bs_bs_banner

Journal of Empirical Legal Studies

Volume 11, Issue 1, 105131, March 2014

The Impact of Neuroimages in the

Sentencing Phase of Capital Trials
Michael J. Saks, N. J. Schweitzer, Eyal Aharoni, and Kent A. Kiehl*

Although recent research has found that neurological expert testimony is more persuasive
than other kinds of expert and nonexpert evidence, no impact has been found for
neuroimages beyond that of neurological evidence sans images. Those findings hold true in
the context of a mens rea defense and various forms of insanity defenses. The present studies
test whether neuroimages afford heightened impact in the penalty phase of capital murder
trials. Two mock jury experiments (n = 825 and n = 882) were conducted online using
nationally representative samples of persons who were jury eligible and death qualified.
Participants were randomly assigned to experimental conditions varying the defendants
diagnosis (psychopathy, schizophrenia, normal), type of expert evidence supporting the
diagnosis (clinical, genetic, neurological sans images, neurological with images), evidence of
future dangerousness (high, low), and whether the proponent of the expert evidence was
the prosecution (arguing aggravation) or the defense (arguing mitigation). For defendants
diagnosed as psychopathic, neuroimages reduced judgments of responsibility and sentences
of death. For defendants diagnosed as schizophrenic, neuroimages increased judgments of
responsibility; nonimage neurological evidence decreased death sentences and judgments
of responsibility and dangerousness. All else equal, psychopaths were more likely to be
sentenced to death than schizophrenics. When experts opined that the defendant was
dangerous, sentences of death increased. A backfire effect was found such that the offering
party produced the opposite result than that being argued for when the expert evidence was
clinical, genetic, or nonimage neurological, but when the expert evidence included
neuroimages, jurors moved in the direction argued by counsel.

I. Introduction
A. Expectations of Inordinate Effects of Neuroimages
In criminal trials, brain images are among the most recent forms of evidence presented for
the purpose of supporting factual arguments about the defendants mental condition. That
images would enhance persuasion should not be surprising when we consider the sizeable

*Address correspondence to Michael Saks, Sandra Day OConnor College of Law, Arizona State University, Box
877906, Tempe, AZ 85287-7906, email: saks@asu.edu. Saks is Regents Professor in the College of Law and Department
of Psychology; Schweitzer is Assistant Professor of Social and Behavioral Sciences, New College of Interdisciplinary
Arts and Sciences, Arizona State University; Aharoni was Postdoctoral Fellow at the University of New Mexico and is
now Associate Behavioral Scientist, RAND Corporation; Kiehl is Associate Professor of Psychology and Neuroscience,
University of New Mexico.
This research was supported by a grant from the John D. and Catherine T. MacArthur Foundation to the Law and
Neuroscience Project.

105

106

Saks et al.

body of research finding that images enhance learning compared to textual-only communication in educational settings (Carney & Levin 2002; Mayer 2009).
Numerous commentators have expressed concern that these images are inordinately
persuasivefar more so than the life history of the defendant, or the interview-based
diagnoses of a clinical psychologist or psychiatrist, or the tests of a neuropsychologist, or
even a radiologists verbal description of the findings of an MRI or fMRI. The notion that
has been advanced is that seeing the image of a brain with their own eyes leads judges and
jurors to believe something that they will be unable to deny: that the suspects brain appears
different, that the observed difference is beyond the control of the defendant, andat the
end of this line of reasoning or intuitionthat the extreme behavior that brought the
defendant to court is not something for which he or she can be held responsible. In this
view, if the brain is the source of behavior, then a defective brain must be the cause of the
deviant behavior. The image, and only the image, it has been suggested, can lead judges
and jurors to this unshakable conclusion (e.g., Batts 2009; Brown & Murphy 2010; Compton
2010; Dumit 1999; Gurley & Marcus 2008; Khoshbin & Khoshbin 2007; Kulynych 1997; Pratt
2005; Rose 2000; Roskies 2007, 2008).
If brain images have far more influence on factfinders than is justified by the
information those images convey, then the courts will be asked to exclude such images.
Rule 403 of the Federal Rules of Evidence, and similar or identical rules in all states,
provides that otherwise relevant evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury . . . . Courts prefer admission of relevant evidence, so exclusion on these
grounds is unusual, but if neuroimages were found to routinely win the acquiescence of
jurors despite adding little or nothing to the information provided by a witness using other
means to explain the findings of an examination of a defendant or, indeed, if the
neuroimages somehow turned the tide in circumstances where evidence in other forms was
not terribly persuasive, then the images would be serious candidates for exclusion.
The neurosciences are revealing growing evidence for neural markers of poor
behavioral control (Aharoni et al. in press) and biomarkers for mental illnesses like schizophrenia (Du et al. 2012) and psychopathy (Ermer et al. 2012, 2013), but it is not the case
that just any neural cause of antisocial behavior is sufficient to legally excuse that behavior.
The ongoing challenge for neuroscientists and lawyers is to reliably distinguish legally
excusable neural causes from inexcusable ones, and to present this information in an
unbiased manner. To date, there is no clear consensus on what criteria ought to be used in
the performance of such tasks.
Moreover, brain images created from techniques such as functional MRI (fMRI) are
the product of a complex set of choices and manipulations: the scanner, its programming,
its measures, the tasks selected for testing patients and subjects, the characteristics of the
sampled participants, the analyses carried out on the resulting time-series data, and the
soundness of the statistics used in the analysis. Interpretation of what constitute abnormal
patterns might require comparison to a normative database. Unfortunately, such normative
databases are rare. Further, many brain functions, and certainly brain structure, change
over time, so a test today may not necessarily tell us anything about acute brain function
(i.e., a psychotic event) at the time of the conduct at issue in a criminal case. In short, no

The Impact of Neuroimages on Sentencing

107

brain image speaks for itself (e.g., Cacioppo et al. 2003; Gazzaniga 2008; Logothetis 2008;
Poldrack et al. 2008; Racine et al. 2005; Society of Nuclear Medicine Brain Imaging Council
1996; Trout 2008; Uttal 2003, 2008; Vul et al. 2009; Vul & Kanwisher 2010).
B. Empirical Evidence Concerning Neuroimage Effects
Empirical support for the possibility that brain images will exert an inordinate impact on
legal decisionmakers is suggested by a variety of studies touching on the cognitive power of
neurological evidence. For example, people give more credence to explanations of psychological phenomena when those explanations include a supposed neurological basis, even
when the explanation itself is weak and illogical. Inclusion of neuroscientific language
appeared to cloud peoples reasoning (Weisberg et al. 2008). Newspaper-style articles that
describe psychological research findings are more persuasive when they include brain
images than when they are accompanied by a bar graph or topographical map of brain
activity (McCabe & Castel 2008). In the context of mock murder trials involving a defense
of insanity as well as a diagnosis of either schizophrenia or psychopathy, jurors were 1.3
times more likely to render a verdict of guilty by reason of insanity when the evidence
included expert testimony that was neuroscience based (an image plus an experts testimony) (Gurley & Marcus 2008). When mock jurors were presented with a criminal trial
scenario that included testimony offered by the prosecution reporting the results of fMRI
lie detection evidence (as well as control conditions presenting other lie detection methods
(polygraph, thermal facial imaging), or no lie detection testimony, significantly more guilty
verdicts were produced by the fMRI lie detection than by the other conditions (McCabe
et al. 2011).
As Schweitzer et al. (2011) have argued, those studies must be regarded as offering
no more than hints about the specific question of whether neuroimages have an inordinate
impact in the setting of criminal trials. The Weisberg et al. (2008) research involved no
images, only verbal explanations. The same is true of McCabe et al. (2011). McCabe and
Castels (2008) study did involve images, but in the tamer context of evaluating newspaper
articlesthat is, where participants are not asked to consider the consequences of yielding
to the suggestions of neuroscientific findings about criminal defendants who have committed harmful acts.1 The Gurley and Marcus (2008) study is placed in a relevant context, but
in their study neuroimages are confounded with other neuroscience information. If what
persuaded their mock jurors was the entire package of neuroevidence, then ascertaining
the role of the neuroimage component of the evidence requires research designs tailored to
separate the effects of images from the effects of other evidence. Without such a design, one
cannot draw inferences about the distinctive effects (if any) of the images.
Schweitzer and colleagues have been conducting a program of research aimed at
testing the effects of neuroimages in various facets of criminal proceedings. The first series

Note that criminal convictions require not only proof of proscribed conduct, but also a culpable mental state (e.g.,
specific intent) and the absence of a persuasive exculpatory defense (e.g., insanity). It is those psychological states and
conditions that are at issue when neuroscience evidence is proffered on behalf of a defendantnot the question of
whether the defendant performed the prohibited act.

