John C Reinard. The Persuasion Handbook: Developments in Theory and Practice. Editor: James Price Dillard & Michael Pfau. Sage Publication. 2002.
The study of Persuasion in the courts has been a major part of communication studies. Aristotle (1941) identified the forensic exchange as a major division of oratory (1358b6) and explained the close link:
Since rhetoric exists to affect the giving of decisions—the hearers decide between one political speaker and another, and a legal verdict is a decision—the orator must not only try to make the argument of his speech demonstrative and worthy of belief; he must also make his own character look right and put his hearers, who are to decide, into the right frame of mind. (1377b21-1377b24)
With the founding of European universities during the Middle Ages, the study of rhetoric (expanded to include law) was at the center of the core curriculum. With a few breaks in the link along the way, this close tie has endured as a standard part of both communication studies and the practice of law. Today, the study of communication and the law has grown from a cottage industry to a major part of the communication field. Each professional organization in the field has an interest group, division, or commission dedicated to legal communication. The American Society of Trial Consultants, which claims many members from psychology and communication studies, has more than 340 members. There is little doubt that legal Persuasion is a major part of the field. This chapter reviews elements of Persuasion in the legal setting and considers theories, basic literature on sources, courtroom messages, jurors as audiences, and the status of the research area. The emphasis is on social science work in the subject, although much intriguing work has been completed using other methods of inquiry.
Theoretic Orientations in Legal Persuasion
Work in legal Persuasion frequently has been driven by practical needs of lawyers. Although this link has kept such studies grounded, it often has placed theory development in a secondary role. Even so, two different strains have dominated theoretic work: the application of existing theories and orientations and the development of conceptualizations specific to the law.
Application of Existing Theory
Although recognizing communication constraints created by legal procedures, Mills (1976) argued for approaching much legal communication as an extension of rhetorical theory, small group communication, and argumentation studies. There can be no doubt that many rhetorical principles apply to courtroom settings. Especially in areas where research specific to legal settings had not been completed, scholars often relied on research in general communication. For instance, Rieke and Stutman’s (1990) excellent textbook covered the use of humor (pp. 69-70) by reviewing non-courtroom research exclusively. Fewer than 20% of the sources cited in their chapter on “Credibility” were specific studies of courtroom behavior. The National Jury Project’s (1986) sourcebook similarly could not rely exclusively on court-specific research; for example, all citations about jury nonverbal communication were from general communication sources (pp. 11-25 to 11-33). Although directly applying communication to the legal setting has intuitive appeal, it also is the case that special constraints in the legal setting limit the reasonable bounds of such applications.
Because legal advocacy is Persuasive argument, Rieke and Sillars (1997) argued that legal case reasoning is frequently analogic and syllogistic: “[major premise:] This is the relevant law…. [minor premise:] The facts of the present case are embraced within those contemplated in the law…. [conclusion:] This case should be decided according to the law stated in the major premise” (pp. 95-96). Because the legal setting also is a forum for Persuasion, some scholars attempted to apply theories of attitude change. Varying degrees of success were found by applying the Linear Discrepancy Model (Boster, Hunter, & Hale, 1991) and the Elaboration Likelihood Model (Floyd, 1998/1999; Wursten, 1986/1987). This approach to legal Persuasion has had the advantages of exploring boundaries of existing theories and identifying predictive relationships.
Other attempts were made to describe the flow of courtroom Persuasion. One promising direction viewed legal Persuasion as a game in which the players used strategies to manage payoffs (Kalai, 1993). Yet other approaches to predicting jury decision making that have been tested and generally found wanting have included multinomial decision making involving decision criteria used by juries, Bayesian probability models of defendant propensities to commit charged crimes, models of jury size and accuracy, juror and deliberation error catching, and simple computer simulation assuming differing degrees of juror resistance to Persuasion (Penrod & Hastie, 1979).
Conceptualizations Specific to the Legal Persuasion
Thinking in legal influence has been dominated by psychologists who have focused attention on two major views of the trial: the story model and varieties of the information acquisition model. In essence, the story model describes the courtroom as an arena in which competing stories vie for the favor of jurors (Hastie, Penrod, & Pennington, 1983, pp. 22-23; Pennington & Hastie, 1986, 1990, 1992). In particular, the story model treats juror decision making as a three-stage cognitive process. During the first stage, jurors attempt to comprehend and organize evidence to create rational versions of the events giving rise to the trial. Jurors frame their explanations of reasonable cause and effect and intentional behavior patterns based on their “general knowledge” of the ways people act. During the second phase, the “verdict-category-establishment” stage, jurors attempt to identify a different verdict alternative and to choose a decision rule that would be required to arrive at a verdict. During the third phase, jurors select the verdict category that seems to be the best match with the plausible stories. Pennington and Hastie (1993) explained that story acceptance depended on the degree to which it covered all of the information presented to jurors, maintained a sense of coherence, appeared to be unique among competing explanations, and established its “goodness-of-fit” with the evidence. The story model seems compatible with other theories of narrative communication. For instance, Bennett and his associates (Bennett, 1978, 1979; Bennett & Feldman, 1981) considered storytelling not just as a way to share information but also as a way for jurors to make sense out of the trial evidence.
Research on the story model has been mixed. Some work observed that jurors often describe trial evidence in story terms (Pennington & Hastie, 1986). Experimental studies that drafted cases as stories found that these adaptations promoted desired verdicts (Pennington & Hastie, 1992; for similar support, see Kuhn, Weinstock, & Flaton, 1994). Hastie and Pennington (2000) observed that the believability of the crime story predicted verdicts even more than did the quality of the evidence. Yet other studies did not support the story model when exploring the predicted role of juror reliance on social problem solving (Becker, 1998/1999). Others found that opening and closing statements were not made increasingly Persuasive by organizing them as stories (McCullough, 1991), even after controlling for the type of organization used in the opposition’s opening and closing statements (Spiecker, 1998/1999). Conceptually, the story model has had its limitations. First, given that some trials deal with more vivid stories than do others, it may be that the model is least useful when non-narrative materials are presented. Second, the model might not be testable given that it is most difficult to identify any predictions that are recognized as clearly contrary to the theory. Nearly any form of trial outcome could be enlisted to support the story model’s view of what makes Persuasive and unPersuasive narratives. These limitations have not been an impediment to trial consultants. One consulting team used the story model to identify coherent “guilty” and “not guilty” stories that could be understandable and potentially Persuasive to jurors (Olsen-Fulero, Fulero, & Wulff, 1989).
A strong alternative to the story model has been the information acquisition model. This approach views jurors as recipients of information who make decisions based on the amount and nature of information with which they are presented. Calling this approach the “meter model,” Lopes (1993) observed that jurors often keep track of subjective values they place on the merit of evidence for each side in a trial. Although finding much to admire in this view, Lopes suggested acceptance of both the meter and story models. Using an averaging model of information integration, Ostrom, Werner, and Saks (1978) noted that mock jurors initially presumed a defendant to be not guilty, but by the end of the trial, the amount and degree of incrimination in the evidence was averaged with the presumption of innocence to reach a decision. Although many researchers were implicitly guided by this approach, they did not always explain their work as such an application. Furthermore, this category of theorizing has not always promoted consistent predictions of jury behavior. Some might consider it an invitation to incoherent collections of research rather than a body of theoretically valuable materials.
Sources of Persuasive Communication in Legal Settings
Many excellent books (Gibson, 1991; Matlon, 1988; Rieke & Stutman, 1990), articles, chapters, and paPers (Boyll, 1991; Devine, Clayton, Dunford, Seying, & Pryce, 2001; Elwork, Sales, & Suggs, 1981; Erlanger, 1970; Linz & Penrod, 1984; Nietzel, McCarthy, & Kern, 1999; Wright & Hosman, 1980) have summarized varying portions of the literature on communication and the law. Unlike those assessing general issues in legal communication, this review examines social science research dealing with Persuasion exclusively. Thus, there are major areas of study that this review excludes. This review of sources considers influences of lawyers, plaintiffs, defendants, victims, and judges.
Lawyers as Sources
Although the aphorism that jurors try the attorneys rather than the clients overstates the matter, there is little doubt that attorney conduct and characteristics affect jury decisions. The credible attorney appears to possess the factors of character, competence, extroversion (Hirsch, Reinard, & Reynolds, 1976), supplemented with a vital factor identified as “likeability” (Rickun, 1977). How attorneys acquire and use such credibility, however, remains an area for serious attention.
Research is very mixed on the impact of the type of lawyer representing defendants. For instance, some inquiry has suggested that privately retained attorneys have a better record than do public defenders. Lizotte (1978) observed that court-appointed attorneys are defeated more often than private attorneys. One study reported that private attorneys lost 48% of cases as compared to 59% for public defenders (Wilbanks, 1985, as described in Wilbanks, 1987, p. 92). Reiman (1995) reported statistical claims that private attorneys were successful as often as 56% of the time, unlike public defenders who secured acquittals or dismissals of charges only 11.4% of the time (p. 117). Another study found that private attorneys have a suPerior record of representing minority group defendants (Wheeler & Wheeler, 1980, esp. pp. 319-323). Some work has indicated that private attorneys secure longer sentences for their clients (Wilbanks, 1985), but a review of California cases found no differences (Truitt, 1997). Other research has suggested that any advantages private attorneys have in reducing rates of conviction may have disappeared during recent years. A review of criminal cases in U.S. district courts and the 75 largest counties in the United States revealed that conviction rates for defendants with private attorneys were nearly identical to conviction rates for those secured by public defenders or other court-assigned attorneys (Harlow, 2000). Overall rates of guilty verdicts (90% in federal courts in 1998 and 75% in county courts in 1996) remained high for defendants, even though convicted defendants who had private attorneys were more likely to receive prison sentences than were those with court-appointed lawyers. Although research has not directly linked attorney effort to actual verdicts, it seems that attorneys were motivated to enhance their Performance when they thought that their cases were likely to go against their clients (Lind, Thibaut, & Walker, 1973).
Attorneys’ Personal characteristics may predict jury decisions. Some research suggests that male attorneys have greater trial success than do female lawyers (Hahn & Clayton, 1996). Yet male attorneys’ physical attractiveness did not influence verdicts (Taylor, 1993). Furthermore, male (but not female) jurors rated the defendant as guilty least often when the prosecutor was female (Pfeifer, 1988). In rape trials, if the defendant is represented by a female attorney, the rate of acquittal increased from 49% to 71% in one study (Villemur & Hyde, 1983) and from 47.5% to 68.8% in another (Yanchar, 1982/1983). Apparently, when a woman was willing to defend a man accused of rape, jurors inferred that the case against the defendant was not particularly influential. Attorney race may affect juror decision making in combination with jury authoritarianism. Among highly authoritarian jurors, Anglos’ verdicts were most influenced by Anglo defense attorneys, and African American jurors were most Persuaded by African American attorneys (Boliver, 1999).
Plaintiffs as Sources
A sense of how deserving and desirable a plaintiff is seems to affect jury assessments. In fact, although social desirability of both plaintiffs and defendants predicted verdicts, only plaintiff desirability predicted size of monetary awards granted (Egbert, Moore, Wuensch, & Castellow, 1992). Jurors in civil cases tended to decide in favor of attractive litigants (Snyder, 1971; Stephen & Tully, 1977). But this preference for attractive plaintiffs seemed to depend on whether people processed information in a rational mode (dedicating energy to the information they heard) or an exPeriential mode (responding to information on an emotional level). Attractive plaintiffs were favored only by individuals in the rational mode (Lieberman, 1998).
Defendants as Sources
Defendant characteristics that have been the object of research attention include the defendant’s race, attractiveness, and various Personal characteristics. Each of these elements is reviewed in this section.
The defendant’s race may affect case outcome, but not under every circumstance. Concerns that ethnic minorities are overrepresented in the prison system (e.g., Clayton, 1983; Radelet, 1981) may complicate the matter. Although studies through the early 1960s showed that African American defendants were found guilty most often and were given the longest sentences (Bullock, 1961; Garfinkle, 1949; Johnson, 1941), research since that time has not always found such a pattern. Indeed, Anglos received longer sentences than did Latinos in Texas noncapital felony trials (Daudistel, Hosch, Holmes, & Graves, 1999), and Anglos on trial in Los Angeles were convicted more often than were African Americans (Petersilia, 1983). In a controversial summary statement, Wilbanks (1987) asserted, “The overwhelming consensus of recent studies (obviously, conviction rates may have been higher for blacks in earlier Periods) has been that blacks are less likely to be convicted than whites” (p. 98). A battle of literature reviews has raged on the subject. Reviewing 20 articles, Hagan (1974) found that defendant race effects explained between 0.4% and 8.0% of the variance on sentences. But studies controlling for the defendant’s prior criminal record and the type of crime did not produce significant effects. Kleck’s (1981) review of 57 empirical studies revealed that in the South the death penalty was imposed in ways that reflected racial discrimination. Otherwise, he stated that discrimination was not widespread in the legal system, a position for which he was attacked, sometimes Personally (see Austin, 1984; Dehais, 1983). A review completed by the Panel on Sentencing Research of the National Academy of Sciences examined 60 studies (Blumstein, Cohen, Martin, & Tonry, 1983) and opined that racial discrimination effects in sentencing were not general but “may play a more important role in some regions, jurisdictions, crime types, or the decisions of individual participants” (Vol. 1, p. 92). On the other hand, some (Hagan & Bumiller, 1983, esp. pp. 31-32) have noted the existence of studies completed after the mid-1960s that controlled for major intervening variables and still reported racial bias. For instance, Hall and Simkus (1975) found that 76% of Native Americans convicted of crimes were sentenced to jail time as compared to 59% of Anglos. One study observed a general tendency toward bias by Anglo mock jurors in a rape case resulting in light sentences for Anglo defendants, especially when jurors were given no pre-deliberation instructions or were presented with strong jury nullification instructions (Hill & Pfeifer, 1992). Other researchers completing controlled work have found evidence of conviction proneness toward African American defendants (Lands, 1986/1987; Townsend, 1996; but see Rector, Bagby, & Nicholson, 1993) and of increasingly severe decisions during the death penalty phase of a trial (Lynch, 1997).
