Laboratory report

CRIMINAL JUSTICE AND BEHAVIOR Skolnick, Shaw / COMPARING EVIDENCE

A COMPARISON OF EYEWITNESS AND PHYSICAL EVIDENCE ON MOCK-JUROR DECISION MAKING

PAUL SKOLNICK JERRY I. SHAW California State University, Northridge

Two studies compared the effectiveness of eyewitness testimony and physical evidence on mock-juror decision making. Jury-eligible participants were randomly assigned to read one of eight versions of a hypothetical murder scenario and were each asked to render a verdict, to rec- ommend a sentencing option, and to make other evaluative judgments of the defendant. In Study 1, either eyewitness testimony or physical evidence was presented, whereas in Study 2, both types of evidence were presented together. Also, in both studies, the strength of evidence varied. Log linear analysis confirmed that mock jurors’ verdicts and evaluative judgments were influenced to a greater extent by physical evidence than by eyewitness testimony. Strong evi- dence produced more guilty verdicts than weak evidence. However, combining strong evidence of both types was no more effective than presenting strong evidence of either type. Implications of these factors for prosecutors and defense attorneys in criminal proceedings are discussed.

Eyewitness testimony plays a key role in the legal system, insofaras investigation of crimes and identification of criminals by law enforcement officials often hinge on such reports. Even more dra- matic is the powerful influence eyewitness testimony in the courtroom can exert on jury verdicts. Loftus (1974) found that eyewitness corrob- oration of circumstantial evidence nearly quadrupled conviction rates by mock jurors, even when it was known that the eyewitness was visu-

614

AUTHORS’ NOTE: The authors wish to express their gratitude to Edward Binder for his skillful role in helping to develop stimulus materials and in collecting and cod- ing the data for statistical analysis. Correspondence regarding this article should be addressed to Paul Skolnick, Department of Psychology, California State University, Northridge, CA 91330. E-mail correspondence may be sent to paul.skolnick@ csun.edu.

CRIMINAL JUSTICE AND BEHAVIOR, Vol. 28 No. 5, October 2001 614-630 © 2001 American Association for Correctional Psychology

ally impaired at the time the crime was witnessed. Comparable find- ings were reported by Wells, Lindsay, and Ferguson (1979) and by Brigham and Bothwell (1983). Loftus (1979) cited statistics from England showing that nearly three fourths of the cases based solely on eyewitness reports resulted in convictions.

The credibility ascribed to eyewitness testimony is thus consider- able, even though psychologists concerned about its accuracy have repeatedly demonstrated that such testimony is vulnerable to a variety of perceptual and memory errors, which can occur at the time the crime was committed or later during police interviews and courtroom questioning. For example, studies of weapon focus (Kramer, Buckhout, & Eugenio, 1990; Loftus, Loftus, & Messo, 1987; Mitch- ell, Livosky, & Mather, 1998; Pickel, 1999; Shaw & Skolnick, 1994, 1999) show that eyewitnesses may ineffectively process features of a criminal perpetrator who is brandishing a weapon. Or, they may mis- takenly identify an innocent but familiar-looking bystander as the criminal, a phenomenon known as unconscious transference (Buckhout, 1974; Ellison & Buckhout, 1981). Other postevent activi- ties such as viewing mug shots of possible culprits or answering ques- tions during initial interviews can also alter eyewitnesses’ memories (Hall, Loftus, & Tousignant, 1984). The interested reader is referred to Bartol and Bartol (1994), Cutler and Penrod (1995), and Wrightsman, Nietzel, and Fortune (1998) for recent reviews of the eyewitness falli- bility literature.

Another type of evidence presented in the courtroom is physical evidence. This broad category includes tangible objects that are offered into evidence such as a weapon allegedly used in a crime, blood-spattered clothing, crime scene and victim photos, and so on. It also includes scientific trace evidence such as DNA, blood, finger- prints, and fibers that is presented by way of expert testimony describ- ing the collection and testing procedures, as well as their probative value. Some scholars have questioned whether jurors are able to understand scientific evidence and accord it its due weight (Bernstein, 1996; Huber, 1991), whereas others are fearful that jurors may attrib- ute a special aura of credibility to scientific evidence, perceiving it as infallible and thus giving it undue importance in their decisions (Imwinkelreid, 1982-1983; Tribe, 1971).