108

Saks et al.

of experiments examined the influence of neuroimagery when neuroevidence was offered

in support of a mens rea defense.2 Four experiments were conducted employing nationally
representative samples of mock jurors, crimes of differing severity, and a series of control
conditions consisting of a testifying neurologist presenting a neuroimage with a series of
control conditions, a testifying neurologist with a graph (instead of a brain image), with no
visual image to accompany his testimony, a clinical neuropsychologist (instead of a neurologist) who based the diagnosis on nonneuroimage tests and examinations, a clinical
psychologist who based the diagnosis on a psychological test, and the omission of all expert
evidence (leaving only the familys testimony to support the claim of a brain disorder)
(Schweitzer et al. 2011). In all, no evidence was found suggesting that neuroimages themselves in any way influenced the jurors verdicts, sentence recommendations, or other
judgments about the defendant over and above the experts verbal testimony. The variations containing neuroscience-based evidence were equally effective at persuading jurors
of the defendants reduced capacity (and influencing their verdicts). Moreover, neuroscience evidence was more effective than either the clinical psychological testimony or the
family history testimony.
In a second set of experiments, Schweitzer and Saks (2011) turned to the more
familiar defense of insanity. Again using a design that permitted evaluation of the distinctive contribution of neuroimage evidence offered as evidence of insanity (experimentally
varying four different legal definitions of insanity), and using a large national sample of
mock jurors, no evidence of any independent influence of neuroimagery was found.
Overall, neuroscience-based evidence was found to be more persuasive than psychological
and family history evidence. Effects were consistent across different insanity standards.
The present study examines the possible impact of neuroimages in the penalty phase
of capital murder trials.
A 2010 Florida trial provides an anecdotal illustration of the possible influence of
neuroimages in penalty-phase hearings (Miller 2010). Grady Nelson was convicted of
murdering his wife by stabbing her 61 times. His history included having previously sexually
assaulted his wifes two daughters and having been convicted of raping a seven-year-old
neighbor. As part of the penalty-phase hearing, the defense presented an expert on
quantitative EEG brain mapping (QEEG), offered to prove brain injury, which was argued
to be responsible for an involuntary predisposition to impulsiveness and violence. The jury
sentenced Nelson to lifetime imprisonment rather than death. At least one juror later
explained the decision as resulting from the neuroscience and accompanying neuroimage
evidence.
One previous jury simulation experiment was conducted to test the impact of
neuropsychological evidence, including brain images, on sentencing recommendations in
capital penalty-phase hearings (Greene & Cahill 2011). Diagnostic evidence of psychosis
was presented, supported by neuropsychological test results and neuroimages. Further,
prosecution evidence of future dangerousness was manipulated, indicating either a high or

That is, a defense in which it is argued that the defendant did not form (because he was mentally incapable of
forming) the intent required to be guilty of the crime charged.

The Impact of Neuroimages on Sentencing

109

low risk. Neuropsychological evidence reduced the frequency of death sentence recommendations compared to the diagnosis-only condition. However, the addition of
neuroimage evidence to the neuropsychological testimony led to no additional reduction
in recommendations for death.
C. Penalty Phase of Capital Trials
Following conviction of a capital crime, the prosecution must first prove the existence of at
least one aggravating circumstance from a specified list of factors (e.g., killing multiple
victims or a child or a police officer, or other factors, notably dangerousness) in order to
warrant consideration of the death penalty. If all the preceding conditions are satisfied, the
trial moves to a penalty phase, where evidence relevant to mitigation as well as aggravation
is presented to the factfinder.
Although proof of aggravating facts is circumscribed by the rules of evidence, defense
proffers of mitigating evidence are admitted with fewer constraintsin part because what
is relevant to persuading the jury to grant leniency is a very broad range of subjects and
sources, and in part because the right of a party charged with a serious crime to put forward
evidence on its own behalf is constitutionally protected. Thus, in contrast to the guilt phase
of the trial, in the penalty phase of a capital case, the scientific quality of neuroscience
evidence and generated images would have lower hurdles to surmount in order to be
presented to the jury.
In weighing and balancing the aggravating and mitigating evidence in order to
decide whether to impose a sentence of death or life in prison, jurors are instructed, among
other details, to each make their own decisions about whether the total of the mitigation
is sufficiently substantial to call for leniency. Sufficiently substantial to call for leniency
means that mitigation must be of such quality or value that it is adequate, in the opinion of
you as an individual juror, to be persuaded to vote for a sentence of life in prison.

II. Psychobiological Evidence Relevant to Mitigation

Many of the evidentiary inquiries about a criminal defendants mental-health status aim to
address some aspect of the defendants past or future ability to conform his/her conduct
to the requirements of the law. Currently, no psychological or biological test is sufficient to
directly establish that an ostensible criminal act was or will be caused by factors outside the
defendants control. However, (1) psychological evaluations have been used to indirectly
suggest the presence of involuntary behavioral disinhibition, evidenced by a relevant pathological diagnosis, and (2) biological evidence has been used to support such clinical
diagnoses. In this section, we describe psychological and biological evidence relevant to two
conditions commonly implicated in behavioral disinhibition: psychopathy and schizophrenia. Both will be represented in the experiments reported in this article.
A. Psychopathy
Psychopathy is a personality disorder characterized by a lack of empathy and remorse,
impulsive and irresponsible conduct, manipulation, and a history of antisocial, often

110

Saks et al.

criminal, behavior (Hare 2003). Although less than 1 percent of the general population
meets full criteria for a psychopathy diagnosis (e.g., Coid et al. 2009), the prevalence of
psychopathy in correctional settings may exceed 20 percent (Hare et al. 1991). The most
well-validated method of assessing psychopathy is the Hare Psychopathy ChecklistRevised
(PCL-R), a multihour archival review and interview (Hare 2003). Because psychopath is
a pejorative term connoting a life-course-persistent condition (see APA Brief 2006; Viljoen
et al. 2010), the diagnosis of psychopathy is typically reserved for adults age 18 or older for
whom clinicians have sufficient historical and other, collateral information to make a valid
diagnosis. However, developmental studies have successfully identified behavioral and
biological risk factors for callous/unemotional (i.e., psychopathic) symptomatology at ages
as young as six (Burke et al. 2007; Frick et al. 2003; Lynam et al. 2007).
1. Scientific Evidence of Affective Problems
Many behavioral studies suggest that affective deficits are a central feature of psychopathy.
For example, Blair and colleagues found that psychopathic offenders were less accurate
than nonpsychopathic offenders at recognizing fearful facial displays (Blair et al. 2004c).
Similar results were obtained using fearful vocal recordings (Blair et al. 2002). Theorists
have argued that one reason healthy adults are often able to successfully inhibit violent
behavior is that this inhibitory response takes input from affective and physiological systems
that detect distress and prepare the appropriate motor plan (Blair 1995). Thus, the limited
ability of psychopathic individuals to recognize and physiologically respond to fearful and
unpleasant stimuli could serve to deprive them of this important regulatory information
and enable them to persist in harmful behavior.
Although biological evidence is not independently necessary or sufficient to diagnose
an individual with psychopathy, biological studies lend support to the behavioral evidence
of an affective deficit underlying psychopathy. For example, physiological studies have
shown that high psychopathy individuals exhibit a reduced startle response to emotionally
unpleasant images (Patrick et al. 1993) and a reduced electrodermal response to distress
cues (Blair 1997).
Functional neuroimaging studies have shown the amygdalaa brain region strongly
associated with emotion processingis less engaged in psychopathic individuals in
response to facial displays of emotion (Gordon et al. 2004), emotional pictures (Muller
et al. 2003, 2008), and emotional memory demands (Kiehl et al. 2001). Finally, structural
neuroimaging studies have shown a negative association between psychopathy level and
amygdala volume (Tiihonen et al. 2000; Ermer et al. 2012).
Evidence of biological abnormalities has become increasingly valuable in validating
behavioral evidence of psychopathic emotional processing deficits, demonstrating that
these deficits are physically instantiated, identifying potential causes of these deficits, and
informing neurocognitive treatment models for such deficits.
2. Scientific Evidence of Inhibition/Impulsivity Problems
Several behavioral studies suggest that psychopathic individuals present deficiencies in the
ability to inhibit a temptation in order to avoid punishment. Initial support for this