But given that ethnic minorities often are poor and tend to be represented by public defenders, it is difficult to accept the notion that racial prejudice is largely inoperative. The controlled research suggests that prejudice is most likely to be found against ethnic minority defendants when cases were tried in the South (particularly in death penalty cases) or in rural areas (Austin, 1981); when jurors in cases involving African American defendants were identified as highly prejudiced (Dovidio, Smith, Donnella, & Gaertner, 1997); when jurors in forcible rape cases were of a different race from the defendant (Feild, 1979; Ugwuegbu, 1976) and when student jurors viewed themselves as dissimilar to the defendant in basic values and background (Reynolds, 1977); when female jurors in an acquaintance rape trial learned that the victim was not a member of their own race (Hymes, Leinart, Rowe, & Rogers, 1993); and when student jurors tried the defendant of a crime that fit a racial stereotype, including an African American accused of burglary or a violent crime (Bodenhausen & Wyer, 1985; Gordon, 1990; Gordon, Bindrim, McNicholas, & Walden, 1988; Rickman, 1988/1989). Elements that seem to inhibit this effect against ethnic minority defendants included when the jury was composed of middle or upPer class student jurors (Nemeth & Sosis, 1973) and when the defendant in an insanity defense had low socioeconomic status (Towers, McGinley, & Pasewark, 1992). In an effort to explore reasons for some of these differing results, Sommers and Ellsworth (2000) suggested that Anglo jurors are motivated to appear unprejudiced. Thus, in cases where race seemed salient, Anglo student jurors would be expected to become vigilant and avoid prejudicial decision making against African American defendants. Contrariwise, when racial issues were not salient to cases, increased guilty verdicts against African American defendants might be enhanced. By manipulating trial summaries presented to student jurors to illustrate different levels of race salience, the researchers found support for their expectations. On the other hand, African American student jurors showed leniency toward African American defendants regardless of the race salience levels. Membership in a racial minority seemed to benefit defendants when an African American defendant pleaded not guilty by reason of insanity (Poulson, 1990); when prosecution evidence in child molestation cases was strong, and mock jurors who believed they would be in the racial minority on the jury were of the same race as the defendant (Kerr, Hymes, Anderson, & Weathers, 1995); and when student jurors tried an African American defendant on a charge that defied a racial stereotype, such as embezzlement or white-collar crime, a crime for which Anglos were most often convicted (Bodenhausen & Wyer, 1985; Gordon, 1990; Gordon et al., 1988; Rickman, 1988/1989).
In criminal cases, as in civil cases, jurors tend to favor the attractive source (Calhoun, Selby, Cann, & Keller, 1978; Darby & Jeffers, 1988; Kulka & Kessler, 1978; Landy & Aronson, 1969; Leventhal & Krate, 1977; McFatter, 1978; Reynolds & Sanders, 1975; Seligman, Brickman, & Koulack, 1977; Solomon & Schopler, 1978; Tieger, 1981; Wyatt, 1982/1983). Looking at defendant character rather than attraction directly, Barnett and Feild (1978) found that the character of the defendant had a significant effect in a case involving rape but played only a minor role in juror sentencing in a burglary case. Yet some interactions of attractiveness with other variables seem to exist. Although attractiveness predicted sentences and even amounts judges set for bail (Downs & Lyons, 1991), it did not always predict verdicts (Erian, Lin, Patel, Neal, & Geiselman, 1998). This information was intriguing for the courtroom Persuader because, in one study (Sigall & Landy, 1972), a participant was made to seem attractive simply by describing him as “loving and warm” instead of “cold and unapproachable.” Similar labeling also was found to be effective for participants in a traffic collision case (Kaplan & Kemmerick, 1974). Yet the type of crime with which a defendant was charged made a difference. When the charge was swindling, where attractiveness might facilitate committing the crime, an attractive defendant was more likely to be convicted than was a nonattractive source (Sigall & Ostrove, 1975). Similar results were found when comparing attractive and unattractive men and women charged with either burglary or swindling (Wuensch, Chia, Castellow, Chuang, & Cheng, 1993). Effects seemed to be most pronounced when jurors were high in authoritarianism (Mitchell & Byrne, 1973) or when student jurors had low socioeconomic status (Nemeth & Sosis, 1973). The impact of defendant attractiveness was reduced by actual jury deliberations (Izzett & Leginski, 1974). Even though deliberations shifted decisions toward acquittal for attractive defendants, they did not affect decisions for unattractive defendants (MacCoun, 1990).
Victim attractiveness also affected decisions. With strong prosecution evidence, unattractive defendants were given the harshest sentences for harming attractive victims (Erian et al., 1998; see also Feild, 1979). Interestingly enough, the effects of defendant attractiveness did not wizen when the trial was long instead of short (Kramer & Kerr, 1989). Related to attractiveness is defendant appearance. One study found that jurors were most receptive to defendants who were neither dressed in prison clothes nor attended by armed guards (Dane & Wrightsman, 1982; Fontaine & Kiger, 1978).
Personal Characteristics of Defendants
Matters such as defendant socioeconomic status and Perceived similarity and sympathy with jurors have been examined in the research. The defendant’s socioeconomic status can influence Perceptions of the case. In one inquiry, research controlling the degree of premeditation of murders and race of defendants, low socioeconomic status defendants received the longest sentences (Osborne & Rappaport, 1985). When physicians from high-status specialties were charged with murder, convictions were greatest. Yet when they were charged with fraud, the highest status specialties received the lowest rate of convictions (Rosoff, 1987).
Defendant status does not always affect verdicts (Adler, 1973; Gleason & Harris, 1976; Gordon & Jacobs, 1969). Yet once high-status figures were found guilty in murder cases, both community and student samples recommended the harshest punishments for these same defendants (Bray, Struckman-Johnson, Osborne, McFarlane, & Scott, 1978). Yet in other types of trials, low-status defendants received the longest sentences (Landy & Aronson, 1969; Reed, 1965). Related to status is juror knowledge of defendants’ “deep pockets” in civil cases. In one study, jurors who learned of the defendant’s insurance increased the plaintiff awards from $33,000 to $36,000, and when the jurors were told to disregard it, the awards jumped to $46,000 (Broeder, 1959). Similar support was found in a dog bite injury suit (Reinard, 1993). This effect was most pronounced when the defendant was a corporation rather than an individual (MacCoun, 1996). In follow-up work, Anderson and MacCoun (1999) noted that jurors who were not Permitted to award punitive damages inflated “pain and suffering” compensation.
Sympathetic defendants and those similar to the jury tended to have advantages in different kinds of cases. In civil cases, jury sympathy for the plaintiff was a strong indicator of decisions (Darden, DeConinck, Babin, & Griffin, 1991). In product liability suits, this sympathy led jurors to favor plaintiffs out of a general desire to promote a healthy social fabric for society rather than simply out of a commitment to follow the law. In criminal cases, defendants who are viewed as dissimilar to jurors tended to be found guilty less often than were others (Dane & Wrightsman, 1982). This effect even outweighed racial biases that may have been operating against minority group defendants in highly charged rape cases (Reynolds, 1977). Jurors also were most likely to convict defendants whose attitudes were Perceived as dissimilar to their own (Griffitt & Jackson, 1973; Shepherd & Sloan, 1979). Defendant sex affected jurors in a criminal case. College students who read case summaries gave men shorter sentences than they gave women for purse snatching (Angira, 1991). In another study contrasting shoplifting and assault cases, female defendants received fewer guilty verdicts from student jurors than did male defendants, although sentence recommendations were not different for the two types of defendants (Dravin, 1982/1983). Empathy with the defendant was tested in a patricide trial in which a child defendant claimed self-defense due to sexual abuse by the victim (Haegerich & Bottoms, 2000). When student jurors were asked to take the Perspective of the defendant and to consider how they would feel under the circumstances of the trial, jurors were most likely to decide not guilty. When a landlord defendant treated tenants coldly and as objects, the plaintiff awards increased (Holt et al., 1997). On the other hand, when criminal defendants already had suffered greatly, the severity of jury decisions was reduced (Shaffer, Plummer, & Hammock, 1986), especially when the defendants’ suffering was linked to the crime with which they were charged (Austin, Walster, & Utne, 1976).
Who the victim is makes a difference in jury decision making, particularly in regard to race and the victim’s conduct before commission of the crime. A literature review on the subject (Kleck, 1981) observed that when the victim was African American, punishment was less severe than when the victim was Anglo, particularly in death penalty cases. Furthermore, the victim’s age did not appear to produce significant effects on verdicts (Nunez, McCoy, Clark, & Shaw, 1999).
The victim’s conduct makes a difference in jury decisions, sometimes in surprising ways. If the victim were called to testify in a rape trial, mock jurors were most likely to return guilty verdicts when the victim avoided eye contact with the defendant (Weir & Wrightsman, 1990). In rape cases, defense attorneys sometimes attempt to place the victim’s past sexual conduct on trial despite great limits on this strategy. When student jurors read case summaries with medical reports and testimony including information about the victim’s past sexual history, the victim’s past reckless conduct did, in fact, reduce guilt ratings and sentence recommendations (Pugh, 1983). Convictions increased if jurors learned that rape victims attempted to resist the attacks (Krulewitz & Nash, 1979; Scroggs, 1976) and if the rapes resulted in pregnancies (Scroggs, 1976). Yet if the defendants were acquaintances rather than strangers, rates of acquittal increased (Calhoun, Selby, & Warring, 1976; Smith, Keating, Hester, & Mitchell, 1976).
Some research investigated the “just world” hypothesis, which holds that if one believes the world is a fair and just place, people pretty much get what they deserve. Thus, jurors may reason that respectable people must have particularly deserved the bad treatment they received given that they were victimized (Lerner, 1965). Some work has supported this notion, including a study completed in both the United States and the United Kingdom where the defendant was most likely to be acquitted if the victim took reasonable precautions to avoid the crime (Kerr, Bull, MacCoun, & Rathborn, 1985). Jones and Aronson (1973) found that respectable victims (women identified as married or as virgins) were blamed more for their victimage in rape cases than were divorced women. Yet the just world hypothesis has not always been supported. Despite their other findings, for instance, Jones and Aronson observed that the longest sentences were recommended when the victims were respectable. Others observed that conviction rates decreased when rape victims were disreputable, variously described as prostitutes (Feldman-Summers & Lindner, 1976) or topless/bottomless dancers, whereas convictions increased when the victims were described as social workers or Roman Catholic nuns (Smith et al., 1976). In compatible work, mock jurors’ empathy for rape victims increased rates of guilty verdicts (Weir & Wrightsman, 1990). Similarly, Kerr and Kurtz (1977) presented student jurors with case summaries and found that the defendants were given the longest sentences when the victims were made to suffer. Contrary to the just world hypothesis, the victim’s status was not denigrated when the respondents believed in a just world.
Judges as Sources
Judges often influence the jury through nonverbal cues. In one case that was the object of misconduct charges, the judge listening to a witness alternately shook his head, rolled his eyes heavenward, and turned around in his chair and stared at the wall for 45 minutes (State v. Jenkins, 1994). Even without such extreme cues, judges actively communicate their sentiments through cues such as smiles, nods, and frowns (Blanck, Rosenthal, & Cordell, 1985). The ways a judge looked at defendants and witnesses were eventually reflected in the jury verdicts among judges in municipal and magistrate courts (Dorch & Fontaine, 1978). There was a moderate correlation (.48) between the amount of a judge’s gaze at the defendant and the severity of the fine the defendant received.
Messages in the Courtroom
Persuasive messages include many elements at trial. This section examines the influence of charges and burden of proof, pretrial publicity, Persuasion during jury selection, opening statements, closing statements, defense strategy, case evidence, direct examination, cross-examination, and judge instructions.
Charges, Requested Civil Remedies, and Burden of Proof
The charge or complaint itself is a proposition that may sensitize some jurors. For instance, mock jurors hearing a murder trial were more likely to return a guilty verdict than those who heard a burglary trial (Paris, 1985). In criminal trials, there is a general pattern that as the severity of the charges and penalties increases, juries become less and less willing to convict, even if the defendants are viewed as guilty (Grofman, 1985). Indeed, increasing the number of options to return reduced charges seems to enhance the chances that jurors will convict on one (Larntz, 1975). Jurors given the option of returning “guilty but mentally ill” rendered that verdict two thirds of the time instead of using the options of “guilty” and “not guilty by reason of insanity” (Poulson, 1990). The sequence in which instructions about different possible verdicts is given has been investigated. In two experiments, when verdict options were introduced in order ranging from the most to the least severe, mock jurors rendered harsher verdicts than when the possible verdicts were arranged from least severe to most severe (Greenberg, Williams, & O’Brien, 1986). To convict on a case involving severe penalties, jurors seemed to demand increased amounts of evidence (Thomas & Hogue, 1976), and they usually reduced their overall rates of conviction (Kerr, 1978). Sometimes trials involve multiple counts of crimes alleged against defendants. Experiments (Tanford, 1985; Tanford & Penrod, 1984) using both actual jurors and student jurors and controlling for similarity of the charges, evidence, and judges’ instructions found that defendants charged with three crimes were most likely to be convicted. Jurors also showed great confusion in keeping track of which evidence related to which charges despite the efforts of judges to provide clear instructions.
