Deliberating on dangerousness and death: jurors' ability to differentiate between expert actuarial and clinical predictions of dangerousness
Introduction
Because a criminal defendant's future dangerousness is an essential aspect of the capital sentencing jury's decision to impose the death penalty in one state in the United States, Texas,1 and is a significant legal consideration in a number of other U.S. states2 (Worrell, 1987), mental health professionals have played a central role in these life or death decisions. Their expertise to proffer such testimony has been uniformly accepted by American courts (Barefoot v. Estelle, 463 U.S. 880, 1983),3 and their testimony has been found to be substantially relied on by jurors in reaching final decisions Marquart et al., 1989, Otto, 1992. Although United States' reliance on capital punishment and the role mental health expert testimony play in these decisions has been soundly criticized by the international community (e.g., Constanzo, 1997), international displeasure appears to have generated little effect of frequency of executions in the states, especially in Texas. Additionally, a substantial body of research suggests that expert predictions of future dangerousness, when based solely on the testifying expert's clinical experience, demonstrate an unimpressive ability to accurately forecast the long-term future behavior of criminal defendants (e.g., Monahan, 1981, Monahan, 1984, Monahan & Steadman, 1994, Otto, 1992).
The inability of mental health professionals to accurately predict future dangerousness has led many within and outside forensic psychology to argue that mental health professionals have no expertise to offer the court in these cases (Faust & Ziskin, 1988), and has led others to begin the process of creating more scientifically sound dangerousness prediction instruments. These actuarial instruments statistically weigh and combine verified risk factors (e.g., psychopathic tendencies, prior arrest history, violence history) in a manner that maximizes their power to predict the specified outcome (recidivism or violent recidivism; Monahan & Steadman, 1994). A large body of research has demonstrated that actuarial instruments outperform clinical predictions in a variety of different areas (see Grove & Meehl, 1996 for a review), and these research findings have been echoed in the future dangerousness arena.4 In the only meta-analysis performed on this topic, Mossman (1994) reanalyzed findings of 58 data sets from 44 published studies and determined that although clinical opinion dangerousness predictions performed at above chance levels,5 actuarial instruments significantly outperformed them for long-term predictions of dangerousness.6
Prior to the development of such actuarial instruments, however, the United States Supreme Court, in Barefoot v. Estelle (463 U.S. 880, 1983), adjudicated the constitutional admissibility of expert clinical opinion predictions of future dangerousness. At issue were two clinical opinion predictions offered by psychiatrists based on the experts' years of experience and in response to a hypothetical question. The Court was unpersuaded by an early summary of empirical research detailing extreme clinical inaccuracy in this area, even though the error rates associated with clinical opinion predictions suggested that the clinicians were incorrect 65–85% of the time (Amicus Curiae Brief for the American Psychiatric Association in Barefoot v. Estelle, 1980; citing Monahan, 1981), and did not overturn the decisions admitting the testimony. The Court held that jurors are capable with the aid of adversary procedures (i.e., cross-examination and competing experts) “… of … separating the wheat from the chaff” (Barefoot, p. 3398) and of appropriately weighing expert testimony on future dangerousness in capital sentencing decisions. The majority reached this decision even though one of the experts in the Barefoot case omnisciently stated that there was “100% chance” the defendant would be a future danger.
Based on existing jury decision-making research, and in light of a recent study (Krauss & Sales, 2001) that directly examined the ability of jurors to weigh clinical opinion expert testimony in the capital sentencing context, the United States' highest court's assumptions in this landmark case appear suspect.