The Impact of Neuroimages on Sentencing

111

characterization derives from a passive avoidance learning study by Lykken (1957). The task
examined psychopathic performance navigating a difficult maze. Navigation errors were
punished by an electric shock. Psychopathic commission errors significantly exceeded
those of control participants despite the increased rate of punishment. This pattern of
disinhibition has been successfully extended to other tasks (Blair et al. 2004a; Hare 1970;
Newman 1987; Schmauk 1970).
Similar patterns have also been observed in gambling tasks measuring risk management abilities. In these tasks, participants are typically instructed to draw playing cards from
any of four decks. Each deck has been preloaded with either a favorable or unfavorable
distribution of point rewards or deductions. Participants must learn by trial and error which
decks will yield maximal rewards. High psychopathy individuals show a consistent tendency
to perseverate on riskier decks and this ultimately results in poorer task performance than
among control participants (van Honk et al. 2002; see also Newman et al. 1987; Siegel 1978).
A number of biological studies support the behavioral evidence for psychopathic
inhibition problems. By increasing adrenaline concentrations in the bloodstream,
Schachter and Latan (1964) were able to reduce psychopathic inhibition errors in a
passive avoidance learning task, suggesting an important role for the physiological stress
response in guiding successful inhibition.
Using a go/no-go task designed to measure response inhibition, Kiehl and colleagues
(2000) observed reduced potentiation of the frontal N275 component in psychopathic
individuals participating in an event-related potential (ERP) study. Complementing this
finding, functional magnetic resonance imaging studies have demonstrated reduced
engagement of the ventromedial prefrontal cortex (vmPFC), orbital frontal cortex (OFC),
and amygdala in psychopathic individuals participating in inhibition-related tasks (Blair
2003, 2004, 2005, 2008). Indeed, their patterns of poor performance on these tasks have
been analogized to that of patients with damage to the ventromedial prefrontal (Koenigs
et al. 2010) and orbitofrontal (Anderson et al. 1999) cortices. Moreover, recent evidence of
cortical thinning in the anterior cingulate and insula, as well as reduced functional connectivity between these regions (Ly et al. 2012), comports with research implicating these
regions in the ability to inhibit an undesirable response (Devinsky et al. 1995; Kosson et al.
2006).
Finally, genetic variation has been associated with impulsive violent behavior, as
commonly seen in psychopathy. One of the strongest genetic predictors of impulsive
violence is the short variant of the monoamine oxidase A (MAOA) allele. This allele is
believed to regulate the transmission of serotonin (Viding & Frith 2006).
These other studies have been used to support arguments that people high in
psychopathy suffer from a neurobiological condition hampering affective and volitional
processing often instrumental to the motivation to control undesirable behavior (Blair
2001, 2012; see also Morse 2008).
3. Psychopathy in the Courtroom
In criminal trials, scientific evidence that psychopathy is a pathological condition is a
double-edged sword. On one hand, defense counsel might present evidence of psychopathy

112

Saks et al.

in an attempt to mitigate responsibility by demonstrating reduced behavioral control. On

the other hand, the prosecution might use this diagnosis in the penalty phase to argue for
future dangerousness, resistance to treatment, and, ultimately, harsher punishment
(Viljoen et al. 2010). Whether justified or not, experimental research has shown that mock
jurors indeed support harsher punishments for defendants described as exhibiting psychopathic traits than for those who were not (e.g., Edens et al. 2003, 2005). However, in an
experimental study of trial court judges, when the cause of a psychopathic defendants
diagnosis was attributable to psychobiological factors, judges granted a more lenient sentence than when the scientific basis of the defendants psychopathy was not introduced
(Aspinwall et al. 2012). To complicate matters, the mere psychopathy label has been shown
in mock jury studies to aggravate, or occasionally mitigate (Cox et al. 2010), verdict outcomes irrespective of the trait evidence underlying this diagnosis (Boccaccini et al. 2008;
Rendell et al. 2010). Thus, observer perceptions of this personality construct might be
independent of and at least as important as the supporting evidence.
B. Schizophrenia
Another disorder strongly connected to problems in behavioral inhibition is schizophrenia.
Schizophrenia is a degenerative mental illness characterized by psychosis and disorganization of thought, emotion, and behavior. A diagnosis of schizophrenia is defined using the
Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) (APA 2000). Although schizophrenia affects less than 1 percent of the general population, the rate of violent convictions
among schizophrenic individuals is substantially higher than that of community controls
(Wallace et al. 2004; Wessely 1998).
1. Scientific Evidence of Affective Problems
Behavioral evidence suggests that schizophrenia is associated with deficits in affective
processing. For instance, schizophrenic individuals exhibit fewer positive and negative
facial displays of emotion than controls in both social and nonsocial contexts (see Kring &
Moran 2008). Moreover, one study reported that in a sample of schizophrenic patients,
those who were least emotionally expressive had poorer interpersonal relationship quality
and social adjustment (Bellack et al. 1990). A third study reported poor recognition of
emotional facial displays in schizophrenic participants (Weiss et al. 2006).
Affective deficits in schizophrenia have been validated using biological measures. In
one study, researchers presented participants with emotionally evocative film clips. They
observed increased skin conductance reactivity accompanied by a reduction in emotional
facial displays among schizophrenic participants compared to control participants (Kring &
Neale 1996). Volumetric studies of the amygdala have observed reductions in gray matter
density in this region by upward of 6 percent (see Wright et al. 2000). Similarly, functional
imaging studies of schizophrenic patients have shown reduced amygdala responsiveness to
fearful facial displays compared to control participants (Phillips et al. 1999). According to
some researchers, affective dysfunction in schizophrenia could hamper normal socialization and appropriate responses to emotional situations and, as a result, could increase risk
of criminal behavior (Weiss et al. 2006).

The Impact of Neuroimages on Sentencing

113

2. Scientific Evidence of Inhibition/Impulsivity Problems

Behavioral studies have shown that schizophrenic patients exhibit deficits in error monitoring and error correction in motor tasks (Malenka et al. 1982; Turken et al. 2003), and
these deficits impair their executive functioning (Silver & Goodman 2007).
Biological studies have found evidence consistent with behavioral inhibition problems. For instance, Kiehl et al. (2000) found a reduced N275 ERP response in schizophrenics compared to control participants in a go/no-go response inhibition task, and this
pattern was paralleled by an increase in commission errors (Nordahl et al. 2001; Kerns et al.
2005; Laurens et al. 2003). Others have shown functional abnormalities in the anterior
cingulate cortex (ACC) of schizophrenics, a brain region underlying the ability to detect
conflicts and correct commission errors. For instance, using positron emission tomography,
researchers found that ACC metabolism was correlated with a high error rate in a Stroop
task among schizophrenic patients but not among controls (Nordahl et al. 2001). Likewise,
fMRI studies with schizophrenic patients have observed reduced engagement of the ACC
during error/conflict-related tasks relative to controls (Kerns et al. 2005; Laurens et al.
2003).
Taken together, studies of schizophrenia make a strong case for both affective
and volitional deficits that could bear on the behavioral inhibition abilities of these
individuals.
3. Schizophrenia in the Courtroom
In criminal trials, evidence of schizophrenia is commonly admitted to support a defense of
not guilty by reason of insanity. Indeed, psychosis evidence has a strong track record in the
majority of successful insanity acquittals (Rice & Harris 1990). In mock jury trials, schizophrenic and psychotic defendants are more likely to be acquitted on the basis of insanity
than are personality disordered defendants (Roberts et al. 1987) and nondisordered
defendants (Gurley & Marcus 2008). However, as with psychopathy, they are also perceived
to be more dangerous on the basis of diagnosis alone, even when evidentiary testimony
about their risk level is controlled (Edens et al. 2004).

III. The Present Experiments

We report two experiments designed to evaluate the added impact, if any, of neuroimage
evidence offered in mitigation in the penalty phase of a capital murder trial. Both experiments employed large, nationally representative samples of jury-eligible, death-qualified
citizens as mock jurors.
Both experiments were designed to facilitate testing of the effects of neuroimages
against other evidence: neuroscience expert testimony sans brain images, evidence of
relevant genetic abnormality (an alternative kind of cutting-edge, high-technology evidence
that led to the same diagnosis as the neurological testing), and clinical psychological
examination (also leading to the same diagnosis). Another control group saw the case
without any expert testimony.

114

Saks et al.

To test the possible impact of images for different abnormalities, the various conditions described above were crossed with diagnoses of psychopathy or schizophrenia as an
additional independent variable. The first experiment also included a set of conditions in
which the diagnosis was that the defendant was healthy.
The second experiment was designed to replicate the first (with the same diagnoses
and the same types of expert evidence) as well as to ask a new question. The new inquiry
involved manipulating which side offered the neuroscience expert evidence. In half the
conditions, the prosecution offered expert testimony (leading to a diagnosis of psychopathy
or schizophrenia), arguing that the evidence is aggravating and supports a verdict of death.
In the other half, the defense offered the expert testimony (leading to a diagnosis of
psychopathy or schizophrenia), arguing that the evidence is mitigating and supports a
verdict of life. Further, the second experiment manipulated dangerousness, with the expert
witness in half the conditions predicting low, and in the other half high, levels of future
dangerousness by the defendant.
In addition to asking mock jurors to render verdicts (death or life in prison, and in
the case of life sentence whether incarceration should be in a prison or a secure treatment
facility), participants were also asked to provide ratings of what they perceived to be the
defendants dangerousness, responsibility for his actions, and amenability to treatment.
A. Experiment 1
1. Methods and Procedure
Our general methods here follow the procedures used for previous studies in this line of
research.3 A marketing firmSurvey Sampling International (SSI)was employed to
gather a demographically diverse sample of U.S. residents 18 years of age or older. After
being recruited by SSI, the participants were directed to an online experiment system that
delivered the experimental materials. Before being allowed to enter the experiment, a
death-qualification question was administered asking whether the participant would be
willing to sentence a person to death if called for and supported by the proper evidence.
Only participants who answered in the affirmative were allowed to continue.4 Overall, 825
participants completed the experiment.5
The median age of participants was 51 years, 58.3 percent of the sample was female,
84.6 percent of the sample was Caucasian, and 36 percent of the sample held at least a
four-year college degree.

See, e.g., Schweitzer and Saks (2011) and Schweitzer et al. (2011).

