was found to increase significantly the understanding of mock jurors, compared with a standard set, and a survey of experienced judges indicated their strong preference for the revised set.
3. A high proportion of jury acquittals is not seen as justified by other participants (judges, lawyers, etc.). Conversely, in about 6 per cent of convictions doubts are expressed by two or more of the other main participants (including the police), indicating that certain juries were too easily convinced of the defendant’s guilt. (The study by Kalven and Zeisel, 1966, which contains evidence of major judge—jury divergence, was described above.) As Baldwin and McConville (1980) point out, we need to know if these questionable verdicts are evenly distributed, across both defendants and offences, or are biased, for example, against minority groups and toward serious crimes.
4. Even when ordered to do so, jurors do not ignore evidence presented which is then ruled as inadmissible. They are affected by pre-trial publicity, and they do not give a reduced weight to witnesses who qualify their testimony under cross-examination (Monahan and Loftus 1982).
5. It may be that juries are less able than professional judges to avoid the pitfalls of eyewitness testimony.
Baldwin and McConville (1980) conclude that the jury trial is an unpredictable method for discriminating between the guilty and the innocent, and that the reverence accorded the system is misplaced and excessive.
Nevertheless, research data from both the USA and Europe indicate that the “adversarial model” (prosecution, defence, judge, and jury) found in the English speaking world is seen as more fair than the “inquisitorial model” (defence lawyers are present at every interrogation, there are separate judges for the examination and the trial, and no jury) used in most European countries (Monahan and Loftus 1982). Perhaps reluctantly, Baldwin and McConville (1980) agree that alternative systems might be equally imperfect. The jury system seems set for a long lease of life.
There is an assumption that ‘jury deliberation is a reliable way of establishing the truth in a contentious matter’ (Stephenson, p.179), but it has been suggested that most jurors have already decided on a verdict before they retire to deliberate, and that the first ballot usually reveals a majority preference (Kalven and Zeisel, 1966). Pennington and Hastie (1990) disagree with this finding and suggest that the process towards achieving a final agreed verdict is much more complex, while Sandys and Dillehay (1995) surveyed 142 ex-jurors and revealed that an average of 45 minutes was spent discussing cases before a first ballot was taken.
In the jury room discussions it seems that those who talk the most are ‘socially successful’ jurors, men, and the foreperson (Ellsworth, 1993). The longer the jury is out the more likely it is they will acquit (Baldwin and McConville, 1980), and as discussion continues the minority are more likely to agree with the majority view.
Various mathematical models have been developed to explain the processes by which juries reach their decisions, but Pennington and Hastie (1993) suggest an alternative model in which jurors actively construct explanations for the evidence presented to them in order to reach a verdict. This dynamic process is therefore entirely dependent on cognitive processes of selection, attention, interpretation and recall which relate to pre-existing schema, and this explains why two jurors presented with the same evidence can reach quite different verdicts. It also suggests that the influence of pre-trial publicity in terms of constructing possible narratives can be highly significant (see Freedman, Martin and Mota, 1998).
Jury decisions are influenced by three groups of factors: jury composition, the decision-making process, and the characteristics of the defendant.
Characteristics of defendent
Size of jury
Majority influence / conformity
Characteristics / attitudes that influence verdict
Influence within jury
Emergence of leader
Individual differences within juries
But specific research into juror variables (like gender, age, race, occupation or personality) found that they are not related to juror decisions. In other words, the verdict cannot be predicted from the make-up of the jury. Research into juries in Birmingham by Baldwin and McConville (1979) believed that ‘no single social factor’, including social class, led to significant effects on the verdicts returned.
However, one personality type that does seem to influence jury decisions is the ‘authoritarian personality’ type. This is a personality type who is very narrow-minded, against change and holds strict conservative views. It is usually measured using a questionnaire called the ‘California F Scale’.
Using the OJ Simpson trial, Chapdelaine and Griffin (1997) found a correlation between the ‘California F Scale’ and the belief in Simpsons guilt, the perception of the fairness of the trial and the severity of the recommended sentence. Thus, the authoritarian personality’ type were more likely to see Simpson as guilty, believe the trial was fair and recommend longer sentences. However, overall, the decisions were based on an appropriate sentence for Simpson rather than the general belief of guilt.