Naturalistic experiments (e.g., Shuman et al., 1994, Sundby, 1997) as well as laboratory-based designs (e.g., Brekke & Borgida, 1988) have demonstrated that jurors' final decisions are strongly influenced by expert psychological testimony when it is presented on a wide variety of psycho-legal issues including: the inaccuracy of eyewitness identification (e.g., Cutler et al., 1989, Loftus, 1986, Wells, 1986), clinical syndromes (e.g., Brekke et al., 1991, Kovera et al., 1997, Schuller & Vidmar, 1992), insanity (e.g., Greenberg & Wursten, 1988, Rogers et al., 1992, Rogers et al., 1990), and future dangerousness (Krauss & Sales, 2001). Although jury decisions are clearly affected by such testimony, the manner in which expert testimony influences jury decisions is poorly understood. Especially troubling is the research that suggests that jurors may not be attending to the content of the expert testimony, but rather are influenced by the mere presence of an expert or his credentials Cooper et al., 1996, Cooper & Neuhaus, 2000, Greenberg & Wursten, 1988. For example, Greenberg and Wursten demonstrated in a mock jury insanity case that jurors' decisions were more influenced by expert testimony offered by a medical expert than a psychological expert even though the testimony offered by both was identical. Yet, other research examining the effects of expert testimony on jurors' beliefs concerning the fallibility of eyewitness identification has produced more equivocal results. In some experimental situations, jurors have been found to be capable of attending appropriately to the content of expert testimony proffered, and such testimony was found to be effective in eliminating pre-existing biases (i.e., the bias that confidence does not equal accuracy in eyewitness identifications) (Cutler et al., 1989), while in other experimental situations such testimony has not been shown to eliminate these biases (Cutler, Penrod, & Stuve, 1988).
Consistent with other jury decision-making research, the one simulation study (Krauss & Sales, 2001), which directly addressed the relationship between expert psychological dangerousness testimony and juror decision-making in capital sentencing, found that mock jurors' final decisions were significantly influenced by expert testimony. Both clinical and actuarial expert testimony7 affected jurors' decisional ratings made immediately following the presentation of the expert testimony. Jurors, however, were more influenced by clinical predictions of future dangerousness than they were by actuarial predictions, and this preference for clinical expert testimony remained even after cross-examination and competing expert testimony manipulations. In fact, after the adversary manipulations, the jurors' confidence in imposing the death penalty in the actuarial expert testimony conditions returned to levels not significantly different from those held prior to the expert testimony while jurors' confidence ratings in the clinical opinion expert condition remained significantly above their pre-expert levels. This continued divergence suggests that adversary manipulations may be more effective in moderating the influence of actuarial expert testimony on jurors' decisions than in moderating the impact of clinical expert testimony. These findings indicate both that the United States Supreme Court's pronouncements about jury behavior in Barefoot may have been incorrect, and that courts must be especially vigilant about the influence of less scientific expert testimony on juror decisions.8
Krauss and Sales (2001) offered two competing explanations for jurors' preference for clinical opinion expert testimony over actuarial expert testimony. They first suggested that jurors may be overwhelmed by actuarial expert testimony because of its complexity and that jurors may lack the means even after adversary procedures to effectively evaluate such statistical information (actuarial complexity hypothesis). Alternatively, they proposed that jurors may hold a pro-clinical bias which causes them to weigh clinical testimony more heavily because there is an intuitive appeal to pronouncements made by a human expert rather than an expert relying on statistics (pro-clinical bias hypothesis).9 Neither hypothesis could be adequately verified based on the results of their study, however.
The present study builds upon and examines an element missing from the Krauss and Sales' (2001) study—juror deliberation. The importance of deliberation conditions in mock jury experiments is becoming well established. Although early empirical work suggested that deliberation did little to change jurors' decisions (Kalven & Zeisel, 1966) and a recent review found few significant effects attributable to deliberations (Bornstein, 1999), some empirical work has demonstrated significant pre- and post-deliberation differences. These effects have occurred when complicated dispositional or legal instructions are presented at trial. In each instance, such information has been found to significantly affect juror decisions post-deliberation even though no effect was shown pre-deliberation Diamond & Levi, 1996, Kerwin & Shaffer, 1994, Wheatmann & Shaffer, 2001. Further, in other studies when different types of expert testimony are proffered at trial, adversarial experts were found to be more influential before deliberation than court appointed experts, this effect disappeared post-deliberation (Brekke et al., 1991). In her review of the jury decision-making literature citing these empirical findings a well as others, Diamond (1997) concluded that one of the five fundamental flaws in past and present jury research was the lack of deliberation conditions in simulation experiments.