Due to limitations of the online experiment system, we were unable to calculate the percentage of participants who
were excluded on the basis of failure to be death qualified.

An initial sample of 999 participants completed the experiment. Consistent with past research (e.g., Schweitzer &
Saks 2011; Schweitzer et al. 2011), participants who completed the experiment in fewer than 300 seconds or scored
below a 67 percent on a recall quiz of the materials were excluded. A sensitivity power analysis indicates that this
sample size is sufficient to detect an effect of 2 = 0.02 for the 3 4 interaction, with greater sensitivity for the main
effects and planned comparisons. Mean time for completion was 936 seconds with a SD of 623 seconds.

The Impact of Neuroimages on Sentencing

115

At the beginning of the experiment, participants were instructed that they were to
read a summary of a criminal case in which the defendant had been convicted of firstdegree murder, and then decide what sentence (death vs. life in prison) the defendant
should receive. At this point, the participants were randomly assigned into one of 13
different experimental conditions. Twelve of these conditions were part of a 3 (Diagnosis:
Healthy, Schizophrenia, Psychopathy) 4 (Evidence: Clinical, Genetic, Neurological,
Neuroimage) factorial design. In addition, a control condition was included in which no
diagnostic information or expert evidence was presented.6 The participants were presented
with one of 13 penalty-phase trial summaries, each about 500 words long, providing
information about the defendants background, testimony from the defendants parents,
and diagnostic/expert evidence corresponding with a participants assigned experimental
condition. A sample case summary is found in the Appendix.7
After reading the case summary, the participants were instructed to provide a sentence (death vs. life in prison without the possibility of parole). Participants were informed
that once they provided their sentence recommendation, they would be unable to change
it. For those participants who chose a sentence of life in prison, a second question asked for
a recommendation of the type of facility in which the defendant should serve his sentence:
a secure psychiatric facility or a traditional prison, which also allowed participants to divide
the defendants sentence between the two facilities in whatever proportion they wished.
Following the sentencing items, participants were asked to provide judgments of two
potential mediatorsmoral responsibility8 and future dangerousness9followed by demographic and individual difference measures.
2. Results
Across all conditions, 56.4 percent of the jurors sentenced the defendant to death, with 43.6
percent selecting life in prison. Of the participants who chose life, an average of 42 percent
of the defendants sentence was to be served at a psychiatric facility, with the remainder in
prison.10 While there was overall disagreement on the sentence, there was considerable

An additional three between-subjects conditions were also included in the experiment for pilot testing future
materials.

To be consistent with previous research, and to control for any confounding that arises from the inclusion of
additional expert evidence that may accompany a neuroimage, the two key evidentiary conditions (neurological vs.
neuroimage) have identical text. Both conditions describe that a neuroimaging technique (e.g., fMRI) was conducted
as part of the diagnostic procedure; however, the neuroimage condition displays that image to the participants as part
of the case summary, while the nonimage conditions display a control image of a courtroom.

For example, The defendant is morally responsible for his actions.

For example, If sent to prison, the defendant would be a danger to guards and prisoners.

10

Of the participants, 18.3 percent recommended that the defendant spend 100 percent of his sentence in treatment,
while 26.3 percent recommended the defendant serve his entire sentence in prison. The remaining participants split
the sentence.

116

Saks et al.

agreement11 on a series of subjective judgment questions that the defendant was indeed
dangerous (99 percent), deserved to be punished (95 percent), that the public needs to be
protected from the defendant (98 percent), and that the defendant is morally responsible
for his actions (87 percent). There was moderate agreement that while incarcerated, the
defendant would be a threat to prison guards or other prisoners (73 percent), that the
defendant was in control of his actions at the time of his crime (70 percent), and that
the defendant could not be treated for his disorder (28 percent agreeing that the defendant could be treated).
To address our first research questionwhether neuroimage-based evidence is more
persuasive than nonneuroimage-based evidencea series of planned comparisons were
conducted comparing the neuroimage-based evidence condition to the general neuroscience condition on sentence recommendations (life in prison vs. death penalty), perceived
responsibility (e.g., whether the defendant was morally responsible for his actions; whether
the defendant was in control of his actions), and future dangerousness (e.g., whether the
defendant would be a danger while in prison; whether the defendant could be treated for
his condition).
For a healthy defendant who lacked any mental defect, evidence that included
neuroimages did not differ in its effect from nonimage-based evidence on responsibility,12
future dangerousness,13 or sentences.14 For a defendant who had been diagnosed with
psychopathy, neuroimage-based evidence did not influence judgments of future dangerousness,15 but it did decrease judgments of responsibility (M = 3.84) compared to the
nonneuroimage evidence (M = 4.26),16 and marginally decreased death sentences (from
62 percent to 47 percent).17 Finally, for a defendant who had been diagnosed with schizophrenia, neuroimage-based evidence again did not influence judgments of future dangerousness,18 but significantly increased judgments of perceived responsibility (M = 4.34)
compared to the nonimage condition (M = 3.94).19 It did not correspondingly increase
death sentences (42 percent without neuroimage evidence, 52 percent with).20 Exploring

11

Responding with either a 4 or 5 on a five-point disagree-agree Likert-type scale.

12

F (1, 99) = 0.006, p = 0.94, 2 < 0.001.

13

F (1, 99) = 0.178, p = 0.67, 2 = 0.002.

14

F (1, 99) = 0.065, p = 0.79, 2 = 0.001.

15

F (1, 105) = 0.375, p = 0.54, 2 = 0.004.

16

F (1, 105) = 4.82, p = 0.03, 2 = .044.

17

F (1, 105) = 2.47, p = 0.12, 2 = 0.023.

18

F (1, 98) = 0.402, p = 0.53, 2 = 0.004.

19

F (1, 98) = 3.97, p = 0.05, 2 = 0.039.

20

F (1, 98) = 0.947, p = 0.33, 2 = 0.010.

The Impact of Neuroimages on Sentencing

117

this effect further, a 2 (Psychopathy vs. Schizophrenia) 2 (Neurological vs. Neuroimage)

ANOVA was conducted on our three primary dependent variables. Consistent with the
above findings, interactions were found on verdicts21 and judgments of responsibility,22
but not on future dangerousness.23 More specifically, simple effects tests revealed that
when neuroscience evidence was presented without an image, psychopaths were judged
more responsible for their actions (M = 4.26) and were more likely to receive a
death sentence (63 percent) than were schizophrenics (M = 3.94, p = 0.08, and 42
percent, p = 0.03, respectively). However, when this very same evidence was presented
accompanied by a neuroimage, psychopaths were judged to be less responsible for their
actions (M = 3.83) than were schizophrenics (M = 4.34, p = 0.02), and the differences in
sentencing disappeared (47 percent for psychopaths, 52 percent for schizophrenics,
p = 0.62).
We next examined how each of these conditions compared to our overall control
condition, which did not present any evidence of a disorder. Within the control condition,
62 percent of participants sentenced the defendant to death, and participants provided a
mean responsibility score of 4.43 out of 5, and a mean future dangerousness score of 3.97
out of 5.
Using a simple contrast that compared the control condition to all other conditions,
we found that healthy defendants did not differ from control defendants on perceived
dangerousness, responsibility, or sentences, regardless of the evidence type.24 Defendants
who had been diagnosed with psychopathy did not differ from control participants when
nonimage neuroscience evidence was presented;25 however, when a neuroimage accompanied the evidence, perceived responsibility significantly decreased26 while dangerousness
and sentences showed a nonsignificant downward trend.27 However, for defendants who
had been diagnosed with schizophrenia, nonimage neuroscience evidence decreased judgments of responsibility, future dangerousness, and death sentences relative to the control
condition.28 However, when a neuroimage accompanied the evidence, those differences

21

F (1, 204) = 3.39, p = 0.07, 2 = 0.016.

22

F (1, 204) = 8.76, p < 0.01, 2 = 0.041.

23

F (1, 204) = 0.763, p = 0.37, 2 = 0.004.

24

All ps > 0.25.

25

All ps > 0.30.

26

Mean difference = 0.421, SE = 0.184, p = 0.001.

For dangerousness, mean difference = 0.277, SE = 0.195, p = 0.156; for sentences, mean difference = 0.146,
SE = 0.098, p = 0.138.

27

For responsibility, mean difference = 0.490, SE = 0.176, p = 0.006; for dangerousness, mean difference = 0.365,
SE = 0.192, p = 0.058; for sentences, mean difference = 0.192, SE = 0.097, p = 0.048.

28

118

Saks et al.