Other research has found that there is a positive correlation between the trial experience of the juror and the severity of the sentence, irrelevant of the type of prior experience. There is a limited amount of research suggesting that men are less likely to convict a rapist.
Among the jury members, those of higher occupational status were more involved and more influential. Hastie et al (1983) set up 69 mock juries and found that the most talkative were dominant in the group — and these were more often male. The 34—56 age group was also seen as more influential.
Where the background of jurors may influence their views is in their reaction to the evidence. For example, jurors from certain areas (like inner cities) may not be surprised or see it as unusual that the defendant was carrying a weapon.
The difficulty with research in this area is that it is illegal to study the jury at work or question jurors afterwards. This means that any research must be with mock juries. It is generally assumed that the jury decision-making process is similar to that of any small group making decisions.
The decision-making process of any group passes through three stages that can be applied to juries (Hastie et al, 1983).
• ‘Orientation process’ — an agenda is set and the evidence is explored.
• ‘Open conflict’ — where differences of opinion become obvious and there is a focus on disputed evidence. This tends to be a move from facts to emotions.
• ‘Reconciliation’ — either the conflict is resolved and a verdict is given or the majority attempt to persuade the others. The main point is that attempts are made to reduce any previous conflicts.
In the main, the majority opinion tends to predominate. For example, Kalven and Zeisel (1966) found that of 215 juries with a majority view at the beginning of the deliberations, only six changed to the minority decision by the end of the discussions. But there is evidence of a ‘leniency bias’ to favour the defendant, the longer are the deliberations. So if there is disagreement, it is easier to persuade jurors of a not guilty than a guilty verdict during the deliberations.
Currently, juries are required to reach unanimous decisions (that is, they all agree on the verdict). This could take longer than a majority verdict (for example, 10:2), and so increase the possibility of ‘leniency bias’.
In Britain a jury does not have to be in total agreement; a majority of 10 to 2 is sufficient to secure a conviction. In the USA the rules vary from state to state, with respect to both the size and unanimity of juries. Nowadays few states allow a less-than-unanimous verdict in criminal trials, although many do so for civil cases.
The effect of using a majority versus a unanimous verdict system was investigated in mock trials by Nemeth (1977). He asked student jurors to make decisions about the guilt or innocence of a person charged with murder and organised the juries so that they included some members initially in favour of acquittal and some who were initially predisposed towards conviction. Some juries were required to reach a unanimous verdict while others were allowed to reach a verdict based only on a two-thirds majority. Some significant differences emerged between the two groups. Compared to the majority-verdict groups, those who had to reach a unanimous decision debated for longer, were more likely to take account of minority views and were more confident in their final decision. More significantly, they were also more likely to change from their original judgement. Hastie et al. (1983) conducted a similar study using 69 mock juries and requiring either a 12 to 0, a 10 to 2 or an 8 to 4 verdict. Compared to the juries requiring a unanimous verdict, those requiring a majority one spent less time discussing and more time voting and they did not continue any deliberation after the required majority had been reached. As in the previous study, these jurors were less confident and less satisfied with the decisions they had made. Perhaps even more disturbing is the observation by Hastie et al. that when a unanimous verdict was not required the jurors who were in the majority used coercive, forceful and bullying tactics to try to persuade those who disagreed with them.
The general opinion of social psychologists is that a majority verdict reduces the effectiveness of juries. There may, however, be some occasions on which it is useful. If, for example, a single bigoted juror insists on sticking to an opinion despite it not being based on any evidence (‘I don’t care what anyone says, you can tell by looking at that lad that he did it’) then at least they can be ignored. Nevertheless, the overall verdict is that expressed by Brehm & Kassim (1996) when they state that ‘it is clear that this procedure weakens jurors who are in the voting minority, breeds close-mindedness, inhibits discussion and leaves many jurors uncertain about decisions’ (page 487).
Generally, very little is known about the psychological processes that happen in real juries as opposed to the mock juries of the psychological experiment (McCabe and Purves, 1974). Myers (1979) chose a somewhat circuitous approach. She considered the cases of about a thousand defendants on felony charges in Indiana. Two thirds of the cases involved trials by jury. Myers obtained qualitative data from a variety of sources:
.file folders of the assistant prosecutor;
.police arrest records;
.telephone interviews to obtain any necessary supplementary data.