Skolnick, Shaw / COMPARING EVIDENCE 615

Other studies have empirically examined how juror-eligible partic- ipants actually evaluate probabilistic or statistical evidence. In a review of the literature, Kaye and Koehler (1991) found that mock jurors tend to underweight statistical evidence when other sources of evidence are available. In four of five studies reviewed by Schklar and Diamond (1999), participants underweighted the evidence relative to normative standards such as Bayes’s theorem (Faigman & Baglioni, 1988; Goodman, 1992; Smith, Penrod, Otto, & Park, 1996; Thompson & Schumann, 1987), whereas one lone study found evidence of juror overweighting (Koehler, Chia, & Lindsay, 1995). Schklar and Dia- mond (1999) suggest that these mistakes may be due to logical errors such as misaggregation and misperception or to expectancy-based errors including prior beliefs about the likelihood of laboratory errors and intentional tampering with evidence. For example, it has been suggested that the failure of jurors to return a guilty verdict in the infa- mous O. J. Simpson murder trial may have been due to expectations the predominantly African American jury had about intentional tam- pering with evidence by police or their beliefs regarding the fallibility of scientific evidence per se. It has been also suggested that, given the jurors’ distrust of the physical evidence, the acquittal verdict was fur- ther hastened by the fact that there were no eyewitnesses to the crime (Cowan & Fairchild, 1997).1 In an empirical study, Schklar and Dia- mond (1999) found evidence to support both types of errors they cited, and they report that attempts to correct these errors have been unsuc- cessful to date.

Besides these errors, Wells (1992) reported that mock jurors were reluctant to base liability judgments solely on statistical evidence. Cooper, Bennett, and Sukel (1996) found that mock jurors pay little attention to the quality of complex scientific evidence, preferring instead to rely on the expertise of the source.

The research reviewed herein suggests that jurors overemphasize eyewitness evidence, even though it may be flawed, and underemphasize physical evidence, even though it may be accurate. Although eyewitness and physical evidence have been studied inde- pendently, there have been no studies directly comparing the impact of each type of evidence on jury decision making. Such a comparison would be interesting because both types of evidence could place a defendant at the scene of a crime and therefore would serve the same

616 CRIMINAL JUSTICE AND BEHAVIOR

evidentiary function. However, eyewitness evidence is based on mem- ory, whereas physical evidence is based on tangible objects and is often accompanied by expert opinion. The present investigation reports on two studies attempting to make this comparison. In both studies, eyewitness evidence was based on the memory of an individ- ual who was present at the crime scene, whereas physical evidence was based on tangible objects allegedly used in the crime and physical trace evidence accompanied by expert testimony. It was predicted that both types of evidence would affect jurors’ decisions, such that stron- ger evidence of either type would increase the rate of guilty verdicts. However, based on review of prior studies, it was predicted that eye- witness evidence would be more influential in determining jurors’ decisions than physical evidence when both types of evidence are equally strong.

STUDY 1

METHOD

Participants

Ninety jury-eligible students participated in partial fulfillment of the requirements for their introductory psychology courses at a state university in California. About one third were men and two thirds were women. Mean age was 19.4, with the sample included diverse representation of race, ethnicity, and major.

Materials and Procedure

Case summary. An authentic-looking 10-page case summary, adapted from practice advocacy materials contained in Seckinger (1992), described the fictitious trial of a male defendant accused of the murder of his former wife. The cover page presented the grand jury indictment against the defendant, who was accused of stabbing his ex- wife to death on the front porch of her home. In each of the four ver- sions, the defendant entered a plea of not guilty to a charge of first- degree murder.

Skolnick, Shaw / COMPARING EVIDENCE 617

A preliminary summary of case background followed, which was identical in all conditions and served as a template to introduce the basic facts of the case. It established that the defendant and his ex-wife had a stormy relationship with many failed attempts to reconcile prior to the murder. Depending on condition, the summary stated that the case was based either on eyewitness or physical evidence. This con- sisted of testimony by seven people as well as four exhibits, which included photos of the murder scene and the dead victim.

In the eyewitness evidence cases, the basic facts were as follows: The victim’s stepmother testified that the victim lived with her and had returned home alone from the movies on the night of the murder at about 10 p.m. The stepmother heard noise on the front porch and saw the attack on her stepdaughter, after which the attacker sped away in a white automobile. A neighbor corroborated the stepmother’s testi- mony. The first police officer at the scene testified the victim was lying in a pool of blood on the front porch. He stated that no weapon was found at the crime scene but that the coroner’s report concluded that the victim had been killed by a serrated, 6-inch hunting knife. Based on an interrogation of the victim’s stepmother and her neighbor, the defendant was subsequently arrested. A bartender testified that the defendant had complained about the failure of his marriage and had consumed several drinks at the Silver Dollar Bar before leaving at about 9:15 p.m. on the night of the murder. The defendant testified that he loved his ex-wife and would never do anything to harm her. He con- firmed that he had a few drinks at the Silver Dollar Bar before going home the night his ex-wife was murdered. A neighbor testified that he had seen the defendant arrive home about 9:30 p.m. that night.