Figure 1: Mock juror sentences, judgments of responsibility, and perceptions of future

dangerousness as a function of diagnosis and evidence type (neuroscience with
neuroimage, neuroscience without image).
Sentences (% Death)

Responsibility

Dangerousness

= Neuroimage
= Neuroscience Only
= Control Condition
Source: Original data; see Method section of Experiment 1 for details.

disappeared, with judgments failing to differ significantly from the control condition.29
These results are illustrated in Figure 1.
We then stepped back and conducted overall comparisons of the four different
evidence types, assessing how clinical and genetically-based evidence influences jurors
relative to neurologically-based evidence. A series of one-way analyses of variance followed
by paired comparison tests with the Tukey HSD correction examined the main effect of the
full evidence manipulation within each of the three diagnoses. In these omnibus tests, when
the defendant was diagnosed as healthy, effects were found on judgments of responsibility30
and future dangerousness,31 but not on sentences.32 Posthoc tests revealed the source of the
significant findings to be the genetic evidence condition, which produced somewhat lower
responsibility ratings (M = 4.34) than the clinical (M = 4.69; p = 0.05), neurological
(M = 4.65; p = 0.18), and neuroimage (M = 4.64; p = 0.14) conditions, as well as lower
future dangerousness ratings (M = 3.32) than the clinical (M = 3.80; p = 0.06), neurological

29

All ps > 0.25.

30

F (3, 211) = 2.50, p = 0.06, 2 = 0.034.

31

F (3, 211) = 4.59, p < 0.01, 2 = 0.061.

32

F (3, 210) = 0.29, p = 0.83, 2 < 0.01.

The Impact of Neuroimages on Sentencing

119

(M = 4.00; p = 0.01), and neuroimage (M = 3.96; p = 0.01) conditions. However, when the
defendant had received diagnoses of psychopathy or schizophrenia, no significant omnibus
effects were found.33
B. Experiment 2
1. Methods and Procedure
This experiment followed the same procedures as the first experiment: SSI was used to
gather a demographically diverse national sample of jury-eligible, death-qualified participants, who were directed to an online experiment system that delivered the experimental
materials. Overall, 882 participants completed the experiment.34 The median age of the
participants was 48 years, 55 percent of the sample was female, 86.3 percent of the sample
was Caucasian, and 31 percent of the sample held at least a four-year college degree.
The evidence presented at the penalty-phase hearing differed from the first experiment by including a manipulation of the defendants future dangerousness: the expert
stated that the defendant would/would not be a threat to others. Further, we added a
manipulation of which party offers the expert evidencethe defense (offering as a mitigating factor) or the prosecution (offering as an aggravating factor).
Upon arriving at our online experiment system, the participants were randomly
assigned to one of 32 different conditions constituting a 4 (Evidence: Clinical, Genetic,
Neurological, Neuroimage) 2 (Diagnosis: Schizophrenia vs. Psychopathy) 2 (Future
Dangerousness: High vs. Low) 2 (Which Party Offers Evidence: Prosecution vs. Defense)
between-subjects design. We again asked participants to provide sentence decisions along
with measures of perceived dangerousness, treatability, and the extent to which they
believed the defendant was in control of his actions at the time of the crime.
2. Results
Across all conditions, 53.5 percent of the jurors sentenced the defendant to death, with 46.5
percent selecting life in prison. Of the participants who chose life, an average of 45 percent
of the defendants sentence was to be served at a psychiatric facility, with the remainder in
prison.35 In addition to sentence, three additional dependent variables were created by

All ps > 0.15, all 2 < 0.02. Put another way, when the additional evidence types (clinical and genetic) are included
in an omnibus ANOVA, the significant Diagnosis Evidence interaction described above is masked.

33

34

An initial sample of 1,151 participants completed the experiment. Consistent with Experiment 1 and past research,
participants who completed the experiment in fewer than 300 seconds and/or scored below a 67 percent on a recall
quiz of the materials were excluded. Individuals who participated in Experiment 1 were not allowed to participate in
Experiment 2. A sensitivity power analysis indicates that this sample size is sufficient to detect an effect of 2 = 0.01 for
the four-way interaction, with greater sensitivity for the main effects, lower-order interactions, and planned comparisons. Mean time for completion was 883 seconds with a SD of 668 seconds.
35

Of the participants, 18.8 percent recommended that the defendant spend 100 percent of his sentence in treatment,
while 19.6 percent recommended the defendant serve his entire sentence in prison. The remaining participants split
the sentence.

120

Saks et al.

averaging items designed to measure perceived dangerousness,36 culpability,37 and a single

item measuring treatability.38
A 4 (Evidence: Clinical, Genetic, Neurological, Neuroimage) 2 (Diagnosis: Schizophrenia vs. Psychopathy) 2 (Future Dangerousness: High vs. Low) 2 (Which Party
Offers Evidence: Prosecution vs. Defense) between-subjects ANOVA was conducted on
the participants sentencing decisions. An overall main effect of future dangerousness
emergeddefendants who were characterized as being dangerous in the future were more
likely to be sentenced to death (59 percent) than those who were characterized as being less
of a danger (47 percent).39 Additionally, a main effect of diagnosis emerged in that
psychopaths were more likely to receive a death sentence (58 percent) compared to
schizophrenics (49 percent).40
Interestingly, however, a significant Evidence Offering Party interaction emerged.41
As illustrated in Figure 2, clinical, genetic, and nonimage neurological evidence produced
a backfire effect. When such evidence was presented by the defense for mitigation, the
defendant was sentenced to death more often, while the same evidence offered by the
prosecution for aggravation led to a lower rate of death sentences.42 But when the expert
evidence contained a neuroimage, the effect reversed, with defense evidence resulting in a
lower rate of death sentences, as would be expected.43
We then sought to identify potential mediators of this effect. To begin, a
binary logistic regression analysis was conducted predicting sentence choice (death
vs. life) from the treatability, dangerousness, and responsibility (in-control-ofbehavior) variables. As expected, all three measures were highly predictive of sentence

36

For example, The defendant is dangerous to society; If sent to prison, the defendant would be a danger to guards
and prisoners; People need to be protected from the defendant.

37

For example, The defendant was in control of his actions at the time of the crime; The defendant is morally
responsible for his actions.

38

The defendant could be treated for his problems.

39

F (1, 846) = 11.96, p < 0.01, 2 = 0.014.

40

F (1, 846) = 6.30, p = 0.01, 2 = 0.007.

41

F (3, 846) = 2.92, p = 0.03, 2 = 0.010.

42

Death sentences are consistently less common in these conditions when the evidence was presented by the
prosecution, though the differences in clinical and genetic evidence conditions are not statistically significant. The
neurological condition does result in a significant difference, F (1, 218) = 6.07, p = 0.02, 2 = 0.027. A simple 2
(Evidence: Neuroimage vs. Others) 2 (Party: Prosecution vs. Defense) interaction test (F (1, 874) = 5.87, p = 0.02,
2 = 0.007) confirms that in the neuroimage condition, a marginally greater proportion of death sentences are
recommended when the evidence is presented by the prosecution (58 percent) versus the defense (48 percent,
p = 0.11); however, in the nonneuroimage conditions, death sentences were more common when the evidence was
presented by the defense (58 percent) versus the prosecution (50 percent, p = 0.03).
43

F (1, 207) = 2.51, p = 0.11, 2 = 0.012.

The Impact of Neuroimages on Sentencing

121

Figure 2: Proportion of mock jurors sentencing the defendant to death as a function of

evidence type (clinical, genetic, neurological without image, neurological with image) and
which party proffered the evidence.

Source: Original data; see Method section of Experiment 2 for details.

decisions.44 Next, our primary analysis was rerun three timesonce with treatability, dangerousness, and responsibility as covariates.45 In the analyses co-varying dangerousness and
responsibility, the interaction remained significant; however, when treatability was entered
as a covariate, the Offering Party Evidence interaction became nonsignificant.46 This
suggests that of the variables tested, treatability is the most likely candidate to be mediating
the impact of the expert evidence on sentencing decisions.
One final effect emerged from our overall analysis: a significant Offering
Party Future Dangerousness interaction was present.47 When the defendant was characterized as nondangerous, the backfire effect identified above is not statistically significant.

44
Culpability, B = 1.14, SEB = 0.110, Wald = 107.46, p < 0.001, OR = 3.13; dangerousness, B = 1.10, SEB = 0.157,
Wald = 49.16, p < 0.001, OR = 3.02; treatability, B = 0.57, SEB = 0.069, Wald = 68.31, p < 0.001, OR = 0.56.

If any of these three items mediate the Party Evidence interaction, we would expect to see an increase in p value
and decrease in effect size in the Party Evidence interaction once the potential mediator is entered as a covariate.

45

46
F (3,844) = 1.73, p = 0.16, 2 = 0.006. Due to the complexity of the experimental design, it would be premature to
state that treatability is the primary mediator of our effect; however, it is the only variable for which we have evidence
of mediation.

47

F (1, 846) = 4.00, p = 0.05, 2 = 0.005.

122

Saks et al.

However, when the defendant was described as dangerous, the backfire effect was present,
with the defense evidence leading to a higher rate of death sentences (64 percent) than
when the prosecution offered the evidence (54 percent).48 No other significant effects
emerged from the overall analysis.