She was on the prosecutor's staff at the time so was also able to conduct informal discussions with prosecutors and court personnel.
She found that the following factors influenced the jury's decision of guilt:
.a weapon was recovered;
.a large number of witnesses were specified;
.the defendant or an accomplice made a statement either concerning involvement in crime or lack of involvement in crime;
.the defendant had large numbers of previous convictions (admissible evidence there);
.the defendant was not employed;
.the victim was young;
.it was a less serious rather than more serious crime.
Juries were not dependent on the following in their decisions:
.eyewitness identification of the defendant;
.recovery of stolen property;
.victim's prior criminality and relationship with the defendant; . past conduct of victim perhaps warranting the injury.
As a general rule, when people explain the behaviour of others, they tend to overestimate the role of personal factors and underestimate the impact of the situation. This is such a pervasive bias that it is known as the fundamental attribution error (Ross, 1977). Many studies have demonstrated that even when the causes of people’s behaviour are very obviously situational, people still attribute it to their personality. For example, Jones & Harris (1967) asked students to read out an essay either favouring or opposing Fidel Castro’s regime in Cuba. Although it was made quite clear to the audience that the essays had been allocated to the students by their teacher and were not their own work, the narrators were still considered to hold the opinions expressed in the essay they had read.
Stanley Milgram placed people in a situation in which they were required, in the name of research, to give a ‘learner’ an increasingly severe electric shock (never actually administered) every time he made an error on a simple memory task (Milgram, 1963). The results of this study were profoundly disturbing: a staggering 65 per cent of people were prepared to administer shocks so severe that they could have been lethal, even though the victim could be heard to scream, yell, complain about a serious heart condition and then fall silent. When people hear of such behaviour, their usual conclusion is that any individual capable of performing such acts is a sadist. Psychiatrists, psychologists and members of the general public interviewed by Milgram prior to the study estimated that fewer than one in a thousand people would give the highest level of shock. But the behaviour of the participants during the study showed they certainly were not being deliberately cruel. The great majority of participants were absolutely distraught, continually questioned the procedure and argued to be released from the study. They were not sadistic — they hated every minute of the study but they felt locked into a situation from which they could not escape. It was the situation, not the disposition of these individuals that was responsible for their potentially cruel behaviour. The assumption that cruel actions are motivated by an ‘evil personality’ is an example of the fundamental attribution error.
Lerner (1980) suggested that people have a need to believe in a world that is just and fair, a world in which people get what they deserve. In this way we can protect ourselves from the unpalatable truth that we could fall victim to the cruel twists of fate.
However, this belief is often strongly challenged by the fact that we regularly hear about events that involve the suffering of innocent victims and the unjust treatment of people. Rather than surrender our belief in a just world, we find ways of interpreting the causes of people’s behaviour in such a way that it restores our faith in a just world. There are several ways in which this can be done.
• We can blame victims for their own misfortune. For example, we can accuse a victim of burglary as being careless and not securing their property adequately.
• We can vilify the character of victims, so they are perceived as deserving of their fate. For example, we may believe that battered wives provoke their abusive husbands or even have a personality type that is attracted to such behaviour.
• We can help the victims or compensate them in some way.
The tendency to disparage victims can be viewed as yet another example of the fundamental attribution error — too much emphasis is put on the person and not enough on the situation. Although believing in a just world is common, there are marked individual differences in the strength of this belief. A strong believer in a just world is relatively unsympathetic to victims of misfortune and is also unlikely to take account of mitigating circumstances that may lead someone to commit a crime. Such individuals are more likely to be biased in their judgements whether they are witnesses or jurors.
One of the most disturbing aspects of the Just World Hypothesis is the attitude that is sometimes expressed towards rape victims. ‘She must have led him on’, ‘She was asking for it walking home that late at night’, ‘What does she expect dressed like that?’ are the kinds of comments often heard in media reports of rape. Bell et al. (1994), using a simulated jury situation, asked male and female college students to consider four different incidents of rape, two of which concerned rape by a stranger and two which involved date rape. Both males and females placed considerably more responsibility for what happened when the victim knew her assailant than when he was a stranger, and males ascribed more blame to the victim in both conditions than did females.