Verdicts are affected by the burden of proof shouldered by prosecutors and plaintiffs. The standard of proof required for a decision against the defendant may be proof “beyond a reasonable doubt,” “by a preponderance of the evidence,” or “clear and convincing evidence.” As the standard of proof increases, rates of guilty verdicts decline (see Kagehiro, 1990). Yet when the judge defined “beyond a reasonable doubt” as one’s being firmly convinced of guilt, rates of guilty verdicts for murder increased (Koch & Devine, 1999). In civil cases, jurors who took notes during the trial made objectively sound decisions on compensatory awards when the evidence modestly or strongly favored the plaintiff (ForsterLee & Horowitz, 1997).
Formal presumptions on the interpretation of evidence are meaningful elements of jurors’ decision making when jurors are made aware of them. Three types of presumptions (conclusive, mandatory, and Permissive) were presented to student jurors as part of criminal trials in which the defendant blameworthiness was manipulated (Schmolesky, Cutler, & Penrod, 1988). The conclusive presumption used to suggest defendant guilt increased the numbers of guilty verdicts, although other presumptions did not. Even so, when ratings of defendant culpability were high, jurors were willing to discount presumptions that tended to benefit the defendant.
In civil cases, the higher the damages plaintiffs request, the more money they seem to receive. Of course, judging whether this pattern is causal required experimental work holding the trial evidence constant. Such work with mock jurors supported these expectations (Hastie, Schkade, & Payne, 1999; Raitz, Greene, Goodman, & Loftus, 1990) across types of Personal injury cases, regardless of victim sex and race (Malouff & Schutte, 1989).
The impact of pretrial publicity remains a concern among constitutional scholars as well as Persuasion researchers. The American Bar Association reported that pretrial publicity descriptions of 27% of criminal suspects were “problematic” (Imrich, Mullin, & Linz, 1995). Not surprisingly, most research indicates that pretrial publicity influences jurors (Constantini & King, 1980-1981; Ogloff & Vidmar, 1994; Padawer-Singer & Barton, 1975; Tans and Chaffee, 1966; Tarrence, 1991/1992; Wright & Ross, 1997; for a meta-analysis, see Steblay, Besirevic, Fulero, & Jimenez-Lorente, 1999), especially when it involves eyewitness identifications (Devenport, Studebaker, & Penrod, 1999), and character evidence (Otto, Penrod, & Dexter, 1994), regardless of whether the trial is short or long (Kramer & Kerr, 1989). Although most pretrial publicity was inimical to defendants (Imrich et al., 1995; Riley, 1973; Tankard, Middleton, & Rimmer, 1979), sometimes it benefited defendants, as when the pretrial publicity reports also included the suggestion that there were racist intentions behind spreading the publicity (Fein, Morgan, Norton, & Sommers, 1997), the publicity involved stories of mistaken identification of an innocent man (Greene & Loftus, 1984), or the publicity involved trials similar to the defendant’s (Greene & Wade, 1988). Yet another study found no effects from pro-defendant pretrial publicity (Riedel, 1993).
Naturally, pretrial publicity affects jurors differently. Mock jurors with high ego levels tended to become less confident in their verdicts (Freundlich, 1984/1985). Among jurors with strong pretrial beliefs in the defendant’s guilt, ceiling effects prevented identifying further effects from prejudicial pretrial publicity (Finkelstein, 1994/1995). One study (Hoiberg & Stires, 1973a) found that lowly intelligent women presented with strongly biased pretrial publicity of a heinous rape were most influenced by it, although men and highly intelligent women were not. Even among men, the publicity sometimes can backfire. Men presented with pretrial publicity portraying Perpetrators of acquaintance rape as predators responded with increased pro-defendant judgments in an acquaintance rape case (Mullin, 1997; Mullin, Imrich, & Linz, 1996).
Attorneys and judges seem assured that intensive voir dire examinations to select jurors protect against pretrial publicity bias (Hans & Vidmar, 1986, pp. 63-78). Attorneys rarely present change of venue motions, and 88% of judges have never ruled on one (Siebert, Wilcox, & Hough, 1970, pp. 4-6). Research does not suggest that active voir dire actually reduces pretrial publicity effects (Dexter, Cutler, & Moran, 1992; Kerr, Kramer, Carroll, & Alfini, 1991; Kerr, Niedermeier, & Kaplan, 1999; Sue, Smith, & Pedroza, 1975; but see Padawer-Singer, Singer, & Singer, 1974). Methodologically, using student jurors tended to inhibit pretrial publicity effects (Steblay et al., 1999), as did publicity related to some types of crimes. Dismissed and retained jurors did not seem to report different verdicts when the pretrial publicity involved emotional or factual material (Kerr et al., 1991; Sue et al., 1975). Another method of control, a continuance or delay for a couple of days, appeared to help partially overcome the effects of pretrial publicity bias (Kramer, Kerr, & Carroll, 1990). Even when a judge admonished jurors to disregard any pretrial publicity, student jurors, especially women, still were influenced by the damaging publicity (Sue, Smith, & Gilbert, 1974).
Despite the results based on experimental and laboratory studies, Bruschke and Loges (1999) launched a broadside attack against such study results. Taking the view that laboratory research on the subject is unrealistic, these authors argued for the use of survey methods. Examining the opportunity of jurors to be exposed to publicity (as tracked by the Lexis/Nexis database of print media), verdicts, and sentences in 134 federal murder trials, they found no significant association between publicity and conviction rates. Indeed, the low-publicity condition was associated with the highest conviction rates. Another study (Bruschke & Loges, 2001) of newspaper publicity related to federal murder and robbery trials in Atlanta, Detroit, and Los Angeles found no publicity effects on verdicts, and after an interaction effect was examined, sentence recommendations were inversely related to the amount of publicity. Hence, when the actual amount of pretrial publicity is experimentally controlled, influences on juror decisions exist. When estimates of juror opportunity to be exposed to print news are used (excluding television news as a source), the damaging effects of publicity seem limited at best.
Voir Dire Messages
Voir dire, or the questioning of potential jurors to select a jury, may be a Persuasive process in which attorneys build rapport, obtain commitments, preview the case (Rieke & Stutman, 1990, pp. 70-71), introduce the client in a favorable light, begin arguing the case, guide the jury in its methods of deliberating (Crump, 1980), and familiarize the jury with relevant factual and legal concepts (Mauet, 1980, p. 31). In practice, statements and questions from judges and attorneys accounted for approximately 60% of the sentences uttered during the voir dire “questioning” of potential jurors (Johnson & Haney, 1994). Defense attorneys used voir dire more aggressively than did prosecutors (Johnson & Haney, 1994), and attorney-led questioning stimulated more candid self-disclosure than did judge questioning (Jones, 1987). Ironically, in California, where great restrictions have been placed on attorney questioning, judges were pleased with their own abilities to conduct voir dire examinations, although they believed that attorneys took too much time with juror questioning (Smith, 1994/1995). Dismissed jurors tended to score low in positive self-disclosure (Wigley, 1986/1987; but see Wigley, 1995); high in verbal aggressiveness, dominance, and contentiousness (Wigley, 1999); and high in communication apprehension with friends (Wigley, 1986/1987). When voir dire questions were completed individually rather than en masse, attorneys raised increased numbers of objections to jurors for cause (Nietzel & Dillehay, 1982). Yet this aggressive questioning still often failed to exclude jurors who were opposed to basic foundations of the U.S. legal system (Johnson & Haney, 1994).
In addition to securing information to identify favorable or unfavorable jurors, questioning strategies frequently included building the case, introducing a theme, and “humanizing” the defendant (Waddell, 1988/1989). At the very least, voir dire processes taught jurors the importance of setting aside their ordinary decision-making processes and deciding cases based on the law (Balch, Giffiths, Hall, & Winfree, 1976). Moreover, jurors exposed to extensive voir dire examination Perceived the defendant as less culpable than did other jurors (Dexter et al., 1992; Moran, Cutler, & Loftus, 1990). Indeed, lawyers were least effective in securing desired verdicts when they avoided the use of voir dire questions strategically designed to influence jurors (Arsenault & Reinard, 1997). Such strategic questions functioned by altering Perceptions of the defendant’s character that, in turn, affected verdicts (Reinard, Arsenault, & Geck, 1998). Another study found that strategic questions asking jurors to overlook the defendant’s undesirable past reduced verdicts of guilt, whereas the absence of any strategic voir dire questions negatively affected defendant credibility ratings (Reinard & Arsenault, 2000). Extending this work into contrasts of strategic questions attempting to promote a sense of juror rapport or empathy with the defense, Reinard, Khalid, and Liso (2001) found that verdicts for the defense were enhanced by the use of questions that requested jurors to reciprocate positively to the defense expressions of trust in them.
Attorneys’ opening statements are supposed to provide a case preview rather than extended argument. But because trials are forums for argument, and distinguishing argumentative from nonargumentative statements can be difficult in practice, one should not be surprised that opening statements have powerful argumentative functions (see Perrin, 1999, esp. pp. 110-132), even though such strategies as fanciful name calling (People v. Johnston, 1994), directly refuting the opposition’s case (e.g., State v. Bell, 1972), and interpreting upcoming evidence (Strong, 1992, pp. 17-19) have been ruled argumentative. Indeed, an examination of 19 opening statements found that all were subject to some objection from the opposition, the plurality of which involved “various types of circumstantial evidence—using a theme, drawing inferences for the jury, characterizing a Person or event in some argumentative way, or discussing the mental condition of a Person” (Perrin, 1999, p. 140). Opening statements also were believed to be influential given that the jurors tended to make decisions early in the trial process (Freundlich, 1984/1985; Pyszczynski & Wrightsman, 1981; Stone, 1969; but see Weld & Danzig, 1940). Pettus (1990) learned that “critical moments” jurors identified as decision points were during first sight of the defendant, opening statements, and presentation of the prosecution arguments. Across varying circumstances, opening statements have not been found as influential as closing arguments (Walker, Thibaut, & Andreoli, 1972; Wilson & Miller, 1968; Zdep & Wilson, 1968). The influence of the prosecution’s opening statement was reduced when the defense argument that followed was strong (Wallace & Wilson, 1969). One might suppose that attorneys prepare opening statements carefully. Yet a review of 50 trials revealed that juries thought defense opening statements were less well prepared than prosecutors’ statements, even though, of course, defense attorneys did not Perceive any inferiority in themselves (Linz, Penrod, & McDonald, 1986). The first opening statement juries hear seems to be particularly important in a couple of ways. First, the first opening statement serves a strong agenda-setting function. The first opening statement appears to “prime the pump” such that jurors’ assessments of witnesses and evidence tend to follow its organization (Bayly, 1988/1989). Second, the length of the first opening statement appears to affect reactions to others that follow. The defense enhanced the likelihood of securing not guilty verdicts if it responded to prosecutors’ brief opening statements with extensive ones (Pyszczynski & Wrightsman, 1981). Curiously, guilty verdicts were increased when the defense made lengthy opening statements in response to lengthy prosecution opening statements. Similarly, the decision to delay an opening statement until halfway though a trial can damage the chances of a defendant’s securing the desired verdict (Wells, Wrightsman, & Miene, 1985). Jurors relied most on the message to which they were first exposed (a primacy effect) when jury deliberations were delayed a week following their hearing both messages (Miller & Campbell, 1959). When there was a 1-week delay before hearing the second message, the message heard last was most influential (a recency effect). Other work did not support the primacy effect, although the recency effect was observed (Insko, 1964; Wilson & Miller, 1968).
Opening statement structure has been examined. It appeared to make no difference whether the opening statement used a “story” format or a legal comparison-expository structure (McCullough, 1991), regardless of whether the attorney represented the plaintiff or the defendant (Spiecker, 1998/1999). If the defense attorney’s opening statement promised evidence exonerating the defendant and then failed to introduce such proof, the prosecution’s reminder to the jury (in this case, presentation of an alibi witness) reduced the defense attorney’s influence on the verdict (Pyszczynski, Greenberg, Mack, & Wrightsman, 1981). Yet if the prosecution failed to remind jurors of the broken promise, the defense opening statement increased the likelihood of an acquittal. In fact, if a potential witness were absent, the mere mention of that fact by the judge or by the attorney affected jury decision making against the side that was expected to produce the witness (Johnstone, 1993/1994).
The effectiveness of attorney delivery during the opening statement was not influenced by whether the attorneys were male or female (Barge, Schlueter, & Pritchard, 1989). The initial manner of delivery seemed to determine Perceptions of the appropriateness of disfluencies in the rest of the message. A lawyer’s great vocal variation stimulated increased Perceptions of dynamism, but it also tended to promote Perceptions that the attorney was not as trustworthy or friendly as others. Contrariwise, other work found that increased fluency and gestures during the opening statement were unrelated to source credibility but vocal variety reduced ratings of character and competence, while increasing volume and variety in speaking rate enhanced Perceptions of friendliness (Rockwell & Hubbard, 1999).
Much lore and a little research have been dedicated to the closing statement. Jurors reported believing that the closing argument was vital, some rating it as second only to the presentation of evidence (Matlon, Davis, Catchings, Derr, & Waldron, 1985). Furthermore, 75% of lawyers believed that the closing arguments could have decisive effects in close cases (Walter-Goldberg, 1985). Disturbingly, in death penalty cases, improper statements made by prosecuting attorneys in the closing statements increased the rates at which jurors voted for the death penalty (Platania, 1995/1996). Significant relationships existed among actual jurors’ favoring the prosecutor’s closing argument and juror recall, belief, and interestingness. Yet no relationship was found among these elements and juror verdicts. Some studies noted a general recency effect favoring the influence of closing arguments over opening statements (Insko, 1964; Miller & Campbell, 1959; Walker et al., 1972; Weld & Danzig, 1940; Wilson & Miller, 1968; Zdep & Wilson, 1968). But in a critique of these primacy-recency studies, Benoit and French (1983, p. 389) warned that many of these studies included conditions for control purposes that were incompatible with actual courtroom practice.