IV. Discussion
A. Past Research
Numerous commentators have expressed concern that neuroimages might be inordinately
persuasive to jurors. Of particular interest are neuroimages offered as evidence in criminal
trials, and especially images of functional abnormality (rather than anatomical damage).
Empirical studies near, though not in the heartland of, that domain of interest have lent
support to the commentators predictions.
Neurological explanations (without images) of psychological phenomena are more satisfying
than non-neurological explanations, news reports of neurological findings are more convincing
when accompanied by brain images than reports without the images, crime photographs produce
more convictions than the same evidence without photographs, and an image of and testimony
about a severely physically damaged brain, especially when caused by recent trauma, was more
likely to produce insanity verdicts than evidence lacking such an image and explanation.
(Schweitzer et al. 2011:362)

A series of experiments focused directly on the area of central concern, however,

found no evidence that neuroimages themselves in any way influenced mock jurors verdicts, sentence recommendations, or other judgments about the defendant over and above
the experts verbal testimony. Those studies did find that neuroscience-based expert evidence was more persuasive than other kinds of expert evidence, but the effect was no more
potent when neuroimages were part of the expert testimony than when neuroexpert
testimony was given without any neuroimages. These findings held true for criminal cases
involving mens rea defenses as well as cases involving insanity defenses. The lack of differences remained consistent and convincing, despite various experimental efforts to detect
effects if they existed.
B. The Present StudiesWhy the Difference from Earlier Studies?
By contrast, the research reported in the present article did find effects of neuroimage
evidence, in the context of a criminal trial, over and above the effects of other kinds of
expert evidence, including the same neuroscience expert testimony without the
neuroimages.
Why the difference between the present findings and those of earlier experiments?
One explanation might be that in the earlier sets of experiments the defense asked mock
jurors to exonerate the defendant on the basis of his mental abnormality. They were

48

F (1, 455) = 4.61, p = 0.03, 2 = 0.010.

The Impact of Neuroimages on Sentencing

123

unwilling to do soregardless of the kind of expert evidence marshaled in support of the

defensebecause they put protection of the public from a plainly dangerous person above
all other considerations, including conceptions of law or justice. That those earlier participants were being practical, not vengeful, is further evidenced by the finding that in large
numbers they supported the idea of having the defendant serve various portions of his
incarceration in therapy rather than undergoing punishment.
The present pair of experiments gives jurors a different choice. Protection from a
dangerous person is a foregone conclusion: this defendant has been convicted of capital
murder and is going to spend life in prison or is going to be executed. This situation allows
jurors to put aside concerns of public safety and move on to thinking about what is just.
What they view as just turns in part on what they believe to be true of the defendants
condition and the causes of his dangerous behavior.
C. When the Defendant is Psychopathic
For defendants diagnosed with psychopathy, when jurors were persuaded that the defendants behavior was the result of a genuine brain-based defect, which through no fault of his
own contributes to his dangerous behavior, jurors were more likely to see him as less
responsible for his behavior and were more likely to spare his life. The most persuasive form
that evidence took was when neuroimages accompanied the neuroscience expert testimony. Responses to defendants whose mitigating expert evidence was other than
neuroimage evidence differed little from jurors responses to mentally healthy defendants.
D. Backfire Effect Further Shows the More Persuasive Impact of Images
The persuasive impact of the neuroimage evidence was further supported by findings
related to which party offered the neuroimage evidence in support of its arguments.
When offered by the defense in mitigation, neuroimage evidence had the effect of
reducing the rate of death sentences. However, when the defense offered other kinds of
expert evidence (clinical, genetic, nonimage neurological) on the same issue, the rate
of death sentences was greater. Looking at the obverse side of the case, when the prosecution offered the same neuroimage evidence as aggravating, it had the effect of increasing
the rate of death sentences, while other forms of expert evidence decreased death sentences. The neuroimage evidence bolsters the argument of whichever side uses it: it helps
the prosecution win more death verdicts and helps the defense prevent death verdicts.
For both advocates, then, the neuroimage evidence was the most persuasive, advancing their arguments and moving verdicts in the direction they sought. Other forms of
evidence in support of the very same arguments backfired, and produced effects opposite
to those each sought.
E. Differences as a Function of Diagnosis
Jurors did not respond to schizophrenics in the same way that they did to psychopaths. First,
all else equal, psychopaths were more likely to be sentenced to death than schizophrenics.
Although neuroimage evidence reduced judgments of responsibility and sentences of

124

Saks et al.

death for defendants diagnosed as psychopathic, it increased judgments of responsibility

for defendants diagnosed as schizophrenic. Moreover, for schizophrenic defendants,
nonimage neurological evidence decreased death sentences and decreased judgments of
responsibility and of dangerousness.
The neuroimage evidence still had a special impact but moved judgments in one
direction for schizophrenics and in the opposite direction for psychopaths. The underlying
difference must have something to do with jurors differing beliefs and feelings about the
two different diagnoses. The nature of those beliefs, and how they interacted with the
neuroimages to lead jurors to opposite responses to the two diagnoses, remains a puzzle.
Solving that puzzle will require more research. Future inquiry likely would focus on
laypersons understanding of the nature of the two diagnoses and differences between
them. One clue might be that, of the several measures of juror perception we collected
(defendants dangerousness, control over his behavior, and treatability), only treatability
mediated between the evidence presented and the verdicts reached by the mock jurors.
Perhaps, upon becoming convinced by neuroimage evidence that a defendant truly suffered from psychopathy, something in their beliefs about the nature of psychopathy led
jurors to see the defendant as more treatable, and in turn more qualified for life imprisonment. Upon becoming convinced by neuroimage evidence that the defendant truly
suffered from schizophrenia, jurors beliefs about schizophrenia led them to view the
defendant as less qualified for imprisonment.

F. From Mock Jurors to Actual Trial Juries

It is worth reminding ourselves that in capital penalty-phase hearings, the jury must reach
a unanimous verdict in order to impose death. While a single juror rarely stands up to the
social pressure of a unanimous majority pushing in the opposite direction, it would take
only a few jurors favoring a sentence of life to block the jury from delivering a death penalty.
The present data suggest that neuroimage evidence is capable of creating a critical
minorityif not a majorityof jurors favoring life over death, at least for defendants
diagnosed with psychopathy.

G. Should Neuroimage Evidence be Admissible?

One of the concerns initially raised about neuroimage evidence was that its impact would
be so overwhelming that jurors would crumple under its persuasive weight. If so, went the
concern, such evidence might have to be excluded as being unduly prejudicial. Even in the
present studies, which do finally find an independent effect of neuroimages, the effect was
a small one.
Nevertheless, the question remains whether these neuroimages change verdicts
because of evidence that is probative or prejudicial. Answering this question empirically
would require more careful measurement of the meaning of the image content, and how
it relates to the verbal testimony of the expert. Ultimately, however, this might be less of
an empirical question than a normative question. Nevertheless, we think the present

The Impact of Neuroimages on Sentencing

125

experimentsalong with earlier ones of this kindsuggest that, so long as neuroimage

evidence is otherwise found to be admissible, there is no reason to exclude it out of fear that
jurors will be unthinkingly deferential to it.
H. Limitations
Of course, all research designs involve tradeoffs. The present studies are mock jury experiments, using brief written and visual materials, online, obtaining individual responses
(rather than group deliberation), and no defendant whose life or death is in the jurors
hands. In an actual penalty-phase hearing, there would be live witnesses presenting much
more evidence in the midst of more formality and tension, involving more detailed aggravating and mitigating evidence and arguments. The obvious advantage of a jury simulation
experiment is the ability researchers have to vary exactly what they are trying to learn
about, and to do so with a high degree of internal validity. The potential disadvantage is
whether findings generalize to the setting of actual trials, that is, the question of external
validity.
Despite obvious differences in procedures, studies comparing mock jury experiments
that ranged along a dimension of verisimilitudefrom the most bare-bones jury decisionmaking studies to those consisting of lavish simulations approaching the procedures of
actual trials as much as a simulation canfound little difference in conclusions about
functional relationships (i.e., whether a dependent variable changed as a function an
independent variable) (Bornstein 1999).
Such a result is less surprising when one considers that in order for study procedures to make a serious difference, verisimilitude would have to interact with a studys
independent and dependent variables in a way that reversed the direction of the functional relationship. For an effect to be stronger in one context and weaker in another is
inevitable; for an effect to be part of a cross-over interaction is less likely and, indeed, not
terribly plausible. Thus, when a number of social-cognitive effects found in laboratory
research using as participants college undergraduates were replicated using survey
research methods employing nationally representative samples, effects were found to
strengthen and weaken, but reversals of direction of an effect bordered on nonexistent
(Schneider et al. 2010; see studies collected at http://electionstudies.org/resources/
papers/pilotrpt.htm). Still, one never knows whether a finding will generalizes to contexts
outside of the research setting unless and until the phenomenon has been replicated in
the additional setting.
Furthermore, since jury (group) decisions have been found to be quite predictable
from knowledge of the distribution of individual preferences at the outset of deliberation
(using appropriate combinatorial models), little is lost when the unit of analysis in a study
consists of jurors rather than juries. As Kalven and Zeisel (1966) put the matter in their own
nonsimulation jury research, the process of deliberation that combines jurors individual
preferences into a group decision is much like developing a photograph: the process itself
is quite interesting, but the outcome has already been determined.
Moreover, an important virtue of the methods employed in the present studies is that
they replicate those of previous research that found no effects of neuroimages in other legal

126

Saks et al.

contexts, and so we avoid confounding study methods with the legal context under study
(mens rea defense, insanity defense, penalty phase). Further, we were able to systematically
vary influences on jurorswhich would be impractical to the point of impossible to
systematically vary in actual trialsenabling us to determine what effects neuroimages did
and did not have on the judgments and decisions of a national sample of mock jurors.
At the end of the day, however, there is no doubt that a further desirable step would
be to present mock jurors, drawn from court jury pools, with videos of more complete
penalty-phase hearings.