Aside from the above basic facts, the eyewitness evidence cases dif- fered as to the following, depending on whether the evidence was strong or weak. In the strong eyewitness condition, the stepmother tes- tified that the porch light was on and a full moon clearly illuminated the attack scene. She was up and alert when she heard noise on the porch. She got a long look at the attacker’s face and identified him with near certainty as her ex-son-in-law. She also positively identified his car. The stepmother’s neighbor, who personally knew the defen- dant, also clearly recognized him when he was standing over the deceased’s body with a bloody knife in his hand. She, too, recognized the defendant’s car. In addition, a doctor from the county coroner’s

618 CRIMINAL JUSTICE AND BEHAVIOR

office testified that after inspecting the body at the crime scene, he found the two eyewitnesses to be calm, emotionally stable, and coher- ent. Subsequent physical examinations found them both to be in good health with excellent eyesight.

In the weak eyewitness condition, the stepmother testified that the porch light was off and the sky was overcast. She was sleeping when she was awakened by noise on the porch. All she could see was a dark, shadowy figure attack her stepdaughter, but based on body build and height, she contended it could have been her ex-son-in-law. She only had a glimpse of the car, but it was of similar color to the one driven by her ex-son-in-law. The neighbor arrived in time to see a vehicle that could have been the defendant’s car speeding away in the darkness. In addition, the doctor from the county coroner’s office testified that the two eyewitnesses were upset, emotionally stressed, and incoherent. The stepmother was groggy from taking sleeping pills and was not wearing her glasses when she saw the scuffle on her front porch.

In the physical evidence cases, the basic facts were as follows: The victim’s stepmother testified that the victim lived with her and had returned home alone from the movies on the night of the murder at about 10 p.m. The stepmother was awakened by noise on the front porch and heard someone attack her stepdaughter in the dark shadows, after which the attacker sped away in a car. Because it was night and overcast, the stepmother was unable to see the assailant’s face nor determine who he was. A neighbor heard a scuffle and arrived in time to see a car speeding away in the darkness. The first police officer at the scene testified the victim was lying in a pool of blood on the front porch. He stated that no weapon was found at the crime scene but that the coroner’s report concluded that the victim had been killed by a ser- rated, 6-inch hunting knife. On the porch, he photographed shoeprints, which the coroner identified as to make and size of shoe. Based on an interrogation of the victim’s stepmother and her neigh- bor, the defendant was subsequently arrested. At his house, a hunting knife and a pair of shoes that matched the descriptions in the coroner’s report were found. A bartender testified that the defendant had com- plained about the failure of his marriage and had consumed several drinks at the Silver Dollar Bar before leaving at about 9:15 p.m. on the night of the murder. The defendant testified that he loved his ex-wife and would never do anything to harm her. He confirmed that he had a

Skolnick, Shaw / COMPARING EVIDENCE 619

few drinks at the Silver Dollar Bar before going home the night his ex- wife was murdered. A neighbor testified that he had seen the defen- dant arrive home about 9:30 p.m. on that night.

Aside from the above basic facts, the physical evidence cases dif- fered as to the following, depending on whether the evidence was strong or weak. In the strong physical evidence condition, a forensic crime expert testified that the shoeprints found at the crime scene, which matched the defendant’s shoes, were from an unusual make and size with only a few such pairs in circulation. Moreover, microscopic examination of the deceased’s entry wounds matched unique shapes and contours on the defendant’s serrated hunting knife. The expert concluded that it was near certain that the shoeprints were from the defendant’s shoes and that the defendant’s knife was the actual mur- der weapon.

In the weak physical evidence condition, a forensic crime expert testified that although the shoeprints found at the crime scene matched the defendant’s shoes, they were from a common make and size with many thousands of such pairs in circulation. Moreover, the deceased’s entry wounds had all been probed during an earlier autopsy, rendering a precise comparison with the defendant’s serrated knife impossible. The expert testified that it could not be reliably concluded that the shoeprints were from the defendant’s shoes nor that the defendant’s knife was the actual murder weapon.

All cases next included closing arguments for the prosecution and the defense, which summarized the earlier testimony. This was fol- lowed by juror instructions, which defined the charge of murder in the first degree, described the reasonable doubt standard of proof, and admonished the mock jurors to consider all of the available evidence.