There are a number of reasons why deliberation conditions may be especially important in understanding jury decision-making beyond simple ecological validity.10 Wheatmann and Shaffer (2001) argue that mock jurors may be more likely to carefully attend to information presented when they will have to justify their opinions to peers (emphasis added). Deliberation conditions also may allow jurors to correct misunderstandings that other members hold, and may allow difficult or complicated information to be better understood when multiple individuals are available to explain it (Diamond & Levi, 1996). Furthermore, Cooper et al. (1996) have conjectured that mock juries fail to attend to the content of testimony presented in juror simulations because they lack the motivation, ability, or time to accurately comprehend the content of the testimony Cooper et al., 1996, Cooper & Neuhaus, 2000.11 Deliberation conditions may address all these concerns by offering a forum for justification of viewpoints, multiple and divergent comprehension abilities, and additional time.
If there are empirical, theoretical, and policy reasons for including deliberation in juror decision-making experiments in general, it is doubly important that deliberation be added to investigations where the impact of expert actuarial dangerousness testimony is to be assessed because such testimony is clearly difficult for laypersons to sufficiently understand and evaluate (Melton, Petrila, Poythress, & Slobogin, 1997). Consistent with the Supreme Court's assumptions in Barefoot and Krauss and Sales' actuarial complexity hypothesis, the addition of deliberation conditions may eliminate the juror preference for clinical expert testimony and support the notion that jurors, when given adequate time and motivation, are able to uncover and take due account of the shortcoming of clinical expert testimony while at the same time recognizing the strengths of actuarial expert testimony.12 Alternatively, a deliberation condition may have no significant effect on juror decisions, supporting the notion that the complexity of the expert actuarial testimony is not the cause of the testimony's limited influence on jurors, but rather that clinical opinion testimony may be more intuitively appealing and receive more weight in the jurors' decisions.
The present study seeks to: (a) replicate the findings of Krauss and Sales (2001) with regard to the effect that different types of expert testimony on future dangerousness exert on mock juror capital sentencing decisions; (b) expand research in this area by examining what role deliberation conditions play in these decisions; and (c) attempt to explain why clinical opinion expert testimony might be preferred over actuarial expert testimony by jurors. To accomplish these goals, the simulated capital sentencing case used in Krauss and Sales was presented to mock jurors using both written and videotaped materials. The influence of expert testimony on mock jurors was assessed using a number of dependent measures (i.e., jurors' ratings of the defendant's dangerousness and jurors' ratings of the expert's influence, level of science, credibility, and confidence) across four different time periods (i.e., after reading a written summary of the case, after the expert testimony, after the cross examination, and after deliberation).13
Section snippets
Subjects
Participants were 114 undergraduates from the Claremont Colleges, 54 women (47%) and 60 men (53%), each of whom participated as an individual juror and as part of 6–14 person jury. The participants described themselves as originating from a variety of different backgrounds, 69% described themselves as Caucasian, 8% as Hispanic, 8% as Asian, 5% as African–American, and 8% did not list an ethnic background. Participants ranged in age from 17 to 27 (M=19.75, S.D.=1.69). Prior to the administration
Juror's overall evaluations of the defendant's dangerousness
An omnibus significance test based on a within and between subjects repeated measures ANCOVA was conducted on the mock jurors' dangerousness confidence ratings across the various time periods (after expert testimony, time 2; after cross-examination, time 3; and after deliberation, time 4) with time as the within subjects variable, and type of expert testimony (actuarial or clinical) and type of cross-examination (ineffective or effective) as factorial variables to determine the overall
Discussion
As predicted and consistent with previous research on expert testimony in this specific context (Krauss & Sales, 2001), this study found a significant effect for expert testimony, and a differential effect for clinical opinion versus actuarial testimony on mock jurors' ratings of a criminal defendant's future dangerousness. Although influenced by both types of testimony, mock jurors were significantly more influenced by clinical opinion expert testimony in their confidence ratings of the
Limitations and conclusions
Like most mock jury simulation experiments this experiment has a number of flaws including, but not limited to: (a) lack of generalizability of results based on a single case, (b) a nonrepresentative jury pool, (c) shortening of procedures, conditions and especially deliberations, (d) the absence of combined cross-examination and competing expert conditions, and (e) perhaps most importantly, verdicts and final decisions that have limited meaning outside the simulation (i.e., participants know
References (54)
- et al.