The message structure may make a difference. A test of the story model revealed that the legal comparison-expository format for the closing statement was superior to the use of narrative structure (McCullough, 1991). Structurally, when a one-sided closing argument was followed by a two-sided closing argument, the two-sided closing argument was most Persuasive (Insko, 1962). In this context, a two-sided presentation does not just mention the opposition but actually responds to it. Student jurors who heard a two-sided summation followed by a one-sided summation favored the first attorney. When jurors were familiar with the case issues, a two-sided presentation was most influential generally, although unfamiliar jurors were most influenced by one-sided case summations (Dipboye, 1977).
Defense Case Strategies
Defendants may deny charges, or they may present “affirmative defenses” that offer outside explanations as excuses. When simply denying the charges, exonerating and incriminating facts about the defendant were most influential on intensifying jurors’ verdict dispositions when the attorney presented the information in a heterogeneous, rather than a homogeneous, order (Kaplan & Miller, 1977). Affirmative defenses seem difficult to argue. For instance, when defendants gave affirmative defenses for retracting confessions, they were more likely to be convicted of murder than when pleas included appeals to the Fifth Amendment or denial of the charges by reference to an alibi (Fischer & Fehr, 1985). Even so, an affirmative defense including testimony to prove mitigating circumstances was effective in reducing mock jurors’ recommended sentence (Suggs & Berman, 1979). When defendants invoked the Fifth Amendment by itself, simulated jurors increased rates of guilty verdicts (Shaffer, Case, & Brannen, 1979). In one experiment, mock jurors placed approximately the same proof obligations on both sides in a murder case, regardless of the defense strategy (simple denial or self-defense) and despite the judge’s instruction about the location of the burden of proof (Posey, 1995/1996). In a medical malpractice lawsuit, damage awards were cut in half when the plaintiff was shown to be partially negligent (Zickafoose & Bornstein, 1999).
Arguing the insanity defense is particularly challenging (see Simon, 1967; Winslade & Ross, 1983) and appears to succeed slightly less than 1% of the time (Cirincione, Steadman, & McGreevy, 1995; McGinley & Pasewark, 1989) and only when defendants suffer from the most severe disorders (Boardman, 1995/1996; Lymburner & Roesch, 1999). Insanity pleas were most successful when a mentally ill defendant was accused of a property crime (Kidd & Sieveking, 1974) as well as when jurors were assured that the defendant accused of a non-murder charge would be in a mental hospital for an extended Period (Carroll, 1981/1982). There are many reasons insanity defenses are risky. In the first place, the insanity defense is often misperceived and unpopular with the public (Borum & Fulero, 1999). In the second place, defining “insanity” has proven to be quite nettlesome, and lawyers and psychiatrists sometimes bewilder each other (Gutheil, 1999). Because insanity is not a medical term describing a particular malady, complicated rules such as the McNaughton, the Durham, the Brawner, Michigan’s Guilty But Mentally Ill statute, and the American Law Institute’s Model Penal Code have formed various notions of insanity. When jurors have been tested on their understanding of the various rules on which they have been instructed, comprehension has averaged between 40% (Arens, Granfield, & Susman, 1965) and slightly more than 50% accuracy (Elwork et al., 1981; Ogloff, 1993). Moreover, student jurors exposed to trial excerpts were unable to identify the defendant’s mental state when a crime was committed despite hearing depositions on the matter (Whittemore & Ogloff, 1995). Although an early literature review (Casper, 1964) claimed that jurors tended to make their decisions based more on the nature of the crime than on the psychiatric evidence, surveys of a large county’s records in Oregon (Steadman, Keitner, Braff, & Arvanites, 1983) and 128 cases in Virginia (Jones, 1995/1996) found a close association between pretrial psychiatric evaluations and verdicts (see also supportive studies by Jeffrey, 1986/1987; Slutsky, 1975/1977). Women who successfully argued insanity differed from others in their past criminal records and in the nature of psychiatric testimony about them at trial (Morris, 1992/1993). Overall, jury assessments appeared to be highly consistent with pretrial psychiatric evaluations, evincing a correlation of .68 (Jones, 1995/1996). Moreover, in a murder trial, path analysis found that mock jurors’ evaluations of the evidence were stronger influences than their initial attitudes (Poulson, Brondino, Brown, & Braithwaite, 1998). Expert testimony and juror reasoning were significant predictors of verdicts (Hernandez, 1985/1986), especially when the jurors had low issue involvement (Wursten, 1986/1987). Similar work has found that student jurors presented with conflicting testimony from expert witnesses in a murder case viewed a medical doctor as more credible than a psychologist (Belon, 1991). Furthermore, an unbiased expert’s testimony (indicated by revealing that the expert often gives legal testimony for both sides) stimulated more verdicts in the direction of the testimony than did the biased expert.
Claims of insanity were most Persuasive when student jurors were presented with evidence that the crimes were committed in a strange (Pickel, 1998) or “bizarre” manner (Boardman, 1995/1996) without a “reasonable” (Pickel, 1998) or “criminal” (Boardman, 1995/1996) motive. Whereas those found not guilty of homicide by reason of insanity usually killed family members or parents, those found guilty usually had killed their spouses or lovers, often while under the influence of drugs or alcohol (Nestor & Haycock, 1997). Women who had killed their children and subsequently were successful in pleading not guilty by reason of insanity tended to attempt suicide after the killings and did not kill the children because they were unwanted or because the acts were revenge against wayward husbands (Holden, Stephenson-Burland, & Lemmen, 1996). Initially, it was found that defendants were most likely to use the insanity defense successfully if they were uneducated, diagnosed as schizophrenic, not drug users, and not young (Pasewark, Jeffrey, & Bieber, 1987). Some research reported that African American defendants, especially males who were diagnosed as schizophrenics (Linhorst, Hunsucker, & Parker, 1998), were most likely to win their insanity defenses (Poulson, 1990). Yet other work found no significant race effects when socioeconomic status was controlled (Towers et al., 1992). Survey work has not settled these matters. A 7-year survey in Knox County, Tennessee, revealed that violent young White men were most successful in their insanity defenses (Windham, 1990/1992). Bogenberger (1989/1990) found that defendants who were successful in their insanity defenses had prior diagnoses of psychosis or insanity, were unemployed, and were not young. When mock jurors in another study heard an insanity defense in a case involving either serious or trivial matters, the jurors were most willing to attribute reduced responsibility to lowly intelligent defendants, especially when the crimes were not serious (McGraw & Foley, 2000). Sometimes an insanity defense is not introduced until the penalty phase of a capital case. In one study, mock jurors heard any of four defense strategies (White, 1987). Although a defense based on mental illness was least effective, a conceptual argument against the death penalty Per se was most Persuasive.
Some jurors are more receptive to the insanity defense than are others. Among participants drawn from actual juror venires, images of insanity and individuals for whom jurors were willing to consider such a defense defied simple identification. But the match between jurors’ images of insane defendants and the actual defense portrayals was important in deciding the cases (Skeem, 1999). Those jurors most willing to accept insanity defenses were young and highly educated, without strong moralistic attitudes (CasPer, 1964), and with high levels of emotional empathy for others (Poulson, Wuensch, Brown, & Braithwaite, 1997). Least willing to accept insanity defenses were jurors with highly authoritarian Personalities, those with negative attitudes toward the insanity defense and sources of expert testimony (Tezza, 1995/1996), and believers in capital punishment (Poulson et al., 1997). Believers in the death penalty discounted insanity pleas based on nonorganic mental deficiencies such as schizophrenia, but they did not differ from other jurors in accepting insanity based on organic disorders such as mental retardation and psychomotor epilepsy (Ellsworth, Bukaty, Cowan, & Thompson, 1984).
Some states have adopted variations of the insanity plea. The “diminished capacity” defense involves claiming that the defendant lacked the ability to premeditate the crime, in which case the defendant is convicted of a lesser offense than the one charged. In a multi-year survey, women pleading diminished responsibility were more successful than were men (Mitchell, 1997). In another variation, some states Permit declaring a defendant “guilty but mentally ill,” in which case the defendant is sentenced to prison following treatment for the mental illness. When jurors were given this alternative, they tended to use it two thirds of the time (Poulson, 1990; see also Bourdouris, 2000). Without this option, even jurors who believed the defendant was insane voted guilty a disproportionate amount of the time (Poulson, Wuensch, & Brondino, 1998). A literature review noted that the guilty but mentally ill verdict option increased confusion among jurors (Palmer & Hazelrigg, 2000) and often was used when innocent defendants were tried by jurors with low levels of control (Seymour, 1986) or as a way to signal reduced blame toward the defendant (Roberts, Sargent, & Chan, 1993). Although the option has tended to be used most for White males who were seriously mentally disturbed (Callahan, McGreevy, Cirincione, & Steadman, 1992), defendants who pleaded straight insanity and were found “guilty but mentally ill” were given more severe sentences than those who were found guilty.
Responding to such a defense can be tricky. Prosecutors usually attempt to show that the defendant premeditated and carefully planned the crime, thus suggesting that the crime was not an impulsive and irrational decision. Yet student jurors used such data to judge the wrongfulness of the conduct but not the defendant’s capacity to control his or her actions (Roberts & Golding, 1991).
Recovered Memory Cases
In some cases, counselors and hypnotists have encouraged children and adult witnesses—often believed to be victims of sexual abuse—to testify about “events” they had “forgotten” or “repressed.” Although this controversial approach has been attacked as little more than the power of suggestion, in a criminal case “repressed” testimony was only slightly less influential on convictions (58%) than was “nonrepressed” testimony (67%) (Key, Warren, & Ross, 1996). Whether jurors believed in such evidence depended on the Perceived strength of evidence, their belief in the repression of memory, and the kinds of mass media news stories they had heard about the credibility of recovered memories (Rosen, 1996/1997). In child abuse cases featuring “repressed memory” evidence, victims who were either 3 or 13 years old at the time were not as believable as 8-year-old victims (Key et al., 1996). In defense against repressed memory cases, presenting counterexperts to testify about false recovered memories blunted the impact of the prosecution evidence (Rotzien, 1997/1998). Recovered memory cases—with their full retinue of expert witnesses—were most influential with female jurors, especially in cases involving female victims (Griffith, Libkuman, Kazen, & Shafir, 1999), and in civil cases with female, highly religious, and highly authoritarian jurors (Schutte, 1994). In different work, many jurors were found to doubt the credibility of hypnosis, as indicated by a sample of actual jurors who judged hypnosis-induced memories as less reliable than immediate recall but more reliable than 1 week’s delayed recall (Greene, 1986).
The Entrapment Defense
One way to respond to criminal charges is by claiming that defendants were induced to engage in illegal behavior by law enforcement officials. Because such a strategy admits the illegal behavior itself, it requires subtle argument, and jurors have had difficulty understanding judges’ instructions on the defense (Borgida & Park, 1988). As a result, juries have relied predominately on evidence of the defendants’ past criminal conduct to reach verdicts. Lewis (1997) found that jurors in a drug case carefully tracked information about the number of times that the defendant refused enticements from law enforcement officers. If the defendant initially turned down suggestions from law enforcement officials, the jury grew increasingly receptive to the entrapment defense.
The Battered Woman Defense
The defense has been made that a “battered woman” is justified in homicide as a form of self-defense brought about by her chronic beating by an abusive spouse. Research is mixed on whether expert testimony has great (Kasian, Spanos, Terrance, & Peebles, 1993) or limited influence on jurors’ decisions (Mechanic, 1996/1997; Schuller & Hastings, 1996). In one study, jurors were most influenced by an expert witness who also was female (Schuller & Cripps, 1998). In another study, jurors were exposed to a wife accused of killing her abusive husband as he attacked her (Terrance, Matheson, & Spanos, 2000). Jurors using the “objective standard” to assess defendant fear were less convinced of the defendant’s innocence than were those using the “subjective standard.”
Survey research has shown that case evidence is the most potent single influence on verdicts (Bayly, 1988/1989; Bridgeman & Marlowe, 1979; Cassidy, 1993/1994; Pettus, 1990; Poulson, Braithwaite, Brondino, & Wuensch, 1997; Weld & Danzig, 1940). In observations of trials and interviews with 331 jurors in sexual assault cases, matters other than evidence influenced jurors most often when the prosecution had weak cases with little hard evidence (Reskin & Visher, 1986; see also Feild, 1979). In follow-up work, juror decisions were dominated by case evidence, particularly physical evidence, although sentiments toward victims and defendants sometimes played a secondary role (Visher, 1987). Experimental work has revealed that case evidence influenced jurors more than did sentiments toward victims (Cather, Greene, & Durham, 1996) and defendants (Clary & Shaffer, 1985). Evidence of a confession, for instance, affected verdicts of student jurors more than did eyewitness identification or character testimony (Kassin & Neumann, 1997). Although student jurors were able to reject confessions that were the result of threats of punishment against the defendants, they were strongly influenced when confessions were either freely given or the products of offers of leniency (Kassin & Wrightsman, 1980). On the other hand, character evidence suggesting that defendants were not the sorts of people who would be likely to commit the alleged crimes did not affect mock jurors’ verdicts unless the prosecution cases included little relevant evidence (Smith, Stasson, & Hawkes, 1998-1999). In criminal cases, verdicts seemed to balance initial bias toward the defendants with the Perceived guilt appearance of the evidence (Kaplan & Miller, 1978). As the importance of one factor increased, the influence of the other factor declined. As jurors’ positive attitudes toward defendants increased, evidence played a less and less important role in predicting convictions (Kaplan & Kemmerick, 1974; Perry, 1976).