Pollard (1992) has reviewed many studies looking at judgements about victims and attackers in depicted rapes. He reports that, generally, females make more pro-victim judgements than males do, as do those people with non-traditional sex-role attitudes. The victim’s dress and past history also have a consistent effect in that the more scantily or provocatively dressed and the more sex partners the victim has had, the less sympathy they receive. Pollard points out that not all results found in experiments are directly relevant to a trial situation: for example, data on sentencing has no direct application since juries don’t recommend sentences. Indirectly though, juries are aware of the likely sentence that can be imposed and this can influence their decisions.
Piaget (1932) reported that children think that somebody has greater responsibility for their actions if the outcome is relatively bad, even if the intentions were good; hence a child helping his father to fill his inkwell who spilt the ink over the carpet was naughtier than another child who deliberately spilt ink on some blotting paper, when told he shouldn’t. In courtrooms hedonistic relevance comes into play when for example a man who falls asleep at the wheel of his car but comes to a halt safely on the verge is judged to be eligible for a small punishment. If, however, the same driver ended up on a railway track and subsequently killed a number of people on a train that ploughs into his car, he is judged to be deserving of a substantial custodial sentence. Logically, in both cases the man’s actions are the same and the punishment should be just as severe in both cases, but our psychology does not allow us to easily agree with this. This principle was tested by Walster (1966); participants were asked to judge how responsible a car owner was for the fact that his hand brake cable snapped. When the car rolled down a hill and badly injured someone the owner was deemed more responsible than when the car did not injure anyone. The experiment has been criticised because responsibility is confounded with being reasonably able to foresee what might happen.
The discounting principle (Kelley, 1971) predicts that we choose the most obvious explanation for somebody’s actions. When we see a famous personality advertise a product we would choose to believe that he or she advocates a product because of a large fee, rather than he or she really recommend the product. So in a court of law we might assume that if a confession is obtained under duress then the jury would disregard the confession. Unfortunately this is not always the case. This could be because the jury believe in a fair world where all police behave correctly at all times, even when interrogating a suspect. Another explanation could be the fundamental attribution error. The reasoning here would be that the defendant confessed therefore he must be guilty, completely disregarding situational factors.
The Story Model
The Story Model proposes that jurors go through three stages in their decision-making.
1 They decide on a story that, in their opinion, provides the best account for the events that occurred. This story will depend on the evidence, on world knowledge and on the personal experiences of the Juror.
2 In the second stage the jurors are required to learn the verdict categories that could apply in that particular case.
3 Finally they have to find the best fit between the verdict category and the story and come to a conclusion about the guilt or innocence of the defendant.
Pennington & Hastie (1986) have tested their model by conducting many studies in which mock jurors are requested to carry out their deliberations out loud. As a result, they have presented substantial empirical evidence in support of the model. Their current research focuses on aspects of the model not yet investigated, such as when exactly the story is created and whether jurors construct several stories and then choose between them, or if they construct just one.
Penrod and Heuer (1997) review the evidence coming from field experiments
involving a fairly large number of different judges, trials, lawyer and jurors.
In one national study, judges gave the jurors permission to take notes as soon
as was practicable in the trial. The majority of jurors took up the option
although as many as a third chose not to in some instances. On average, taking
into account the civil and criminal trials, just over half a page of notes were
taken each hour of the trial. While studies do not show a spectacular effect of
note-taking, some of Penrod and Heuer's conclusions are of particular
. Note-taking does not interfere with a juror's ability to keep up with the proceedings.
. Note-taking seems to be neutral in regard to the prosecution and
defence cases in terms of its effects.
. Note-takers do not concentrate more on the evidence in their notes than on the other evidence available.
. Note-taking jurors are not more satisfied with the trial, the judge or the verdict than those who do not take notes.
. The notes are at best a small assistance to remembering aspects of the evidence.
. The notes taken tend to be accurate records as far as they go.