Dependent measures. Mock jurors were next requested to indicate a not guilty or a guilty verdict on a juror verdict form. Those who found the defendant guilty were asked to indicate which of three sen- tencing options they preferred: Life in prison with possible parole, life in prison without parole, or the death penalty.

A juror opinion form then requested several additional judgments, each on a 9-point scale. Mock jurors rated how responsible the defen- dant was from 1 = not at all responsible to 9 = totally responsible. A second item asked them to rate the overall strength of the evidence

620 CRIMINAL JUSTICE AND BEHAVIOR

against the defendant from 1 = weak to 9 = strong. Mock jurors also rendered a personal evaluation of the defendant by rating him on 10 personality traits, including trustworthiness, reliability, competence, ethics, considerateness, attractiveness, intelligence, warmth, sensitiv- ity, and industriousness. For each trait, higher ratings signified a more positive evaluation. Ratings for the 10 traits were averaged for each mock juror to yield a personal evaluation score.

Finally, demographic attributes of the mock jurors were recorded on a juror background form requesting their gender, age, year in col- lege, university major, and ethnic background. After completing the postquestionnaire, mock jurors were debriefed as to the nature of the experiment, given credit for their participation, and dismissed. The entire procedure was completed by participants in about 30 minutes.

Design

Individuals were randomly assigned to read one of four versions of a hypothetical murder scenario. In two eyewitness evidence condi- tions, eyewitness evidence was strong or weak with no physical evi- dence available. In two physical evidence conditions, physical evi- dence was strong or weak with no eyewitness evidence available. This produced a 2 × 2 (type × strength of evidence) factorial design. Each of the four groups resulting from this procedure contained 22 to 24 participants.

RESULTS

A 2 × 2 analysis of variance (strength × type of evidence) provided a manipulation check on strength of evidence. Mock jurors in the strong evidence conditions rated the evidence significantly stronger than did those in the weak evidence conditions, F(1, 86) = 11.55, p < .001. The mean for the strong evidence condition was 7.4 (SD = 1.8) and for the weak evidence condition, 5.9 (SD = 2.4). Although not statistically reliable, the interaction between strength and type of evidence approached significance, F(1, 86) = 3.36, p = .07. A follow-up analy- sis revealed that the means in the strong eyewitness and physical evi- dence conditions were identical (M = 7.4, SD = 1.6 and 2.1 respec- tively), whereas the means for weak eyewitness (5.1, SD = 2.4) and

Skolnick, Shaw / COMPARING EVIDENCE 621

physical evidence (6.7, SD = 2.3) conditions differed, F(1, 44) = 5.90, p < .02.

A 2 × 2 hierarchical log linear analysis (strength × type of evidence) was conducted on mock juror verdicts. Table 1 presents the data. Overall, mock jurors judged the case to be a strong one: They found the defendant guilty in 68 (75%) cases and not guilty in the remaining 22 (25%) cases. There was a significant main effect for strength of evi- dence, indicating that the defendant was found guilty more often when the evidence was strong (40 cases, 91%) than when it was weak (28, 61% cases), χ2(1, N = 90) = 12.55, p < .001. The main effect for type of evidence was likewise significant, with the defendant being found guilty more often when the evidence was physical (37 cases, 84%) than when it was based upon eyewitness testimony (31 cases, 67%), χ2(1, N = 90) = 4.29, p < .04. The interaction between the two variables was not significant in this analysis.

If mock jurors found that the defendant was guilty, they were then asked to recommend a sentence. Their choices were life with the pos- sibility of parole, life without parole, or death. A second 2 × 2 hierar- chical log linear analysis (strength × type of evidence) was conducted on sentencing for the 68 mock jurors who voted guilty. Although there were no significant main effects or interactions in this analysis, the number of mock jurors preferring the three sentencing options were 22, 32, and 14, respectively.

Mock jurors were asked to rate the defendant on a 9-point scale measuring attribution of responsibility. A 2 × 2 (strength × type of evi- dence) analysis of variance was conducted on this dependent variable. Results revealed a significant main effect for strength of evidence. When the evidence was strong, defendants were perceived as more responsible than when the evidence was weak (M = 7.8, SD = 1.8 ver- sus M = 6.5, SD = 2.7), F(1, 86) = 7.71, p < .007. The main effect for type of evidence and the interaction of strength × type of evidence failed to reach significance.