What makes evidence strong or weak remains an open issue. One explanation was offered by Pettus (1990), who summarized some of her interview data: “The clear and well told story is considered effective evidence, whereas the unclear, nonsensical story is considered ineffective evidence” (p. 92). The tone set by key witnesses can make a difference. When testimony against a defendant in a murder case was introduced in an opinionated manner, judgments of guilt increased. But when the testimony was presented in an unopinionated way, student jurors were likely to find the defendant guilty of the lesser charge of manslaughter (Ludwig & Fontaine, 1978). Beyond verdicts alone, jurors’ overall interpretations were affected by the placement of judges’ instructions to jurors (Elwork, Sales, & Alfini, 1977). When the judge explained the burden of proof before the presentation of evidence, conviction rates declined (Kerr et al., 1976). Furthermore, when an active judge commented on testimony from a doubtful eyewitness by issuing limiting instructions, summarizing relevant testimony, and explaining the role of eyewitness testimony, the jury disregarded the troubled material during its deliberations (Katzev & Wishart, 1985).
Witness Evidence. Research on the presentation of evidence by witnesses involves the study of testimony from defendants, children, eyewitnesses, and experts. In addition, the impact of inadmissible evidence has been investigated.
Defendant testimony may be presented in a manner that affects juries. In one of the few controlled experiments on the subject, three stereotypical signals of lying (fidgeting, avoiding eye contact, and disfluencies) were used in a deposition by an African American defendant accused of breaking and entering (Pryor & Buchanan, 1984). Not surprisingly, student jurors rated the defendant as least guilty when the signs of nervousness were minimized. Other work found reduced defendant believability when the deposition was presented in a case involving a minor crime but not a major crime (Feldman & Chesley, 1984). A defendant’s decision not to testify is not supposed to affect jurors, but it does. In experimental work, jurors over 60 years old thought that the defendant’s failure to testify indicated an attempt to hide guilt (Zeiglar, 1978). Yet fewer than 15% of jurors under 30 years old held such a view. Highly dogmatic jurors were particularly harsh on such defendants (Shaffer & Case, 1982). Field studies show a mixed record. Defendants who took the stand in Indiana were most likely to be convicted (Myers, 1979), whereas defendants who testified in Utah were least likely to be convicted (Werner, Strube, Cole, & Kagehiro, 1985).
Children called as witnesses are most Persuasive when they appear confident, intelligent, and ready to provide details (Goodman, Goldings, & Haith, 1984). The literature suggests that juries tend to stereotype children as essentially honest but easily manipulated by parents or lawyers, sometimes unable to distinguish reality from imagination, and generally unsophisticated thinkers (Goodman et al., 1984). Nevertheless, when a child’s testimony is uncorroborated or attacked successfully during cross-examination, its Persuasiveness can be largely eliminated. In sexual abuse cases, otherwise dramatic influences of child witnesses appear to be ameliorated somewhat by the use of defense expert witnesses who can place the testimony in context (Kovera & Borgida, 1996). On the other hand, corroborating expert testimony enhanced the impact of the children’s testimony (Kovera, Gresham, Borgida, Gray, & Regan, 1997), especially when the expert did not present material about child sexual abuse accommodation syndrome (Kovera, Levy, Borgida, & Penrod, 1994). Survey work with actual jury venires revealed that decisions were strongly influenced by the confidence of the mental health professionals who assessed the credibility of children’s reports (Corder & Whiteside, 1988).
The child’s age makes a difference. When a 6- or 12-year-old child testified in a sexual abuse case or in a Personal injury lawsuit, expert corroboration of the testimony enhanced Persuasiveness (Nightingale, 1993). This corroboration had no influence when the witness was 9 years old. When the witness was 14 years old, the child’s influence was not particularly enhanced by expert corroboration. Yet others (Crowley, O’Callaghan, & Ball, 1994) found no relationship between children’s ages and influence, although expert testimony alone was Persuasive. Female student jurors were most likely to rate a child’s testimony as credible and to vote guilty in child abuse cases. It seemed to make no difference whether the child’s testimony was presented live or on videotape (Hagen, 1990/1991).
Eyewitness testimony can be quite influential with jurors (Lindsay, 1994), with as many as 83% of mock jurors unduly influenced by such evidence (Brigham & Bouthwell, 1983). Yet the fallibility of eyewitness testimony is well established and need not be reviewed here. Lindsay, Lim, Marando, and Cully (1986) exposed mock jurors to various numbers of eyewitnesses testifying for the prosecution and defense in a purse-snatching case. Although unopposed witnesses were most effective, internally inconsistent testimony (including alibi testimony) also was Persuasive for the defense. Whether the eyewitness identified or failed to identify the suspect made a difference. Student jurors were exposed to two types of nonidentification information (eyewitness testimony and fingerprint evidence) in situations where there were one or two eyewitnesses or contradictory evidence (McAllister & Bregman, 1986). Regardless of the evidence type, identifications influenced decisions more than did nonidentifications (i.e., statements that a Person was not at a location). In other inquiry, when eyewitness evidence and fingerprint evidence were contrasted, positive examples of each were most influential on verdicts, and the absence of interactions indicated that the effects were independent of each other (Bregman & McAllister, 1987). Similarly, when an alibi eyewitness claimed that the defendant either was or was not at another location when a crime was committed, mock jurors tended to discount nonidentification information if it failed to meet their expectations (McAllister & Bregman, 1989). The confidence of eyewitnesses seemed directly related to their influence (Cutler, Penrod, & Dexter, 1990; Cutler, Penrod, & Stuve, 1988; Fox & Walters, 1986; Lindsay, Wells, & O’Connor, 1989; Penrod & Cutler, 1995; Sporer, Penrod, Read, & Cutler, 1995). Even though confidence can be manipulated by lawyers’ preparation of witnesses (Luus & Wells, 1994), jurors seem to be relatively unaffected by elements such as cross-examination and judges’ instructions regarding the fallibility of eyewitness testimony. In another study (Wells, Lindsay, & Ferguson, 1979), even though juror Perceptions of witness confidence were unrelated to witness accuracy, these Perceptions predicted nearly 50% of the variance in jurors’ willingness to believe witnesses. As it turned out, witnesses often expressed their confidence not through nonverbal cues but simply by making assertions such as “I am 100% sure” of an identification. Even when the levels of evidence incrimination (corroboration with physical evidence) and Perceived confidence of the eyewitness were experimentally manipulated, both factors produced significant and independent effects on mock jury verdicts (Moore & Gump, 1995). The type of information to which the witness testified made a difference. In three experiments, Whitley and Greenberg (1986) found that mock jurors assessed eyewitness accuracy in identifying and describing the criminal suspect based on Perceptions of witness expertise. On the other hand, eyewitness confidence affected witness Persuasiveness when the issue involved the description of the crime. For instance, eyewitnesses to an auto-pedestrian incident were most influential when increasing the details of their testimony (Bell & Loftus, 1988). Disturbingly, when accurate eyewitnesses had a difficult time recalling trivial elements of the identifications, jurors became less confident of them as sources of information (Wells & Leippe, 1981).
To reduce the influence of faulty eyewitness testimony, three techniques have been explored. First, psychologists have been called to offer advice to jurors, although courts have been reluctant to embrace such an approach (see Raimo, 1987). One study (Wells, Lindsay, & Tousignant, 1980) found that having a psychologist testify that eyewitness confidence had nothing to do with identification accuracy increased doubt in the eyewitness testimony. Including testimony from a psychologist explaining the fallibility of eyewitness testimony reduced conviction rates from 47% to 35% in an assault case and from 68% to 43% in a murder case (Loftus, 1980). Such results have been supported with adult jurors (Hosch, Beck, & McIntyre, 1980), but not always (Warner, 1987/1988). To reduce rates of guilty verdicts, the information did not even have to be presented by a psychologist (Cutler, Dexter, & Penrod, 1989; Paris, 1985; but see Fox & Walters, 1986). Sometimes judges’ cautionary instructions about eyewitness testimony affected jury decisions (Ramirez, Zemba, & Geiselman, 1996). Using a court-appointed expert did not seem to make a difference. Yet when expert testimony exposed problems with identifying the suspect from a police lineup, no significant juror skepticism about lineup testimony occurred (Devenport, 1996/1997; Devenport, Cutler, & Penrod, 1998). Moreover, psychologists’ comments were not influential when unrelated to the specific testimony under consideration (Maas, Brigham, & West, 1985). Second, some have suggested that active voir dire could be a safeguard, but research did not indicate that juror overestimation of eyewitness credibility was reduced by this method (Narby & Cutler, 1994). One would be surprised if voir dire inhibited eyewitness influence given that community members were somewhat likely to confuse the influence of stress, fear, and witness confidence with the credibility of such testimony (Rahaim & Brodsky, 1982). Third, a view has been argued that eyewitness fallibility can be revealed by active cross-examination. Student jurors exposed to experienced and inexperienced lawyers engaging in cross-examination continued to rate the testimony of accurate and inaccurate eyewitnesses as equally credible (Lindsay et al., 1989). Lawyers’ abilities made no difference. In other words, when an eyewitness was discredited, the eyewitness was still more Persuasive on verdicts than when no eyewitness testified at all (Whitley, 1987). Without the direct contradiction of an eyewitness by another eyewitness, the original testimony remained very influential (Leippe, 1985).
Expert witnesses tended to be influential, especially when the testimony was presented early in the trial and when it was specifically linked to the case under consideration (Brekke & Borgida, 1988). In contrasts of expert witness varieties (e.g., physicians, psychiatrists, psychologists, chemists, document examiners, polygraph examiners, police, eyewitnesses, firearms experts, accountants, appraisers), members of the professions were most Persuasive with actual jurors, although other experts could be influential when the relevance of their testimony could be established (Saks & Wissler, 1984). Jurors were strongly influenced by testimony from psychiatrists (Mc-Mahon, 1974). In somewhat related inquiry, judges and lawyers considered forensic evidence on mental health to be most useful when it dealt with clinical diagnoses and analyses of whether legal thresholds had been met (Redding, Floyd, & Hawk, 2001). In another study, jurors were Persuaded by expert presentation of data about groups of people most likely to commit rape (Brekke & Borgida, 1988). In a civil case, expert economic testimony of the plaintiff’s loss of wages and benefits seemed influential to juries (Greene, Downey, & Goodman-Delahunty, 1999), whereas juries tended to discount the lawyer’s recommendations about “pain and suffering” awards. Expert psychologists’ race and sex did not affect credibility or Perceived effectiveness in the eyes of student jurors (Miyatake, 1998/1999). Other research (Dravin, 1982/1983) found that although female expert witnesses were Perceived as most credible by student jurors in shoplifting and assault cases, no significant influence was seen on actual jury verdicts. In medical malpractice cases, although one study found that the most Persuasive expert was an African American woman (Memon & Shuman, 1998), another study found that male experts were Perceived as more authoritative than female experts (Monroe, 1993/1994).
Not all expert testimony is influential. Although expert testimony about repressed memory was Persuasive to mock jurors who heard only the testimony favoring the recovered memory notion (Reince, 1991), such expert testimony was not Persuasive to male jurors or to those who were unsympathetic to the recovered memory concept (Rotzien, 1997/1998). In a rape case, when expert witnesses dueled over whether the victim had rape trauma or had made up her charge of rape as a result of borderline Personality disorder, the introduction of expert witnesses by the prosecution backfired and the acquittal rate increased (Burnstein, 1995/1996).
Inadmissible evidence sometimes is presented to jurors. Although few claim that jurors can “unlisten” to what they have heard, the most common remedy is for judges to instruct jurors to disregard the questionable material. Yet such instructions are not always effective. In the first place, few jurors seem able to remember the instructions to disregard the testimony (Hoffman & Brodely, 1952). In the second place, jurors seem able to disregard only mundane information (Hirsch et al., 1976; Konopka et al., 1974; Miller & Fontes, 1978; Poole, Lefebvre, Miller, & Fontes, 1975; Reynolds, 1977). Retracted confessions (Kassin & Wrightsman, 1981), evidence of a past criminal record (Greene & Dodge, 1995; Hans & Doob, 1976; Pickel, 1995; but see Cornish & Sealy, 1973), and emotional materials remained influential and were amplified by judges’ instructions (Edwards & Bryan, 1997). Similar effects also were found on Perceptions of witness credibility (Horn, 1976). Whereas some work using mock jurors observed that anti-defendant information was not ignored (Lenehan & O’Neill, 1981), another study found that pro-acquittal inadmissible testimony also was not ignored (although pro-conviction evidence sometimes was) (Thompson, Fong, & Rosenhan, 1981). Disturbingly, the little research on the subject indicates that judges are not suPerior to juries in ignoring biasing or inadmissible material (Landsman & Rakos, 1994; Wells, 1992).