Similarly, research into the effectiveness of allowing jurors to ask questions using similar methods led Penrod and Heuer to the following conclusions:
. Jurors do understand the facts and issues better if they ask questions.
. Allowing questions seems to make no difference to the jurors', judge's and lawyers' satisfaction with the trial and verdict.
. If a juror asks an inappropriate question, lawyers will object and the jury does not draw inappropriate conclusions from this. However, generally speaking, the jurors ask perfectly appropriate questions.
In other words, research of this sort tends to demonstrate little or no negative consequences of the potential innovations of note-taking and juror questions, and modest, at best, improvements.
Jurors and conformity
Experiments with mock juries find that if two-thirds of the group agree then that decision will eventually win. Any less, and the group becomes deadlocked. Generally in psychology it has been of interest how the majority win through. The research on conformity pressures show that the minority will change for a number of reasons.
Conformity is when an individual gives up his or her personal views under group pressure. Within a jury situation, two types of conformity may occur: normative and informational. In the latter case, the individual conforms to the group norms because he/she does not know what to do and looks to the group for guidance. It may be that the individual does not understand the legal technicalities of the case. With normative conformity, the individual outwardly conforms to avoid rejection by the group or to gain rewards from them, but inwardly disagrees. This is a more superficial form of conformity.
Solomon Asch undertook the classic study of normative conformity in the 1950s. He asked individuals to say in front of a group of people which two out of four drawn lines were the same length. The answer was obvious. But a number of the group members deliberately gave the wrong answer. Would the ‘real participants’ (the others in the group were working for the experimenter) follow their own judgements and give the obvious correct answer or conform to the group with the wrong answer? The participants conformed in about one-third of the trials.
Asch (1951) developed this experiment to pin down the factors involved in conformity. He found the best group size to gain conformity was 7:1. Also, greater conformity was produced by the more difficult the task, or the higher the status of group members.
Another decision-making process observed in groups is known as group polarization. Stoner (1961) was the first to experimentally study group polarization, but it was called ‘risky shift’ at that time. Participants were given twelve dilemmas to think about by themselves, and then asked as a group to discuss the dilemmas. Stoner found that the group decisions were riskier than the individual decisions. Moscovici and Zavalloni (1969) coined the phrase ‘group polarization’ to show that group decisions would always be more extreme (either riskier or more cautious) than the individual decisions.
Group polarization can also be noted with federal judges either sitting as a ‘three’ or ‘one’ during a trial. The three judges gave 65% of their judgements as ‘libertarian’ (less harsh) compared to 30% by the individual judges.
The opinions of the majority are hugely influential in the final decisions made within a group and juries are no exception to this rule. Hastie et at. (1983) found that if the first decision favoured acquittal, then in 86 per cent of cases a not guilty verdict was returned. If at the outset the majority favoured a guilty verdict, then in 90 per cent of cases, that was the final outcome.
Social psychologists explain group polarisation in terms of the need to make sense of the social world. Many people hold the view that if most people think something is correct then it probably is. Interviews with jurors and studies of mock juries show that the majority has a powerful influence over the minority. The extent of this influence is surprisingly strong, as shown by Asch (1955). If people are prepared to deny the clear evidence of their own eyes on a simple task, as 75 per cent of the participants in this study did at least once, consider how much more powerful the influence of the majority may be when considering whether they are right or wrong in a far more ambiguous situation. The experience of being in a small minority, especially a minority of one, can create considerable anxiety, which is reduced by conforming to the majority decision.
Where group polarization doesn’t apply
Generally group polarization does not apply in situations where individuals are arguing for something they believe in. However, the effect is also influenced by identification with the defendant, the victim or the authorities. This process is through the social identity theory, where individuals ‘self-stereotype’ themselves based on the groups they feel they belong to (Wetherell, 1987). For example, a former police officer on the jury may identify with the police officers giving evidence, and be influenced in his or her verdict that way.
The decision-making process of a group can also show evidence of ‘groupthink’. Janis (1972) sees ‘groupthink’ as the pressure to reach a consensus that leads to an extreme or bizarre decision. This situation could particularly happen to the jury in a long, highly publicized case.