A final 2 × 2 analysis of variance was conducted on mock jurors’ personal evaluation of the defendant. The only significant effect in this analysis was strength of evidence. When the evidence was strong, mock jurors evaluated the defendant less positively than when the evi- dence was weak (M = 3.6, SD = .90 versus M = 4.1, SD = 1.3), F(1, 86) = 3.94, p < .05. Thus, personal evaluations were dependent on

622 CRIMINAL JUSTICE AND BEHAVIOR

strength of evidence but were unrelated to any other factor. Additional analyses of the demographic information about the mock jurors revealed no significant findings.

DISCUSSION

Consistent with our hypothesis, the data show that both eyewitness and physical evidence are important factors in mock-jury decision making. In spite of the facts that (a) eyewitness testimony is known to be highly influential in forming jurors’ opinions and (b) jurors tend to underrate the importance of physical evidence, our data indicate that mock jurors’ decisions are influenced more by physical evidence than by eyewitness testimony. A direct comparison between the two types of evidence shows a significantly greater number of guilty verdicts were obtained with physical evidence than with eyewitness evidence.

One problem with this unexpected finding is that the manipulation check suggested that strength of evidence was confounded with type of evidence. Specifically, mock jurors rated weak physical evidence as stronger than weak eyewitness evidence. Perhaps, the superiority of physical over eyewitness evidence was due to the fact that mock jurors responded to this difference in perceived evidence strength rather than evidence type. To eliminate this possibility, a second study was con- ducted using the results of pilot-testing drawn from an independent group of participants used to develop protocols in which weak physi- cal and eyewitness evidence and strong physical and eyewitness evi- dence would each be rated identically on strength of evidence.

Skolnick, Shaw / COMPARING EVIDENCE 623

TABLE 1: Mock Juror Verdicts as a Function of Strength and Type of Evidence

Verdict

Type of Evidence Strength of Evidence Guilty Not Guilty

Eyewitness Strong 20 3 Weak 11 12

Subtotal 31 15

Physical Strong 20 1 Weak 17 6

Subtotal 37 7

Total 68 22

Study 1 presented mock jurors with only one type of evidence in each scenario, either physical or eyewitness. Owing to this limitation, the influence of one type of evidence on the other could not be assessed. Moreover, in actual criminal cases, both types of evidence are often available and are used by prosecutors and/or defense attor- neys. To accommodate these issues, Study 2 combines both types of evidence within each scenario. The factors are physical evidence (strong versus weak) and eyewitness testimony (strong versus weak). It was predicted that both types of evidence would affect mock-juror verdicts. In view of the fact that the confound between strength and type of evidence in Study 1 rendered the results inconclusive, no pre- diction was made in Study 2 as to whether eyewitness or physical evi- dence would have a stronger effect on mock jurors’ judgments.

STUDY 2

METHOD

Participants

One hundred twelve jury-eligible students participated in partial fulfillment of the requirements for their introductory psychology courses at a state university in California. None of the students had participated in Study 1. About one third were men, and two thirds were women. Mean age was 19.60 years, and the sample included diverse representation of race, ethnicity, and major.

Design and Procedure

Individuals were randomly assigned to read one of four versions of a hypothetical murder scenario. The case materials, juror instructions, and dependent measures were identical to those used in Study 1 with the following exceptions. First, in Study 1, each scenario contained either eyewitness or physical evidence. Type of evidence (eyewitness versus physical) and strength of evidence (strong versus weak) were varied factorially. In contrast, Study 2 combined both types of evi- dence within each scenario. The factors were eyewitness testimony

624 CRIMINAL JUSTICE AND BEHAVIOR

(strong versus weak) and physical evidence (strong versus weak). This yielded a 2 × 2 (eyewitness × physical evidence) factorial design. Each of the four groups resulting from this procedure contained 28 participants. Second, in Study 1, there was a confound between strength and type of evidence. Specifically, the weak physical evi- dence condition was rated stronger on the strength of evidence scale than the weak eyewitness evidence condition. To eliminate this prob- lem, the evidence in both weak cases was weakened, particularly the physical evidence, in an attempt to equate the two weak evidence cases.

RESULTS

A check on the manipulation of strength of evidence indicated that the manipulation was successful. Mean rating for the case where both types of evidence were strong was 7.9 (SD = 1.9), whereas mean rating for the case where both types of evidence were weak was 4.9 (SD = 2.6). In the other two cases, where one type of evidence was strong and the other was weak, mean ratings were identical, both being 6.9 (the SD was 1.60 when eyewitness evidence was strong and physical evi- dence was weak; the SD was 2.20 when eyewitness evidence was weak and physical evidence was strong). A 2 × 2 analysis of variance on these data produced the expected two main effects (both p < .001) and no interaction.