If there is a pattern to the matter, it seems that jurors use inadmissible materials along with other case arguments and evidence when they believe such information is true and relevant (Mosmann, 1998). Hearsay evidence introduced to mock jurors during expert testimony was readily dismissed, and the judge’s instructions further reduced the Perception that the witness’s conclusions were supported (Schuller & Paglia, 1999). Moreover, jurors were most likely to reject inadmissible materials when the judge gave detailed explanations why coerced confessions were both unreliable and unjust (Kassin & Wrightsman, 1981); when wiretap evidence in a murder trial was explained to be unreliable (as opposed to reasoning that it violated due process protections) (Kassin & Sommers, 1997); in the unlikely event that the admonition preceded the inadmissible material (Kassin & Wrightsman, 1979); when jurors were given reason to be suspicious about the motives of the advocates and witnesses who introduced the inadmissible evidence (Fein, McCloskey, & Tomlinson, 1997); when an official source (Reinard, 1989), especially a police officer (Reinard, 1981) or a county government fingerprint expert (Reinard, 1985), introduced the inadmissible materials; when the jurors had negative attitudes toward the criminal justice system and the police (CasPer & Benedict, 1993); when inadmissible testimony involved a “mild” violation of due process guarantees (Fleming, Wegener, & Petty, 1999); and when jurors in some cases deliberated before rendering decisions (Kerwin & Shaffer, 1994; Weld & Danzig, 1940; but see Reinard, 1985, 1986a). Some jurors seemed more willing than others to reject inadmissible materials. Sommers and Kassin (2001) found that when student jurors reading a trial summary were high in need for cognition, they not only were willing to discount inadmissible materials identified as unreliable but also tended to “overcorrect” by assessing defendant guilt lower than did others. In related but different work using adult jurors, Sutton (1979) found that inadmissible evidence about the character of the accused was curvilinearly related to guilt ratings and that testimony alleging the low moral character of the defendant in a rape case was most easily discounted.
It seems that inadmissible testimony is more damaging to some types of defendants than to others. In an effort to include fairly emotional information, Reynolds (1977) presented simulated jurors with a rape trial involving an Anglo victim. He found that when controlling for homophily and race of the defendant, inadmissible testimony damaging to the defendant predicted 22% of the variance in jury verdicts. Wolf and Montgomery (1977) presented mock jurors with a criminal trial about a barroom brawl in which the defendant was alleged to have stabbed the victim with a broken glass. In the experiment, the questionable evidence was ruled either admissible, inadmissible, or inadmissible accompanied by the judge’s instruction to disregard the evidence. The researchers reported that biasing effects were eliminated when the judge ruled the material inadmissible, whereas the testimony was amplified when the judge ordered jurors to disregard it. Reinard and Reynolds (1978) found that raising an objection to inadmissible testimony in a criminal trial amplified it despite the judge’s ruling. When the objection was overruled, the bias was greatest. The seeming contradiction about the impact of judges’ instructions was explained by Loh (1985), who drew a distinction among limited admission of criminal background, limited admissions, and total exclusion of such materials. Careful reading of the research revealed the following:
Conviction rates were highest under unlimited admission, lowest under complete exclusion, and in-between under limited admission … Instructions on limitation are effective (in the sense of reducing guilty verdicts) when compared to the admission of the evidence without any instructions. They obviously are not effective or are less effective, however, when compared to the exclusion of the evidence. (p. 27)
In civil cases (Cox & Tanford, 1989), inadmissible testimony by itself was most Persuasive when favoring the defendant instead of the victim or others in the trial. Adult jurors presented with inadmissible evidence were unable to disregard inadmissible evidence when objections to it were sustained unless the judge also gave a specific instruction on the matter (Schaffer, 1984/1985). Evidence of insurance coverage by the defendant led to increased awards in cases involving auto collisions (Broeder, 1959) and attacks by a homeowner’s escaped pet dog (Reinard, 1993). In each instance, the judge’s instructions seemed to amplify the illicit impact of such evidence.
How inadmissible testimony produces its effects has been investigated. Developing a causal model of the impact of inadmissible testimony, Reinard (1989) found that the sentence recommendations were a function of the verdict and the character Perceptions of the defendant. The nature of the inadmissible testimony, the type of witness introducing it (official or nonofficial source), and the Perceived character of the defendant influenced verdicts.
Manner of Evidence Presentation. Not surprisingly, sensational materials possessed by prosecutors may influence decisions if judges Permit their introduction. Sensational crime scene photos seemed to influence jury decisions (Russo, 1992), although to produce such an effect, grisly video recordings had to be relevant to the case (Kassin & Garfield, 1991). Juries easily discounted recreations, as in the case of video reenactments in a wrongful death civil case (Fishfader, Howells, Katz, & Teresi, 1996). Attorneys in one sample expressed their opinions that jurors were suspicious of animated presentations of information if they were overly realistic because such materials created an impression that the attorneys were attempting to manipulate the facts (Cusick, 1994/1995). Yet the use of prototypes or models of crime scenes seemed to enhance Persuasion (Filkins, 1996/1997).
Jurors often have a difficult time in dealing with scientific and statistical information, resulting in confusion and reduced influence of such evidence. In one study involving DNA evidence, jurors (particularly women) separated or combined probabilities incorrectly and gave such probabilistic evidence less influence than it should have received (Schklar & Diamond, 1999). Juries sometimes do not know what to do with statistics. For instance, Thompson and Schumann (1987) examined a case in which blood typing evidence showed that only 1 Person in 100 shared the actual Perpetrator’s blood type. Fully 60% of the jurors decided for the defendant when the defense responded by arguing that such numbers meant that in the city of 100,000 people, 1,000 people had the same blood type. Using tortured statistical sophistry, the defense attorney opined that because the defendant was only 1 of the 1,000 people in the city with the culprit’s blood type, that fact meant that there was only 1 chance in 1,000 that the defendant committed the crime. The problem of dealing with statistical evidence has been vexing, and proper interpretation of statistics presented by expert witnesses remains at an unsatisfactory level (see Fienberg, 1989, esp. pp. 149-189).
The language and manner of trial participants can influence jurors. Non-Latino jurors tended to judge defendants who testified through translators as more likely to be guilty than English-speaking defendants (Stephan & Stephan, 1986). Similar results were found for Thai defendants speaking through translators. Dialects also introduced similar effects. Members of a jury pool in Hawaii who listened to direct examination of witnesses, including speakers of Hawai’i Creole English, rated speakers of Standard English as most credible (Takakawa, 1999). The phrasings of defendants also affected jurors. A defendant who loudly denied the charges and intemperately protested against them was likely to be Perceived as more guilty than a defendant who testified with a tone of moderation (Yandell, 1979). In a case involving a husband’s assisted suicide of a terminally ill wife, an emotional explanation of how the defendant helped in the suicide amplified the sympathy or hostility that the jurors had toward the means of suicide (disconnecting a respirator or shooting through the wife’s skull) (Pfeifer, Brigham, & Robinson, 1996). Yet a study of 13 trials showed no changes in credibility or believability when the witness made repeated use of disclaimers in reaction to assertions from the questioning attorneys (Stutman, 1986b).
Occasionally, witnesses appear in some form of disguise or identity masking, such as when government agents or organized crime informants present evidence. One study attempted to check the effect of electronic masking of videotaped testimony (Towell, Kemp, & Pike, 1996). None of the methods (placing a gray circle over the witness’s face, using pixelation masking, using negation of the witness’s face, repeating out of synchrony video images of the courtroom, or placing a static image of a witness’s face before jurors as the audio portion of the testimony was played to them) affected witness credibility ratings, but all methods except the negation and out of synchrony conditions impaired jurors’ ability to remember the testimony.
Another area of interest has been the use of powerful and powerless speech. Powerless speech consists of language that reduces dominance in an exchange such as unnecessary intensifiers, tag questions, hedges, and indirect language. Attorneys seemed to emphasize powerful speech styles. Parkinson (1981) found that successful prosecutors were verbally assertive, took command of the courtroom, spoke at length, asked large numbers of questions, usually employed direct language in the indicative case, and referred directly to witnesses. On the other hand, unsuccessful prosecutors tended to use many expressions of politeness, excessively correct grammatical speech, and increased numbers of conditional statements. Successful defense attorneys tended to use large amounts of abstract and ambiguous language, increased legal jargon, reduced numbers of adverbs, and reduced numbers of words expressing affects. Losing defense attorneys tended to use language with much demonstrative pronoun use and increased grammatically correct speech. Whether these elements were contributors to success or artifacts remained unknown. Powerful speech styles by witnesses for the plaintiff stimulated increased awards (Erickson, Lind, Johnson, & O’Barr, 1978) and made witness testimony increasingly believable (Stutman, 1986b) and credible (Chapman, 1993/1994). In particular, dysfluent witnesses, witnesses with poor grammar, and witnesses who used lower class language styles were viewed as less credible than those whose speech approximated Standard English (Conley, O’Barr, & Lind, 1978). A combination of powerful verbal and nonverbal cues also affected Perceptions of witness credibility (Lisko, 1992/1993; see also Lind & O’Barr, 1979; Scherer, 1979). A meta-analysis of powerless speech research in both courtroom and non-courtroom settings found an average correlation effect size of .23 with Persuasion, corresponding to 5.3% of the variance explained (Burrell & Koper, 1998). Yet not all research has supported this effect on awards (Bradac, Hemphill, & Tardy, 1981). Manipulating the use of hedges and intensifiers, Wright and Hosman (1987) exposed participants to testimony from male and female witnesses in a traffic collision case. Although verdicts were not measured, women who used the hedges gained in credibility. Similar results were found in other courtroom applications (Hosman & Wright, 1987).
Direct Examination and Cross-Examination
Direct examination is the chance for one side to introduce evidence to support essential claims. One might imagine that the opposition would have its own plan for undermining this set of claims through cross-examination. But in a study of rape trials, whatever the direct examination covered tended to be reviewed in the same order in all subsequent questioning (Sanford, 1987). The agenda-setting function of the direct examination is quite strong. Although attorneys might wish to read a physician’s deposition into the record, controlled experimentation in an industrial accident case revealed that direct testimony was most influential on the awards granted (Jacoubovitch, Bermant, Crockett, McKinley, & Sanstad, 1977). Court-appointed experts in a rape trial were not more influential than witnesses called by attorneys and subjected to detailed cross-examination (Brekke, Enko, Clavet, & Seelau, 1991). Based on observations of actual trials, Antieau (1998/1999) found that attorneys using indirect language in direct examination enhanced jurors’ favorable impressions of witnesses. The phrasing of questions made a significant difference in the answers secured and, as a consequence, the potential influence produced. In one study, simply asking a car crash eyewitness, “About how fast were the cars going when they smashed into each other?” increased mock jurors’ Perceptions of the speed of the cars more than when the collision was described as cars that “hit” or “contacted” each other (Loftus & Palmer, 1974). Similar work involving car crashes revealed that using unmarked adverbs (e.g., “How fast was the car going?”) during witness questioning induced witnesses to report that the car was moving faster and that more damage was done than when the attorney asked questions with very specific phrasing (Lipscomb, McAllister, & Bregman, 1985). Jurors also seemed to pay attention to the practical implications of witness answers. If the witness made a statement with indirect language (e.g., “After I heard the shot, I went to the telephone”), jurors completed the implication in their own minds (e.g., inferring that the witness made a phone call, probably to the police) (Harris, Teske, & Ginns, 1978). Nonverbal elements sometimes can make a difference. One writer has suggested that nonverbal cues are potentially influential but that only those who are sensitive to such cues, typically women, are likely to be influenced by them (LeVan, 1984). Another legal scholar (Dombroff, 1988, p. 340) asserted that a lawyer’s tone of voice may be the most significant factor causing objections to be raised to questions asked during examinations.
Cross-examination can be very influential, sometimes even reversing juror decisions (Spanos, Dubreuil, & Gwynn, 1991-1992). In the case of eyewitnesses, defense attorney cross-examination that exposed inconsistencies in testimony reduced rates of conviction among mock jurors (Berman & Cutler, 1996). These effects were great even when the inconsistencies dealt with Peripheral rather than central case facts (Berman, Narby, & Cutler, 1995). Even among highly credible witnesses, exposing inconsistencies in testimony resulted in discounting the evidence (Devine & Ostrom, 1985). The phrasing of cross-examination questions can influence jurors. A lawyer’s asking for very brief responses to specific questions stimulated the impression that witnesses were not as competent, intelligent, or assertive as those whose answers were not so constricted by the cross-examiner (Conley et al., 1978). Cross-examining attorneys often use leading questions that suggest desired answers. In one study of 42 cross-examination sessions, student jurors exposed to extensive use of leading questions during cross-examination were more likely to believe accurate than inaccurate witnesses, whereas nonleading questions were most likely to stimulate belief in inaccurate witnesses (Wells et al., 1979). Even so, the use of leading questions and questions designed to increase control over the witness resulted in decreased amounts of the witness’s testimony (Stutman, 1986a). Among the “dirty tricks” of cross-examination is posing a question that presumes information not in evidence and asking the witness to react to it. When mock jurors heard an expert witness presented with a presumptive question casting as persions on a rape victim’s character, the target’s credibility was unaffected (Kassin, Williams, & Sanders, 1990). But the expert witness’s credibility was reduced, even when the content of the question had been denied or when the question produced an objection and jurors were instructed to ignore it.
Apparently, invasions of witness space during direct examination and cross-examination are not particularly effective. Field observations revealed that attorneys tended to shorten the distance between themselves and witnesses more during cross-examination than during direct examination (Brodsky, Hooper, Tipper & Yates, 1999). But the impact on juries seemed to be a negative one. The authors suggested that proxemics in interpersonal settings simply are not the same as the structural proxemics in the courtroom setting.
Judges are considered the captains of the court. Although lawyers’ lore probably exaggerates this effect, there is little doubt that judges’ comments affect jurors. Jurors have been found to return verdicts designed to please the judge (O’Mara, 1972). In one experiment with jury-eligible adults, even when admonished to disregard the judge’s behavior and form their own opinions, jurors returned verdicts in accordance with the judge’s bent (Hart, 1995). Even elements of judges’ nonverbal activity during trials affected jury attention as well as Perceptions of witness credibility and attractiveness (Badzinski & Pettus, 1994).