There are certain conditions that facilitate ‘group-think’: the group is highly cohesive (that is, it has strong group identity), it is insulated from outside information, it is under pressure to make a decision, all the options are not assessed, the group feels the responsibility in making a very important decision, and there may be a dominant, directive leader.
Janis has studied ‘groupthink’ with political decisions, a famous example being President Kennedy’s decision in the 1960s with the Bay of Pigs. This was the area suggested for the invasion of Cuba by American troops, but it was the worst place to do such a thing (because it was marshland).
A completely different approach to understanding group decisions comes from psychodynamics. These ideas come from the work of Sigmund Freud and place great emphasis on the unconscious processes in the group. The jury has a task to perform, but it is ‘as if another group, comprising the same individuals was operating simultaneously, at the unconscious level’ (Morgan and Thomas, 1996).
Thus, the jury decision may be more about these processes than the actual evidence. For example, a jury may convict a child murderer out of the unconscious fear for their children rather than on the evidence. However, this view of the group has little empirical support.
Another area of interest is whether all members of the juries contribute equally to the decision. In a group where there is no individual recognition of effort, research has noted a phenomenon called ‘social loafing’. Williams et al (1981) define this as a ‘reduction in individual effort on a collective task (in which one’s outputs are pooled with those of other group members) compared to when working either alone or coactively’.
In a classic experiment, Latane et al (1979) asked participants in a sound studio to make as much noise as possible, and the level was measured. First, they did this alone. Then the participants were told that they were linked to other studios and the joint noise levels would be measured. In fact, this was not true. But when the participants thought they were part of a group of six, their individual level of noise dropped by around a quarter, compared to when they believed they were alone. In other words, they were putting in less individual effort to the group cause. Applying this research to the jury situation, the larger the jury, the less individual effort each jury member will put in.
Because it is impossible to study real juries, it is assumed that all these psychological processes of groups are at work. A limited number of studies have tried to assess the accuracy of the jury decision by comparing it with the judge’s view. In one American study of 3,500 cases, there was around 80% agreement between the judge and jury on the verdict (Kalven and Zeisel, 1966).
The strength of majority influence
Smith & Mackie (1995) suggest that there are several factors which operate in favour of the majority position.
• Firstly, when the majority is offering a certain opinion the arguments are more numerous and almost certainly more varied. The minority of the jurors who disagree with the majority are likely to be presented with a set of persuasive arguments that they had not previously considered and this is likely to move them towards the majority view (Hinsz & Davis, 1984).
• Secondly, when information and opinions are shared by several members of a group, they tend to be discussed for longer than views held by one person. Stassér & Stewart (1992) devised a situation in which the majority of members of a group shared certain information but other information was only given to a single individual. Even though the group were specifically instructed to discuss all the information, they focused almost entirely on that which was shared, to the almost total exclusion of the non-shared information. This is another reason for majority influence: their opinions receive considerably more discussion.
• Thirdly, majority arguments are usually more compelling than minority ones. When several people make the same arguments, it has more impact than one offered by a single individual. People generally believe that if several individuals have come to the same conclusion, then it is more likely to be the correct one. Therefore the arguments that are advanced by the majority of jurors tend to be very persuasive and move the group further towards the extreme of their original opinion.
• Fourthly, majority views tend to be expressed more convincingly than minority views. With the benefit of knowing that most people are on their side, members of a majority use a forthright and argumentative style that is compelling (Kerr et al., 1987). Perhaps it is for this reason that members of a majority are viewed as being more confident, logical and intelligent than those of a minority (McLachlan, 1986), resulting in the likelihood that they will convert them. Once again, the end result is group polarisation.
In summary, the majority view is more likely than the minority one to be accepted because this view is expressed more often, is discussed more and in greater depth, seems more compelling and is expressed more persuasively than the opinions of the minority. This quantitative and qualitative advantage leads to group polarisation.