A 2 × 2 hierarchical log linear analysis (eyewitness × physical evi- dence) was conducted on mock-juror verdicts. Table 2 presents the data. Overall, mock jurors judged the case to be a strong one: They found the defendant guilty in 83 cases (69%) and not guilty in the remaining 37 (31%) cases. There was a significant main effect for eye- witness evidence, indicating that the defendant was found guilty more often when the eyewitness evidence was strong (47 cases, 78%) than when it was weak (36 cases, 60%), χ2(1, N = 120) = 5.44, p < .02. The main effect for physical evidence was likewise significant, with the defendant being found guilty more often when the physical evidence was strong (51 cases, 85%) than when it was weak (32 cases, 53%), χ2(1, N = 120) = 15.29, p < .001. Notably, the number of guilty verdicts when both types of evidence were strong (26 cases, 87%), was not appreciably greater than when eyewitness evidence was strong (21 cases, 70%) or when physical evidence was strong (25 cases, 83%);

Skolnick, Shaw / COMPARING EVIDENCE 625

χ2(2, N = 72) = 0.58, p = n.s. There was no interaction obtained between the two independent variables.

If mock jurors found that the defendant was guilty, they were then asked to recommend a sentence. Choices were life with the possibility of parole, life without parole, or death. A second 2 × 2 hierarchical log linear analysis (eyewitness × physical evidence) was conducted on sentencing for the 83 mock jurors who voted guilty. Although there were no significant main effects or interactions in this analysis, the number of mock jurors preferring the three sentencing options were 16, 45, and 22, respectively.

Mock jurors were asked to rate the defendant on a 9-point scale measuring attribution of responsibility. A 2 × 2 (eyewitness × physical evidence) analysis of variance was conducted on this dependent vari- able. Results revealed significant main effects for both eyewitness and physical evidence. When the evidence of either type was strong, defendants were perceived as more responsible than when the evi- dence was weak; for eyewitness testimony (M = 7.2, SD = 2.5 versus M = 6.2, SD = 2.7), F(1, 116) = 5.01, p < .027, and for physical evi- dence (M = 7.4, SD = 2.5 versus M = 5.9, SD = 2.7), F(1, 116) = 10.09, p < .002). Additional analyses revealed that this effect was stronger for physical evidence (η2 = .08) than for eyewitness evidence (η2 = .04). The interaction between the two variables was not significant in this analysis.

A final 2 × 2 analysis of variance was conducted on mock jurors’ personal evaluation of the defendant. The only significant effect in this analysis was physical evidence. When physical evidence was strong, mock jurors evaluated the defendant less positively than when the evi-

626 CRIMINAL JUSTICE AND BEHAVIOR

TABLE 2: Mock Juror Verdicts as a Function of Eyewitness and Physical Evidence

Verdict

Eyewitness Evidence Physical Evidence Guilty Not Guilty

Strong Strong 26 4 Weak 21 9

Weak Strong 25 5 Weak 11 19

Total 83 37

dence was weak (M = 3.3, SD = 1.4 versus M = 4.0, SD = 1.4), F(1, 116) = 8.00, p < .006. Thus, personal evaluations were dependent on physical evidence but not on eyewitness evidence. Additional analy- ses of the demographic information about the mock jurors revealed no significant findings.

OVERALL DISCUSSION

Consistent with our hypothesis, the data show that both eyewitness and physical evidence are important factors in mock-juror decision making. However, of the two types of evidence, our data indicate that mock-juror decisions are influenced more by physical evidence than by eyewitness testimony. In Study 1, a direct comparison between the two types of evidence shows a significantly greater number of guilty verdicts obtained with physical evidence than with eyewitness evi- dence. Study 2 supported this conclusion by finding more guilty ver- dicts when physical evidence was strong than when eyewitness evi- dence was strong. This effect was obtained even when the confound between strength of evidence and type of evidence that existed in Study 1 was eliminated. Consistent with the conclusion that physical evidence exerts more influence on mock-juror judgments than does eyewitness evidence, a stronger effect size was obtained for physical evidence than for eyewitness evidence on attribution of responsibility. Furthermore, personal evaluations of the defendant were dependent only on physical evidence, not on eyewitness testimony.

Contrary to what has been suggested by previous research, our findings taken together show that mock jurors are influenced more by direct physical evidence than by eyewitness testimony. This finding should provide encouragement to lawmakers and researchers who are aware of the literature suggesting that physical evidence is generally more reliable than eyewitness testimony. Because the physical evi- dence in our studies was presented in direct, simple language, it is unclear whether the present findings apply to those courtroom cases in which physical evidence is presented in complex, mathematical ter- minology. Tribe (1971) suggests that jurors in such instances may be overinfluenced by complex, quantitative evidence to which they ascribe an “aura of precision.” However, more recent evidence dis-

Skolnick, Shaw / COMPARING EVIDENCE 627

putes Tribe’s conclusion and suggests that jurors may not always overvalue statistical evidence (Koehler, 1993).