The judge’s instructions are designed to influence jurors to promote the cause of justice, but they have often featured jargon, passive voice, and odd syntax (Buchanan, Pryor, Taylor, & Strawn, 1978). Indeed, 45% of judges’ instructions were misunderstood (Charrow & Charrow, 1979). Even in capital cases, where courts have paid great attention to refining instructions, jurors tend not to understand them, and closing arguments by attorneys have seemed impotent to overcome the defect (Haney & Lynch, 1997).
In trials involving technical evidence, judges sometimes introduce instructions before presentation of such evidence. Jurors given instructions before hearing the evidence were able to make clear compensation distinctions among four plaintiffs involved in the case (ForsterLee, Horowitz, & Bourgeois, 1993).
Jury attitudes toward the law have affected the impact of judges’ instructions (Pryor, Taylor, Buchanan, & Strawn, 1980). Because juries are charged with attempting to advance the cause of justice, some hold that juries have the right to ignore (nullify) a law that they believe is unfairly applied. A controversy involves what happens if the judge informs jurors of the opportunity to nullify. Some work found that judges who introduced the nullification option actually influenced jurors to return reduced numbers of guilty verdicts (Davis, 1998; Pfeifer et al., 1996). Other research did not find this effect (Niedermeier, Horowitz, & Kerr, 1999). Willingness to entertain the nullification option depended on juror attitudes toward the law, Perceived physical attractiveness of the defendant, and schemata about the crimes involved (Clark, 2000).
Jurors as Audiences
Because few researchers have been allowed in jury rooms, most evidence on jury deliberations derives from simulations or interview studies. Nevertheless juror elements affect their decision making on the evidence and their dispositions toward various judgments.
Decision Making on the Evidence
Common folk wisdom and some research (MacCoun & Kerr, 1988; McCoy, 1997/1998; Reinard & Arsenault, 2000) hold that jurors rarely change their verdicts as a result of deliberations. This thinking is based on the simple observation that most juries start with a majority sentiment. Because the odds favor jurors in the minority changing their minds, any statistics would naturally indicate that most juror decisions were unchanged as a result of deliberations. But of course, there is more to this story. Jurors in the majority generally were effective in influencing others (Marcus, Lyons, & Guyton, 2000), a fact that gave the appearance of polarizing the direction of mock jury verdicts. This pattern was observed when evidence was presented in a heterogeneous order (Kaplan & Miller, 1977; see also MacCoun & Kerr, 1988) and when jurors were not sound reasoners (Weinstock, 1999). This influence appeared as jurors articulated their own arguments and as the influence of witnesses and trial participant credibility decreased (Sonaike, 1977).
Yet deliberating jurors may be restricted in their discussion of issues. Interviews with jurors in Oregon and California revealed that the capital sentencing instructions limited the realm of legitimate discussion and discouraged jurors from considering the moral aspects of their decision making (Haney, Sontag, & Costanzo, 1994). To trace the flow of Persuasive communication among jurors, efforts have been made to identify Persuasive functions. Kaplan (1983) explained that deliberating jurors might influence each other with messages that serve either of two functions. First, informational influence takes the form of information sharing and direct Persuasive argumentation. Second, normative influence consists of pressures to conform to the group. According to this model, informational influence occurs as a result of responding to the mass of information. Whereas informational influence is directed at changing subjective or private beliefs, normative influence is focused on changing public beliefs. Although one might speculate to the contrary based on small group communication research, verdicts were unrelated to how much a juror talked, the juror’s sex, or the juror’s initial confidence in his or her opinion (McCoy, 1997/1998).
A tradition of lawyers’ lore has produced contradictory advice about juror dispositions and has often substituted opinions for answers to empirical questions (see Fulero & Penrod, 1990a, 1990b). Although jury selection—and nearly all of work called scientific or systematic jury selection—is rooted in juror demography, survey research has not always supported the notion that such matters influence either procedural or outcome variables (Bridgeman & Marlowe, 1979). In their development of the Juror Bias Scale, Kassin and Wrightsman (1983) found that conviction proneness was predictable from an interrelated set of factors, including authoritarianism, acceptance of a relaxed view of the “reasonable doubt” standard, conservative political views, belief in a just world view of reality, and adoption of an internal locus of control. In fact, when one looks at the nature of the jury variables, most research has shown that these elements accounted for very small proportions of effects. Moreover, when scholars have looked at the power of systematic jury selection to identify receptive juries, support has ranged from partial (Brand, 1985/1986) to supportive of the “survey approach” (Frederick, 1984). Controlled experimental studies contrasting various techniques, however, found that “scientific jury selection” was not particularly effective in selecting favorable juries (Baker, 1984/1985; McGowen, 1981). Even so, research interest has remained in investigating jury characteristics such as juror sex, social and economic status, age, race/ethnicity, prior experience, Personalities, and death qualification.
The influence of juror sex has produced a mixed record. Some work found that men were more conviction prone than women (Angira, 1991; Elkins & Philips, 1999; Simon, 1967; Steffensmeier, 1977), some studies found that women were more conviction prone and recommended more severe sentences than men (Austin et al., 1976; Griffitt & Jackson, 1973; Scroggs, 1976), and some studies found no differences on verdicts and recommended sentences (McCoy, 1997/1998; Nemeth, Endicott, & Wachtler, 1976) unless the prosecution evidence proved to be weak (Rosen, 1996/1997). Student jurors were least likely to find a defendant of their own sex guilty (Stephan, 1974). In civil cases, women generally seemed most plaintiff prone (McCullough, 1991).
In rape cases, women were both more likely to return guilty verdicts and to recommend severe sentences than were men (Beauvais, 1982/1983; Calhoun et al., 1978; Feldman-Summers & Lindner, 1976; Lyons & Regina, 1986; Miller & Hewitt, 1978; Rumsey & Rumsey, 1977; Spanos et al., 1991-1992; Yanchar, 1982/1983; but see Jones & Aronson, 1973; Scroggs, 1976; Smith et al., 1976; Kahn et al., 1977), especially when defendants and victims were from different races (Hymes et al., 1993). Not only were men less likely to find rape defendants guilty in the presence of strong evidence (McNamara, Vattano, & Viney, 1993), but men were increasingly likely to believe that the rape victims were partially to blame for the rapes (Calhoun et al., 1978) and that the victims may have been careless or behaved provocatively (Smith et al., 1976). Women were more likely than men to accept evidence of a recovered memory used against a defendant (ForsterLee, Horowitz, Ho, ForsterLee, & McGovern, 1999; Griffith et al., 1999; Key et al., 1996; Rotzien, 1997/1998). In non-rape cases, females were more likely than men to accept an affirmative defense such as insanity or self-defense (Posey, 1995/1996).
Men and women differed in a variety of other dispositions. In cases involving the insanity defense, women were more likely than men to accept such a defense, especially when the defendant was female (Towers, 1996/1997). Women also were more likely than men to favor the plaintiff in a sexual harassment case (Gowan & Zimmerman, 1996), to recommend increased damage awards in a medical malpractice suit (Bensko, 1995/1996), to find a child witness credible in a child abuse case (Crowley et al., 1994), and to render guilty verdicts in child abuse cases (Hosch, Chanez, Bothwell, & Munoz, 1991). Men tended to be more influenced by expert testimony than did women (Kovera, McAuliff, & Hebert, 1999). In general, women were more likely to conform to the views of the majority of the jury than were men (Prager, 1995/1996).
Juror Social and Economic Status
Upper class jurors tended to be more conviction prone than lower class jurors (Rose & Prell, 1955), especially as the disparities between the juror and defendant socioeconomic status grew (Adler, 1973). In civil cases, the situation is not so clear. In wrongful death suits, although juror sex, age, marital status, and occupation were unrelated to damages awarded, juror beliefs and attitudes about monetary damages influenced verdicts (Goodman, Loftus, & Greene, 1990). Juror income also was only weakly related to amounts awarded (Hastie, Schkade, & Payne, 1998). The political background of the jurors also seemed to make a difference, with political conservatives awarding reduced damages in a malpractice suit (Bensko, 1995/1996). Highly educated jurors tended to return more guilty verdicts than did others (Reed, 1965). When the jurors and the defendant had the same religious affiliation, decisions tended to be more lenient than when their religions were different (Kerr et al., 1995).
Although sometimes it is believed to play a major role in predicting decisions, age does not seem to produce consistent effects. Some work indicated that young jurors had an increased tendency to acquit the defendant (Stephen & Tully, 1977), but others found no significant effects (Goodman et al., 1990). During deliberations, age groups did not differ in their rates of changing their verdicts (Prager, 1995/1996).
A common attorney belief is that African American jurors are acquittal prone (Turner, Lovell, Young, & Denny, 1986). Although highly authoritarian jurors had a preference for defense attorneys of their own race (Boliver, 1999), when defendant socioeconomic status was controlled, neither the defendant’s race nor the jurors’ race affected verdicts or sentence recommendations (Towers et al., 1992). When African Americans were on the jury, no negative biases against African American defendants were significant (Tinsley, 1991/1992). In a child neglect case, Anglos and Latinos did not differ in their judgments (Hosch et al., 1991). In civil cases, ethnicity was slightly related to amounts awarded, with Anglos awarding slightly less than did members of ethnic groups (Hastie et al., 1998). In product liability cases, minority members and poorly educated jurors from the community most often found for the plaintiff (Bornstein & Rajki, 1994).
Juror Prior Experience
Jurors who previously served on juries may have distinct views of what jury service involves (Durand, Bearden, & Gustafson, 1978), but the influence of this experience on verdicts remains inconclusive. In one study of 143 criminal trials in Kentucky, jurors with experience in either civil or criminal trials rendered more severe verdicts in criminal cases than did first-time jurors (Himelein, Nietzel, & Dillehay, 1991). Another analysis of 206 criminal cases revealed a similar effect for small juries composed of a majority of experienced jurors (Werner et al., 1985), as did a review of 175 criminal trials throughout a year (Dillehay & Nietzel, 1985). To test whether this effect could be verified experimentally, students either were or were not given experience as mock jurors and subsequently were put on mock juries (Kassin & Juhnke, 1983). No significant differences were found between verdicts of experienced and inexperienced jurors.
Juror Personality Characteristics
Although one’s degree of general or social intelligence was unrelated to verdicts (Becker, 1998/1999), defense-prone jurors tended to be high in autonomy and most likely to advocate acquittal when given instructions on nullification (Laird, 1997/1998), high in attributional complexity (similar to cognitive complexity) (Pope & Meyer, 1999), and low in alienation (Claghorn, Hays, Webb, & Lewis, 1991). Locus of control research is mixed. Some studies in criminal venues showed that jurors whose internal locus of control was high were least likely to convict (Kurdys, 1983/1984), especially when the jurors shared similar attitudes with the defendant (Kauffman & Ryckman, 1979). Yet other research has found no effect on verdicts regarding locus of control (Villemur & Hyde, 1983). In civil cases, jurors with a high internal locus of control viewed plaintiffs as contributing increasing amounts of liability, and, hence, they reduced awards (Rowland, 1991; Sosis, 1974). Jurors who scored high in boredom proneness were more critical in their verdicts of a male defendant in a civil case than were others (Salmons, 1995/1996). When a child abuse case was presented to student jurors, those jurors who scored high on the “hostility-guilt” scale did not differ from others in verdicts, although they tended to recommend shorter sentences than did others (Ackerman, McMahon, & Fehr, 1984).
Authoritarian Personalities tend to be resistant to influence unless the source is an authority figure. Although a meta-analysis found that authoritarianism predicted guilty verdicts (Narby, Cutler, & Moran, 1993), many studies did not show this effect (Shay, 1987/1988; Weir & Wrightsman, 1990). Many times the authoritarianism appeared to produce its effects as part of an interaction with other variables. In a case involving an insanity defense, highly authoritarian jurors tended to render high rates of guilty verdicts (Tezza, 1996). Yet jurors with high emotional empathy tended to render more not guilty by reason of insanity verdicts than did others. Other studies found that high authoritarians tended to oppose insanity defenses (Cutler, Moran, & Narby, 1992) and were willing to consider the prosecution’s expert testimony on repressed memory evidence in a civil case (Schutte, 1994). Highly authoritarian mock jurors who were presented with a defendant who possessed dissimilar attitudes to their own recommended harsh verdicts and sentences (Mitchell & Byrne, 1973). When the defendants were from races or socioeconomic groups different from those of the jurors, high authoritarians gave the harshest verdicts and sentences to those who were similar to them (McGowen & King, 1982). In one study, mock jurors who scored high in authoritarianism were most malleable and willing to change their views of the verdicts during jury deliberations, but authoritarians were not more prone than others to return guilty verdicts (Lamberth, Krieger, & Shay, 1982). In another study, highly authoritarian mock jurors hearing a murder trial most often rendered guilty verdicts and recommended increasingly severe sentences. But these jurors showed increased verdict changes during deliberations (Bray & Noble, 1978). Yet other studies found that high authoritarians were willing to change their minds during deliberations (Berg & Vidmar, 1975). The seeming contradictory results for authoritarians were explained by noting that individuals with this character trait were prone to comply with those whom they believed to be in positions of authority. Thus, the conviction proneness would be most pronounced for those defendants of low status and least pronounced for those of high status. When a defendant simply used the defense that the action was taken in an effort to obey a suPerior authority, highly authoritarian jurors tended to respond more favorably than did lowly authoritarian jurors (Hamilton, 1978). Akin to high authoritarians, juries composed of dogmatic Personalities tended to return more severe decisions than did others (Shaffer et al., 1986). Highly dogmatic jurors presented with a homosexual defendant in a murder case viewed the defendant’s case less favorably than when the defendant was straight (Shaffer & Case, 1982). Actual jurors in criminal cases who were highly dogmatic tended to identify with victims or with authority figures on either side, and this bias was reflected in their verdicts (Wyman, 1984/1985).