Psychiatrists and the parameter of expert testimony
International Journal of Law and Psychiatry
(1992) - et al.
Effects of ultimate opinions on juror perceptions of insanity
International Journal of Law and Psychiatry
(1990) - Barefoot v. Estelle, No. 82-6080, 463 U.S. 880...
The ecological validity of jury simulations: is the jury still out?
Law and Human Behavior
(1999)- et al.
Expert psychological testimony in rape trials: a social-cognitive analysis
Journal of Personality and Social Psychology
(1988) - et al.
Of juries and court appointed experts: the impact of nonadversarial versus adversarial expert testimony
Law and Human Behavior
(1991) - Brief for the American Psychiatric Association as Amicus Curiae in Support of Petitioner, in Barefoot v. Estelle, No....
The heuristic model of persuasion
Just revenge: costs and consequences of the death penalty
(1997)- et al.
Complex scientific testimony: how do jurors make decisions?
Law and Human Behavior
(1996)
The hired gun effect: assessing the effect of pay, frequency of testifying, and credentials on the perception of the expert
Law and Human Behavior
The eyewitness, the expert psychologist, and the jury
Law and Human Behavior
Juror decision making in eyewitness identification cases
Law and Human Behavior
Illuminations and shadow from jury simulations
Law and Human Behavior
Improving decisions on death by revising and testing jury instructions
Judicature
The expert witness in psychology and psychiatry
Science
Clinical versus actuarial predictions of violence in patients with mental illness
Journal of Consulting and Clinical Psychology
The psychologist and psychiatrist as expert witnesses: perceived credibility and influence
Professional Psychology: Research and Practice
Comparative efficiency if informal (subjective, impressionistic) and formal (mechanical, algorithmic) prediction procedures: the clinical–statistical controversy
Psychology, Public Policy, and Law
Prospective replication of the violence risk appraisal guide in predicting violent recidivism among forensic patients
Law and Human Behavior
Effects of involvement on persuasion: a meta-analysis
Psychological Bulletin
The American jury
Mock juror versus mock juries: the role of deliberations in reactions to inadmissible testimony
Personality and Social Psychology Bulletin
Does expert psychological testimony inform or influence juror decision making? A social cognitive analysis
Journal of Applied Psychology
The effects of clinical and scientific expert testimony on juror decision-making in capital sentencing
Psychology, Public Policy, and Law
Cited by (43)
Effects of the "intimate conviction" instruction on the processing of judicial information
2013, Revue europeenne de psychologie appliqueeCitation Excerpt :This automatic processing may constitute just an initial phase, followed by a more analytical, rational processing procedure whereby first impressions can be corrected (Epstein, 1994). CEST has formed the backbone of many studies of how judgments are reached (Borum, 1996; Gardner, Lidz, Mulvey, & Shaw, 1996; Krauss & Lee, 2003; Krauss & Sales, 2001; Mossman, 1994; Quinsey, Harris, Rice, & Cormier, 1998). Some of these have shown that “mock jurors”1 spontaneously exhibit greater sensitivity to testimony of a clinical nature than to information that is more objectively based (e.g., statistics on offender dangerousness, the risk of repeat offending).
A confidence-credibility model of expert witness persuasion: Mediating effects and implications for trial consultation
2011, Consulting Psychology JournalThe bias in judgement: when “naïve” knowledge challenges expert knowledge in criminal trials
2022, Journal of Forensic PracticeWhen punishment goals moderate and mediate the effect of clinical reports on the recidivism risk on prison sentences
2021, Psychiatry, Psychology and Law