References
Aharoni, E., G. M. Vincent, C. L. Harenski, V. D. Calhoun, W. Sinnott-Armstrong, M. S. Gazzaniga, &
K. A. Kiehl (in press) Neuroprediction of Future Rearrest, Proceedings of the National Academy of
Sciences.
American Psychiatric Association (2000) Diagnostic and Statistical Manual of Mental Disorders, 4th ed.,
text rev. Washington, DC: Author.
American Psychological Association & Missouri Psychological Association, Brief to the Supreme Court
of the United States, Roper v. Simmons, 543 U.S. 551 (2006).
Anderson, S. W., A. Bechara, H. Damasio, D. Tranel, & A. Damasio (1999) Impairment of Social and
Moral Behavior Related to Early Damage in Human Prefrontal Cortex, 2(11) Nature Neuroscience 1032.
Aspinwall, L. G., T. R. Brown, & J. Tabery (2012) The Double-Edged Sword: Does Biomechanism
Increase or Decrease Judges Sentencing of Psychopaths? 337 Science 846.
Batts, S. (2009) Brain Lesions and Their Implications in Criminal Responsibility, 27(2) Behavioral
Sciences & the Law 261.
Bellack, A. S., R. L. Morrison, J. T. Wixted, & K. T. Mueser (1990) An Analysis of Social Competence
in Schizophrenia, 156 British J. of Psychiatry 809.
Blair, R. J. R. (1995) A Cognitive Developmental Approach to Morality: Investigating the Psychopath,
57 Cognition 1.
(1997) Moral Reasoning and the Child with Psychopathic Tendencies, 22(5) Personality &
Individual Differences 731.
(2001) Advances in Neuropsychiatry: Neurocognitive Models of Aggression, the Antisocial Personality Disorders, and Psychopathy, 71(6) J. of Neurology, Neurosurgery & Psychiatry 727.
(2003) Neurobiological Basis of Psychopathy, 182 British J. of Psychiatry 5.
(2004) The Roles of Orbital Frontal Cortex in the Modulation of Antisocial Behavior, 55 Brain
& Cognition 198.
(2005) Applying a Cognitive Neuroscience Perspective to the Disorder of Psychopathy, 17
Development & Psychopathology 865.
(2008) Fine Cuts of Empathy and the Amygdala: Dissociable Deficits in Psychopathy and
Autism, 61(1) Q. J. of Experimental Psychology 157.
(2012) Cortical Thinning and Functional Connectivity in Psychopathy, 169 American J. of Psychiatry 684.
Blair, R. J. R., D. G. V. Mitchell, A. Leonard, S. Budhani, K. S. Peschardt, & C. Newman (2004a) Passive
Avoidance Learning in Individuals with Psychopathy: Modulation by Reward But Not by Punishment, 37 Personality & Individuals Differences 1179.
Blair, R. J. R., D. G. V. Mitchell, K. S. Peschardt, E. Colledge, R. A. Leonard, J. H. Shine, L. K. Murray,
& D. I. Perrett (2004b) Reduced Sensitivity to Others Fearful Expressions in Psychopathic
Individuals, 37 Personality & Individual Differences 1111.

The Impact of Neuroimages on Sentencing

127

(2004c) Reduced Sensitivity to Others Fearful Expressions in Psychopathic Individuals, 37

Personality & Individual Differences 1111.
Blair, R. J. R., D. G. V. Mitchell, R. A. Richell, S. Kelly, A. Leonard, C. Newman, & S. K. Scott (2002)
Turning a Deaf Ear to Fear: Impaired Recognition of Vocal Affect in Psychopathic Individuals,
111(4) J. of Abnormal Psychology 682.
Boccaccini, M. T., D. C. Murrie, J. W. Clark, & D. G. Cornell (2008) Describing, Diagnosing, and
Naming Psychopathy: How Do Youth Psychopathy Labels Influence Jurors? 26 Behavioral
Sciences & the Law 487.
Bornstein, B. H. (1999) The Ecological Validity of Jury Simulations: Is the Jury Still Out? 23 Law &
Human Behavior 7.
Brown, T., & E. Murphy (2010) Through a Scanner Darkly: Functional Neuroimaging as Evidence of
Criminal Defendants Past Mental States, 62 Stanford Law Rev. 1119.
Burke, J. D., R. Loeber, & B. B. Lahey (2007) Adolescent Conduct Disorder and Interpersonal
Callousness as Predictors of Psychopathy in Young Adults, 36 J. of Clinical Child & Adolescent
Psychology 334.
Cacioppo, J. T., G. G. Berntson, T. S. Lorig, C. J. Norris, E. Rickett, & H. Nusbaum (2003) Just Because
Youre Imaging the Brain Doesnt Mean You Can Stop Using Your Head: A Primer and Set of
First Principles, 85 J. of Personality & Social Psychology 650.
Carney, R. N., & J. R. Levin (2002) Pictorial Illustrations Still Improve Students Learning from Text,
14(1) Educational Psychology Rev. 5.
Coid, J., M. Yang, S. Ullrich, A. Roberts, & R. D. Hare (2009) Prevalence and Correlates of Psychopathic Traits in the Household Population of Great Britain, 32 International J. of Law &
Psychiatry 65.
Compton, S. E. (2010) Not Guilty by Reason of Neuroimaging: The Need for Cautionary Jury
Instructions for Neuroscience Evidence in Criminal Trials, 12 Vanderbilt J. of Entertainment &
Technology Law 333.
Cox, J., D. S. DeMatteo, & E. E. Foster (2010) The Effect of the Psychopathy ChecklistRevised in
Capital Cases: Mock Jurors Responses to the Label of Psychopathy, 28 Behavioral Sciences & the
Law 878.
Devinsky, O., M. J., Morrell, & B. A. Vogt (1995) Contributions of Anterior Cingulate Cortex to
Behaviour, 118 Brain 279.
Du, W., H. Li, V. D. Calhoun, S. Ma, T. Eichele, K. A. Kiehl, G. D. Pearlson, & T. Adali (2012) High
Classification Accuracy for Schizophrenia with Rest and Task fMRI data, 6 Frontiers in Human
Neuroscience 145.
Dumit, J. (1999) Objective Brains, Prejudicial Images, 12 Science in Context 173.
Edens, J. F., L. H. Colwell, D. M. Desforges, & K. Fernandez (2005) The Impact of Mental Health
Evidence on Support for Capital Punishment: Are Defendants Labeled Psychopathic Considered More Deserving of Death? 23(5) Behavioral Sciences & the Law 603.
Edens, J. F., D. M. Desforges, K. Fernandez, & C. A. Palac (2004) Effects of Psychopathy and Violence
Risk Testimony on Mock Juror Perceptions of Dangerousness in a Capital Murder Trial, 10(4)
Psychology, Crime & Law 412.
Edens, J. F., L. S. Guy, & K. Fernandez (2003) Psychopathic Traits Predict Attitudes Toward a Juvenile
Capital Murderer, 21(6) Behavioral Sciences & the Law 807.
Ermer, E., L. M. Cope, P. K. Nyalakanti, V. D. Calhoun, & K. A. Kiehl (2012) Aberrant Paralimbic Gray
Matter in Criminal Psychopathy, 121(3) J. of Abnormal Psychology 649.
(2013) Aberrant Paralimbic Gray Matter in Incarcerated Male Adolescents with Callous
Conduct Disorder, 52(1) J. of the American Academy of Child & Adolescent Psychiatry 94.
Frick, P. J., E. R. Kimonis, D. M. Dandreaux, & J. M. Farrell (2003) The 4-Year Stability of Psychopathic
Traits in Non-Referred Youth, 21 Behavioral Sciences & the Law 713.
Gazzaniga, M. S. (2008) The Law and Neuroscience, 60 Neuron 412.
Gordon, H. L., A. A Baird, & A. End (2004) Functional Differences Among Those High and Low on
a Trait Measure of Psychopathy, 56 Biological Psychiatry 516.

128

Saks et al.