“Death-qualified” jurors are those identified as willing to apply the death penalty. Interestingly, jurors who were death qualified showed increasingly harsh judgments in non-capital trials as well. Such jurors were more conviction prone than were other jurors (Bronson, 1970; Cowan, Thomson, & Ellsworth, 1987; Filkins, Smith, & Tindale, 1998; Haney, 1984; Jurow, 1971; Moran & Comfort, 1986; but see Russell, 1991). Some research found that death-qualified women were most likely to convict the defendant (Seguin & Horowitz, 1984). A meta-analysis revealed that belief in the death penalty was positively related to a juror’s returning a verdict of guilty (Allen, Mabry, & McKelton, 1998). As is the case with most jury variables, the effect size was small, accounting for an average of only 3% of the variance.
An Interim Status of Legal Persuasion Research
Social science inquiry in legal Persuasion has grown to a stage where general claims of research paucity no longer can be made. Particularly in the areas of the impact of trial participants, types of evidence/testimony, and jury dispositions, considerable progress has been made over the years. Somewhat curiously, however, research on the impact of voir dire as Persuasion, on the Persuasiveness of general argument strategies, and on organization of materials remains in its infancy. It is with some irony that one observes that Persuasion in the legal setting has involved more inquiry into non-message factors than into central message elements. For scholars of legal communication, the priority on the influence of message cues would seem particularly invited.
Because many studies of legal Persuasion have reflected variable by variable inquiry, advances in the study of Persuasion in the law have suffered from the failure to guide much research by relevant theories or conceptualizations. To be sure, the story model has focused much research, but the failure to provide consistent support for this theory may be taken as a general sign that the search for theoretic alternatives still is required. Some have thought of enlisting the contributions of other theories designed for the general study of Persuasion, and there is little question that such an approach may be of some benefit. Yet it also must be recognized that the legal setting is a genre of communication (the forensic setting) that also is typified by distinct forms of language, lines of argument, and case construction forms. Thus, it would make sense for legal Persuasion theorists to search for new analogies to guide research or, at least, to adapt existing theories to the unique influences of the legal setting. Such projects require movement beyond simple appropriation of influence theories to the full development of new applications.
Despite the contributions of research in legal Persuasion, considerable criticism— often from practitioners—has been directed at the entire domain of study. The primary charges involve issues of research using mock jurors and the realism of laboratory research in legal Persuasion. Each of these issues invites comment.
Much research on legal Persuasion has enlisted mock juries in simulated trials. One might ask whether results drawn from such samples are generalizable to actual jurors. McGaffey (1983) put the question briskly: “Many of our early experiments were limited to college sophomores. Now we are trying to see if real people react differently” (p. 251). Surely, there is some merit in this reservation, especially among research projects studying jury characteristics and demographics. Of course, despite witticisms and doubts to the contrary, students are people, and they also are called to serve on juries. But most jurors are not students, and a reasonable concern for generalizability should be a concern of serious researchers and practitioners alike. There are two approaches to this matter. First, one may dismiss all research using simulated jurors. This step would leave a tidy collection of survey research on actual jurors to guide drawing some conclusions. Such work is identified in this chapter for those wishing to take this Perspective. In the areas where survey research has not been completed, attorneys might rely on their own opinions or on those of consultants. With no real test of such opinions, however, it would be difficult to draw any general conclusions with confidence. For all the limits one might place on simulated jury research, it would be difficult to prove that sampling mock jurors is a priori inferior to reliance on opinions alone. It might be mentioned that occasionally claims are made that consultants have completed research to answer many relevant questions but that they cannot reveal such work due to proprietary contracts. But it is difficult to assess such assertions or the quality of the research. Indeed, as one skeptical scholar explained, “If you ask the trial consult-ant[s] for their data, with all due respect, they don’t really have it” (Simon, 1983, p. 287). On such a matter, the “jury” may still be out. Second, researchers might examine the limits of laboratory results by the use of corroborative surveys to test boundaries of relationships found with mock jurors. This approach would reveal an interesting pattern. There appear to be few studies to show the generalizability of student jurors’ findings on impact of organismic variables (e.g., Personality, age, socioeconomic status). Nevertheless, when defendant source and message variables are examined (e.g., inadmissible testimony, compliance with the judge’s instructions), relationships found among student jurors tend to be found among adult juries as well. Researchers have long found that community and student jurors looking at defendant characteristics seemed to deliver the same verdicts and sentences (Bray et al., 1978). When the impact of message variables is assessed, student jurors do not seem to report results different from adult juror samples (Bray et al., 1978; Feild & Barnett, 1978; Nunez et al., 1999; Reinard, 1985; Tanford, 1985; see also a review by Bornstein, 1999) or from those of actual jury venires (Cutler et al., 1989; Cutler, 1990; Zickafoose & Bornstein, 1999). Regarding source and victim variables, on the other hand, noteworthy differences have been found between student and adult populations. In one study, adult and student jurors differed in the Perceived Persuasiveness of an expert witness in a battered woman defense case (Schuller & Hastings, 1996). Also involving a battered wife incident, student jurors were found to hold fewer stereotypes about battered women than were adult jurors from voter registration lists (Aubrey & Ewing, 1989). Another study revealed that adult and student jurors differed in attribution assessments of an assault victim who was portrayed as either a male or a female (Howard & Leber, 1988). Student jurors’ attributions of the source sex variable were more influential than those of adult jurors. Attributions of adult jurors were more influenced by testimony information than was the case with student jurors. Additional research seems invited to identify where results with student jurors can and cannot be replicated with adult jurors. Hence, of the two options to the sampling issue, the second choice seems to make the most temperate use of available information.
The realism of mock trials has been criticized. Although experimental exposure to trial reenactments (often abbreviated for the practical purposes of data collection) may be required for control purposes, realism concerns may limit the external validity of research findings. This criticism is particularly troublesome in experiments where either of two limitations exists. First, when jurors are presented with trial summaries rather than reenactments, the trial communication effects might not find broad application. One group of researchers found that mock jurors returned verdicts that were different from that returned by the actual jury from which the trial manipulation was taken (Bermant, McGuire, McKinley, & Salo, 1974). It seems that researchers must commit themselves to the use of trial reenactments in some medium. Some practitioners doubt that any manipulations employing trials abbreviated in any manner can be realistic enough to justify drawing conclusions. Yet the research suggests that salient variables might not decay as rapidly as critics fear. Contrary to a popular expectation, for instance, a study found that as the length of the trial increased, the influence of trial variables, such as defendant attractiveness, did not seem to be reduced in impact (Kramer & Kerr, 1989). Of course, the presence of uncontrolled variables in field settings might ameliorate the influence of nearly any factor under the right circumstances, but in controlled research such effects could be identified and studied. Moreover, such claims about the richness of uncontrolled variables also place limits on conclusions drawn by practitioners and consultants. A second problem in some research is the failure to include juror deliberations. In their summary of simulation research in legal communication, Miller and his associates concluded that trial simulations were useful provided that jury deliberations were included and that due attention was paid to the types of jurors and the realism of the trial (Miller, Fones, Boster, & Sunnafrank, 1983). Other research has found that jury deliberations sometimes can overcome biasing effects created by defendant attractiveness (Izzett & Leginski, 1974) and some forms of inadmissible testimony (Kerwin & Shaffer, 1994; Weld & Danzig, 1940). Sound research in legal Persuasion appears to require that designs include deliberations as a distinctive element in developing juror reactions. Some critics have doubted that research even is possible. For instance, this claim was made at a national conference on legal advocacy: “Whatever model is used, the researcher will never be able to find out what really goes on during the deliberative process of even one real jury because seclusion and secrecy are themselves significant factors in the jury process; if we observed or bugged the jury room, we would change the process” (McDermid, 1983, p. 284). If taken literally, this concern might lead one to conclude that no social science research should ever be done in legal Persuasion and rarely in any applied area of the social sciences. Furthermore, such a position may be an example of using the ideal in research to exclude the adequate. Of course, this concern for an empirical study’s external validity is not new to social science researchers in any field. Empirical researchers in any field constantly address such issues. But it should be remembered that external invalidity—even if found—does not mean that the results of a study are unsound. The relationships found in laboratory research are real relationships among variables, and the data do not go away because of attacks on their breadth of application. Instead, external invalidity limits the situations in which the results may be applied. Thus, concerns for external validity should be accompanied by alternative explanations of applications for the data by practitioners and critics before they are casually dismissed. In passing, it might be mentioned that some social scientists take the position that because they are strictly researchers of behavior, it is not their job to solve problems in the legal setting and so questions of such applications are not their concern. Even so, it remains the case that most legal Persuasion researchers were drawn to the field out of a legitimate sense that relevant research in legal Persuasion holds the exciting potential to enrich understanding of both processes of Persuasion and processes of justice. The concern for external validity gives rise to a third opportunity. The concern for realism among consultants, lawyers, and researchers might be taken as an invitation for consultants and lawyers to sponsor research with funds sufficient to support development of increasingly realistic materials with professional actors and professional video production. Because research in this area is typically unsupported financially, compromises to realism might be eliminated if those who wish to benefit from the research help to support its completion. In reality, much—if not most— experimental research has relied on trial recordings presented to jurors. One might wonder whether the absence of a live trial adversely affects realism. Yet the use of videotaped trials does not seem to alter jury verdicts either in civil suits (Miller & Fontes, 1978) or in Perceptions of children’s testimony (Davies, 1999).
Research and theorizing in legal Persuasion remains a dynamic process. Even so, it seems that the research to date has emphasized that major trends are visible in the areas of the influence of trial participants, evidence types, and jury variables. For legal Persuasion researchers, however, a gap remains in the development of relevant theory and inquiry into the message variables of research. Such a condition may be taken both as a suggestion for the direction of future work and as a limitation on claims of accomplishment thus far.
1. It should be mentioned that this chapter is limited to trial-related Persuasion, although of course lawyers often have to Persuade their clients during interviews. Furthermore, negotiation and alternative dispute resolution issues are excluded from this discussion largely for practical reasons of space limitations. Similarly, parole board decisions, juvenile justice hearings, and family court hearings are excluded.
2. In his review of the theoretic foundations of communication in the practice of law, Gibson (1991, pp. 31-35) identified the following areas of relevant communication theory: storytelling, temporal density, dramatism, legal action stages, argumentation, linguistics, psychometrics, and language. Although each of these approaches has played a part in communication studies, especially enriching the scope of rhetorical-critical studies in the field, social science inquiry dealing with influence in the legal setting has been dominated by the conceptual discussions listed here.
3. Despite predictions from the Elaboration Likelihood Model, argument and evidence quality—associated with the central route to Persuasion—were most influential among those jurors with a low need for cognition.
4. In this review, a number of studies are excluded such as the following: studies that do not deal with actual trials or at least summaries of them (e.g., Efron’s  study of attraction among college students in an academic discipline setting; Kassin, Reddy, & Tulloch’s  study of evidence order), studies on detection of deception, studies on the role of polygraph evidence, studies of mistaken eyewitness identification (e.g., Cutler & Penrod, 1995), studies of reactions to police reports (e.g., Dean, Holliday, Mack, & Thomas, 2000), studies of the Wells (1992) effect on cognition, studies on the legal systems of nations other than the United States (e.g., Wall & Schuller, 2000; Clark, 1994), studies of courts in which lawyer-advocates do not operate (e.g., small claims court, [Hollihan, Riley, & Freadhoff, 1986]), studies involving jurors under 18 years of age (e.g., Hoiberg & Stires, 1973b), and studies of comprehension of judges’ instructions (e.g., Morier, Borgida, & Park, 1996).
5. A popular figure has been the assertion that 80% of jurors make up their minds during the opening statement (see Perrin, 1999, p. 124). Apparently, this often cited statistic was based on a misreading of the American Jury Project of Kalven and Zeisel (1966, esp. p. 488). These authors found that in nearly 90% of the cases, the jurors’ initial opinions on entering the jury rooms did not change. In addition, because more than 80% of the criminal cases were decided in favor of the prosecution, there was opportunity for the statistic to take on a life of its own as it was retold as trial lore. In 1988, Zeisel responded to users of the misrepresented statistic and encouraged them to use caution (Zeisel, 1988).
6. Some studies claimed to investigate inadmissible testimony but really did not. For instance, Carretta and Moreland (1983); Johnson (1994); Simon (1967); Sue, Smith, and Caldwell (1973); and Werner, Kagehiro, and Strube (1982) presented participants with newspaper-style summaries of trials in which inadmissible testimony was included. Participants did not actually hear trials in which the evidence was or was not presented.
7. Not all consultants’ advice or systematic jury selection is based on juror background alone. Some emphasize the use of nonverbal cues. Vaughan (1986a, 1986b) advised selecting jurors based on the shapes of various body parts. Starr (1979) adapted general nonverbal work to suggest assessing juror dispositions by (among other things) body type, manner of walking to the jury box, and the distance prospective jurors stood from each other during breaks. Such approaches have been controversial. Although they praised scientific jury selection as a form of “community or courtroom research involving systematic attention to the problem of bias,” Rieke and Stutman (1990, p. 78) also accused many commercial research firms of passing off “canards of nonverbal behavior as proprietary wisdom” (p. 79) and compared many jury consultants with astrologers relying on “stereotypical signs, ranging from particular nonverbal behaviors to the dress or cosmetics worn by prospective jurors” (p. 79). Making unqualified translations from one area of research to another is troubling at best (Reinard, 1986b), and Rieke and Stutman (1990) declared, “Evidence for generalizing from juror idiosyncrasies does not exist, and claims based on this evidence are generally without foundation” (p. 79).