It is interesting to note that both of these studies show that each type of evidence (eyewitness and physical) affects the kinds of judgments that actual jurors are required to make (verdict decisions) but not the kinds of judgments they normally are not required to make (sentenc- ing preferences). This finding lends considerable ecological validity to the present research. The fact that the same pattern of findings was obtained whether the two types of evidence were presented alone (as in Study 1) or together (as in Study 2) permits added confidence in the generalizability of the results. These facts notwithstanding, it must be pointed out that the present findings are limited by the facts that the mock jurors were college students who were not death-qualified, responding to written trial materials pertaining to only a single case, without the benefit of deliberation. In spite of these restrictions, sev- eral recent studies do suggest that mock-jury research has consider- able applicability to actual jury settings. For example, Bottoms and Goodman (1994) obtained similar findings in sexual assault cases using both written and videotaped stimuli, and Cutler, Penrod, and Dexter, (1990) obtained comparable findings in a study of eyewitness reliability using both college students and experienced jurors.

Both studies found that strong evidence of either type produced more guilty verdicts than weak evidence. Although this result seems self-evident, a comment regarding the strength of evidence manipula- tion is relevant. The strength of eyewitness evidence was manipulated by varying on several dimensions the credibility of eyewitness reports (e.g., the stepmother was alert versus groggy, viewing conditions were favorable versus unfavorable), whereas the strength of physical evi- dence was manipulated by varying its probative value (the probability that the defendant’s shoes and knife were used in the crime). Even though increasing an eyewitness’s credibility and increasing the pro- bative value of evidence both led to more guilty verdicts, it is impor- tant to realize that these are not the same. Subsequent research should be addressed to differentiating more carefully between these two aspects of evidence strength.

One caveat is suggested by our research. Even though jurors may be more influenced by physical than by eyewitness evidence, the latter still carries considerable weight in forming their opinions.2 If either

628 CRIMINAL JUSTICE AND BEHAVIOR

eyewitness or physical evidence is strong by itself, the likelihood of obtaining a guilty verdict is almost as great as if both types of evidence are strong. An implication is that prosecutors need only concentrate on their strongest type of evidence, whereas defense attorneys must be concerned about refuting both.

NOTES

1. Aside from explanations based on jurors’ expectancies and the lack of any eyewitness tes- timony, there is evidence that the acquittal verdict in the Simpson case was due to his celebrity status, as well as same-race leniency by the predominantly African American jury (Brigham & Wasserman, 1999; Skolnick & Shaw, 1997).

2. Although the current study found that mock jurors were more influenced by physical than by eyewitness evidence when they are of equal strength, it is still possible that jurors may over- weight eyewitness evidence and/or underweight physical evidence relative to normative stan- dards such as base rates of eyewitness accuracy or standard statistical criteria.

REFERENCES

Bartol, C. R., & Bartol, A. M. (1994). Psychology and law (2nd ed.). Pacific Grove, CA: Brooks- Cole.

Bernstein, D. E. (1996). Junk science in the United States and in the Commonwealth. Yale Jour- nal of International Law, 21, 123-182.

Bottoms, B. L., & Goodman, G. S. (1994). Perceptions of children’s credibility in sexual assault cases. Journal of Applied Social Psychology, 24, 702-732.

Brigham, J. C., & Bothwell, R. K. (1983). The ability of prospective jurors to estimate the accu- racy of eyewitness identifications. Law and Human Behavior, 7, 19-30.

Brigham, J. C., & Wasserman, A. W. (1999). The impact of race, racial attitude, and gender on reactions to the criminal trial of O. J. Simpson. Journal of Applied Social Psychology, 29(7), 1333-1370.

Buckhout, R. (1974). Eyewitness testimony. Scientific American, 231, 23-31. Cooper, J., Bennett, E. A., & Sukel, H. L. (1996). Complex scientific testimony: How do jurors

make decisions? Law and Human Behavior, 20, 379-394. Cowan, G., & Fairchild. H. (1997). The O. J. Simpson trial: Research and theory on the dynamics

of ethnicity. Journal of Social Issues, 53(3), 409-591. Cutler, B. L., & Penrod, S. D. (1995). Mistaken identification: Eyewitnesses, psychology, and the

law. New York: Cambridge University Press. Cutler, B. L., Penrod, S. D., & Dexter, H. R. (1990). Juror sensitivity to eyewitness identification

evidence. Law and Human Behavior, 14, 185-191. Ellison, K., & Buckhout, R. (1981). Psychology and criminal justice. New York: Harper & Row. Faigman, D., & Baglioni, A. (1988). Bayes’ theorem in the trial process. Law and Human Behav-

ior, 12, 1-17.