Usually majority influence persuades those holding a minority view to accept the majority view. However, upon rare occasions a minority can persuade a majority to accept their viewpoint. Moscovici (1976, 1980, 1985) maintains that the success of minorities is dependent on the behavioural style of the individuals involved. If the minority is consistent and flexible and their arguments are relevant, then they may eventually win over the opinions of the majority. The first of these factors, the consistency with which the group defends and advocates its position, is the most crucial. This consistency must be maintained between the minority group and over time. If the minority members agree amongst themselves and continue to do so, they may persuade the majority to question its own assumptions and seriously consider those of the minority. To be successful, those people in the minority must not appear to be rigid and dogmatic but flexible in their approach and willing to discuss the reasons why they disagree with the majority.
Just as we looked at why majorities are influential, it is appropriate to consider why some minorities are successful in converting others to their point of view. Nemeth (1977) suggests that when majorities are faced with a consistent minority sticking to their guns, they are puzzled and try to work out why they are so convinced they are right and so determined to express publicly these unpopular views. The majority is therefore prepared to scrutinise these minority views and, on occasions, be convinced by them. Even when minorities fail to sway people initially, they may start a questioning process, which disconcerts the majority and may eventually lead to change.
One of the reasons why a minority is rarely influential may be the fact that not all jury decisions have to be unanimous.
Leadership of the jury
Within the jury one psychological process concerns leadership of the group. The jury elects a foreperson. Research has found that the person who is elected as foreperson tends to be of a higher social class, or has previous experience of jury work, or sits at the head of the table or speaks first (Strodtbeck and Lipinski, 1985). In practice, this often means a man. In a San Diego study, Kerr et al (1982) found that 90% of forepersons were male, though the majority of jurors were female. However, the foreperson may not be the actual leader of the group.
So who emerges as the leader? General research on leaders has tried to establish if there is a certain type of person who always emerges as the leader. This is known as the ‘great person’ theory (or originally as the ‘great man’ theory, because all the research was on male leaders). For example, Mann (1959) reviewed hundreds of studies, but no clear patterns emerged.
Leaderless group discussion
The alternative view is to concentrate on the situational approach and see who emerges as the leader in a particular situation. Experimentally this is tested by the Leaderless Group Discussion (LGD) technique, which involves a group of strangers discussing a particular topic. They are observed to see who emerges as the leader, and group members are asked afterwards who they think became the leader. For example, in highly ambiguous situations (where there is no clear-cut answer), the most talkative individuals emerge as leaders. This is called the ‘blabbermouth rule’ in some books. However, other research suggests that it is often chance who becomes the leader.
In a LCD experiment, the researchers led some participants to believe that they were making important comments, and these participants emerged as the leaders (whether they were most talkative overall or not).
The effectiveness of juries
In an attempt to increase the objectivity and effectiveness of juries, Wrightsman et al. (1994) have put forward a set of recommendations based on empirical evidence and case studies of actual trials:
1 Jury exemptions should be restricted so that juries are more representative of the whole population.
2 The number of peremptory challenges that can be made should be limited, and during voir dire the questions should be framed so that they increase the likelihood of an honest response from jurors and reduce the attorneys’ ingratiation techniques. Judges, rather than attorneys, should do most of the questioning.
3 To prevent the influence of inadmissible evidence, the trial should be videotaped and edited to remove objectionable material (the researchers acknowledge that this may be impracticable).
4 During deliberation the jury should have access to a transcript or videotape of the trial to which they can refer when questions arise.
5 Jurors should be permitted to take notes during the trial or refer questions to the witnesses.
6 Instructions should be clearly worded, given in writing as well as oral form and delivered at the beginning and at the conclusion of the trial. In complex trials in which several verdicts must be given, the questions that the jurors must answer should be put into a logical sequence.
Kevin Brewer, 2000, Psychology and Crime, Heinemann, 0-435-80653-X.
Diana Dwyer, 2001, Angles on Criminal Psychology, Nelson Thornes, 0-7487-5977-8
Philip Feldman, 1993, The Psychology of Crime, Cambridge, 0-521-33732-1.
Julie Harrower, 2001, Psychology in Practice – Crime, Hodder & Stoughton, 0-340-84497-3
Clive R. Hollin, 1989, Psychology and Crime, Routledge, 0-415-01807-2
Dennis Howitt (2006) Forensic and Criminal Psychology, Pearson Prentice Hall, ISBN 0-13-129758-9
PSYCHOLOGY IN COURT: AN OVERVIEW by BRIAN HUGHES*