Skolnick, Shaw / COMPARING EVIDENCE 629

Goodman, J. (1992). Jurors’ comprehension and assessment of probabilistic evidence. Ameri- can Journal of Trial Advocacy, 16, 361-389.

Hall, D. F., Loftus, E. F., & Tousignant, J. P. (1984). Postevent information and changes in recol- lection for a natural event. In G. L. Wells & E. F. Loftus (Eds.), Eyewitness testimony: Psy- chological perspectives (pp. 124-141). New York: Cambridge University Press.

Huber, P. W. (1991). Galileo’s revenge: Junk science in the courtroom. New York: Basic Books. Imwinkelreid, E. (1982-1983). The standard for admitting scientific evidence: A critique from

the perspective of juror psychology. Villanova Law Review, 28, 554-571. Kaye, D. H., & Koehler, J. J. (1991). Can jurors understand probabilistic evidence? Journal of

the Royal Statistical Society Series A, 154, 75-81. Koehler, J. J. (1993). The normative status of base rates at trial. In N. J. Castellan, Jr. (Ed.), Indi-

vidual and group decision making: Current issues (pp. 137-149). Hillsdale, NJ: Lawrence Erlbaum.

Koehler, J. J., Chia, A., & Lindsay, S. (1995). The random match probability in DNA evidence: Irrelevant and prejudicial? Jurimetrics Journal, 35, 33-58.

Kramer, T. H., Buckhout, R., & Eugenio, P. (1990). Weapon focus, arousal, and eyewitness memory: Attention must be paid. Law and Human Behavior, 14, 167-184.

Loftus, E. F. (1974). Reconstructing memory: The incredible eyewitness. Psychology Today, 8, 116-119.

Loftus, E. F. (1979). Eyewitness testimony. Cambridge, MA: Harvard University Press. Loftus, E. F., Loftus, G. R., & Messo, J. (1987). Some facts about “weapon focus.” Law and

Human Behavior, 11, 55-62. Mitchell, K. J., Livosky, M., & Mather, M. (1998). The weapon focus effect revisited: The role of

novelty. Legal and Criminological Psychology, 3, 287-303. Pickel, K. L. (1999). The influence of context on the “weapon focus” effect. Law and Human

Behavior, 23(3), 299-311. Schklar, J., & Diamond, S. S. (1999). Juror reactions to DNA evidence: Errors and expectancies.

Law and Human Behavior, 23(2), 159-184. Seckinger, J. H. (1992). State v. O’Neill, problems and case file (5th ed.). Notre Dame, IN:

National Institute for Trial Advocacy. Shaw, J. I., & Skolnick, P. (1994). Sex differences, weapon focus, and eyewitness reliability.

Journal of Social Psychology, 134(4), 413-420. Shaw, J. I., & Skolnick, P. (1999). Weapon focus and sex differences in eyewitness identification:

Arousal vs salience. Journal of Applied Social Psychology, 29, 2328-2341. Skolnick P., & Shaw, J. I. (1997). The O. J. Simpson criminal verdict: Racism or status shield?

Journal of Social Issues, 53(3), 503-516. Smith, B. C., Penrod, S. D., Otto, A. L., & Park, R. C. (1996). Jurors’ use of probabilistic evi-

dence. Law and Human Behavior, 20, 49-82. Thompson, W. C., & Schumann, E. (1987). Interpretation of statistical evidence in criminal tri-

als. Law and Human Behavior, 11, 167-187. Tribe, L. (1971). Trial by mathematics: Precision and ritual in the legal process. Harvard Law

Review, 84, 1329-1376. Wells, G. L. (1992). Naked statistical evidence of liability: Is subjective probability enough?

Journal of Personality and Social Psychology, 62, 739-752. Wells, G. L., Lindsay, R.C.L., & Ferguson, T. (1979). Accuracy, confidence, and juror percep-

tions in eyewitness identification. Journal of Applied Psychology, 64, 440-448. Wrightsman, L. S., Nietzel, M. T., & Fortune, W. H. (1998). Psychology and the legal system (4th

ed.). Monterey, CA: Brooks-Cole.

630 CRIMINAL JUSTICE AND BEHAVIOR