Greene, E., & B. S. Cahill (2011) Effects of Neuroimaging Evidence on Mock Juror Decision Making,
30(3) Behavioral Sciences & the Law 280.
Gurley, J. R., & D. K. Marcus (2008) The Effects of Neuroimaging and Brain Injury on Insanity
Defense, 26 Behavioral Sciences & the Law 85.
Hare, R. D. (1970) Some Empirical Studies of Psychopathology, 18(1) Canadas Mental Health 4.
(2003) Manual for the Hare Psychopathy Checklist-Revised, 2nd ed. Toronto: Multi-Health Systems.
Hare, R. D., S. D. Hart, & T. J. Harpur (1991) Psychopathy and the DSM-IV Criteria for Antisocial
Personality Disorder, 100 J. of Abnormal Psychology 391.
Kalven, Harry, & Hans Zeisel (1966) The American Jury. Boston, MA: Little, Brown.
Kerns, J. G., J. D. Cohen, A. W. MacDonald 3rd, M. K. Johnson, V. A. Stenger, H. Aizenstein, & C. S.
Carter (2005) Decreased Conflict- and Error-Related Activity in the Anterior Cingulate Cortex
in Subjects with Schizophrenia, 162(10) American J. of Psychiatry 1833.
Khoshbin, L. S., & S. Khoshbin (2007) Imaging the Mind, Minding the Image: An Historical Introduction to Brain Imaging and the Law, 33(23) American J. of Law & Medicine 171.
Kiehl K. A., A. M. Smith, R. D. Hare, & P. F. Liddle (2000) An Event-Related Potential Investigation
of Response Inhibition in Schizophrenia and Psychopathy, 48 Biological Psychiatry 210.
Kiehl K. A., A. M. Smith, R. D. Hare, A. Mendrek, B. B. Forster, J. Brink, et al. (2001) Limbic
Abnormalities in Affective Processing by Criminal Psychopaths as Revealed by Functional Magnetic Resonance Imaging, 50 Biological Psychiatry 677.
Koenigs, M., M. Kruepke, & J. P. Newman (2010) Economic Decision-Making in Psychopathy: A
Comparison with Ventromedial Prefrontal Lesion Patients, 48(7) Neuropsychologia 2198.
Kosson, D. S., S. Budhani, M. Nakic, G. Chen, Z. S. Saad, M. Vythilingam, D. S. Pine, and R. J. R. Blair
(2006) The Role of Amygdala and Rostral Anterior Cingulate in Encoding Expected Outcomes
During Learning, 29(4) NeuroImage 1161.
Kring, A. M., & E. K. Moran (2008) Emotional Response Deficits in Schizophrenia: Insights from
Affective Science, 34(5) Schizophrenia Bulletin 819.
Kring, A. M., & J. M. Neale (1996) Do Schizophrenic Patients Show a Disjunctive Relationship Among
Expressive, Experiential, and Psychophysiological Components of Emotion? 105 J. of Abnormal
Psychology 249.
Kulynych, J. (1997) Psychiatric Neuroimaging Evidence: A High-Tech Crystal Ball? 49 Stanford Law
Rev. 1249.
Laurens, K. R., E. T. C. Ngan, A. T. Bates, K. A. Kiehl, & P. F. Liddle (2003) Rostral Anterior Cingulate
Cortex Dysfunction During Error Processing in Schizophrenia, 126(3) Brain 610.
Logothetis, N. (2008) What We Can Do and What We Cannot Do with fMRI, 453 Nature 869.
Ly, M., J. C. Motzkin, C. Philippi, G. R. Kirk, J. P. Newman, K. A. Kiehl, & M. Koenigs (2012) Cortical
Thinning in Psychopathy, 169 American J. of Psychiatry 743.
Lykken, D. (1957) A Study of Anxiety in the Sociopathic Personality, 55(1) J. of Abnormal & Social
Psychology 6.
Lynam, D. R., A. Caspi, T. E. Moffitt, R. Loeber, & M. Stouthamer-Loeber (2007) Longitudinal
Evidence that Psychopathy Scores in Early Adolescence Predict Adult Psychopathy, 116 J. of
Abnormal Psychology 155.
Malenka, R., R. W. Angel, B. Hampton, & P. A. Berger (1982) Impaired Central Error-Correcting
Behavior in Schizophrenia, 39 Archives of General Psychiatry 101.
Mayer, R. E. (2009) Multimedia Learning, 2nd ed. New York: Cambridge Univ. Press.
McCabe, D. P., & A. D. Castel (2008) Seeing is Believing: The Effect of Brain Images on Judgments of
Scientific Reasoning, 107 Cognition 343.
McCabe, D. P., A. D. Castel, & M. G. Rhodes (2011) The Influence of fMRI Lie Detection Evidence on
Juror Decision Making, 577 Behavioral Sciences & the Law 566.
Miller, G. (2010) Brain Exam May Have Swayed Jury in Sentencing Convicted Murderer, December
14 Science Insider, available at http://news.sciencemag.org/scienceinsider/2010/12/brain
-exam-may-have-swayed-jury.html.
Morse, S. J. (2008) Psychopathy and Criminal Responsibility, 1 Neuroethics 205.

The Impact of Neuroimages on Sentencing

129

Muller, J. L., M. Sommer, K. Dohnel, T. Weber, T. Schmidt-Wilcke, & G. Hajak (2008) Disturbed
Prefrontal and Temporal Brain Function During Emotion and Cognition Interaction in Criminal Psychopathy, 26 Behavioral Sciences & the Law 131.
Muller, J. L., M. Sommer, V. Wagner, K. Lange, H. Taschler, C. H. Roder, G. Schuierer, H. E. Klein, &
G. Hajak (2003) Abnormalities in Emotion Processing Within Cortical and Subcortical Regions
in Criminal Psychopaths: Evidence from a Functional Magnetic Resonance Imaging Study Using
Pictures with Emotional Content, 54 Biological Psychiatry 152.
Newman, J. P. (1987) Reaction to Punishment in Extraverts and Psychopaths: Implications for the
Impulsive Behavior of Disinhibited Individuals, 21 J. of Research in Personality 464.
Newman, J. P., C. M. Patterson, & D. S. Kosson (1987) Response Perseveration in Psychopaths, 96(2)
J. of Abnormal Psychology 145.
Nordahl, T. E., C. S. Carter, R. E. Salo, L. Kraft, J. Baldo, S. Salamat, L. Robertson, & N. Kusubov (2001)
Anterior Cingulate Metabolism Correlates with Stroop Errors in Paranoid Schizophrenia
Patients, 25(1) Neuropsychopharmacology 139.
Patrick C. J., M. M. Bradley, & P. J. Lang (1993) Emotion in the Criminal Psychopath: Startle Reflex
Modulation, 102 J. of Abnormal Psychology 82.
Phillips, M. L., L. Williams, C. Senior, E. T. Bullmore, M. J. Brammer, C. Andrew, S. C. Williams, & A.
S. David (1999) A Differential Neural Response to Threatening and Non-Threatening Negative
Facial Expressions in Paranoid and Non-Paranoid Schizophrenics, 92 Psychiatry ResearchNeuroimaging 11.
Poldrack, R. A., P. C. Fletcher, R. N. Henson, K. J. Worsley, M. Brett, & T. E. Nichols (2008) Guidelines
for Reporting an fMRI Study, 40 NeuroImage 409.
Pratt, B. (2005) Soft Science in the Courtroom?: The Effects of Admitting Neuroimaging Evidence
into Legal Proceedings, 1 Penn Bioethics J. 1.
Racine, E., O. Bar-Ilan, & J. Illes (2005) fMRI in the Public Eye, 6(2) Nature Reviews Neuroscience
159.
Rendell, J. A., M. T. Huss, & M .L. Jensen (2010) Expert Testimony and the Effects of a Biological
Approach, Psychopathy, and Juror Attitudes in Cases of Insanity, 28 Behavioral Science & Law
411.
Rice, M. E., & G. T. Harris (1990) Predictors of Insanity Acquittal, 13 International J. of Law &
Psychiatry 217.
Roberts, C. F., S. L. Golding, & F. D. Fincham (1987) Implicit Theories of Criminal Responsibility
Decision Making and the Insanity Defense, 11 Law & Human Behavior 207.
Rose, N. (2000) The Biology of Culpability: Pathological Identity and Crime Control in a Biological
Culture, 4 Theoretical Criminology 5.
Roskies, A. L. (2007) Are Neuroimages Like Photographs of the Brain? 74(5) Philosophy of Science 860.
(2008) Neuroimaging and Inferential Distance, 1(1) Neuroethics 19.
Schachter, S., & B. Latan (1964) Crime, Cognition, and the Autonomic Nervous System, in D.
Levine, ed., Nebraska Symposium on Motivation, pp. 22173. Univ. of Nebraska Press.
Schmauk, F. J. (1970) Punishment, Arousal, and Avoidance Learning in Sociopaths, 76(3) J. of
Abnormal Psychology 325.
Schneider, D., M. DeBell, & J. A. Krosnick (2010) Using the American National Election Study Surveys
to Test Social Psychological Hypotheses, in K. Trzesniewski, M. B. Donnellan, & R. E. Lucas,
eds., Obtaining and Analyzing Archival Data: Methods and Illustrations. Washington, DC: American
Psychological Association.
Schweitzer, N. J., & M. J. Saks (2011) Neuroimage Evidence and the Insanity Defense, 29 Behavioral
Sciences & the Law 592.
Schweitzer, N. J., M. J. Saks, E. R. Murphy, A .L. Roskies, W. Sinnott-Armstrong, & L. M. Gaudet (2011)
Neuroimages as Evidence in a Mens Rea Defense: No Impact, 17 Psychology, Public Policy, & Law
357.
Siegel, R. A. (1978) Probability of Punishment and Suppression of Behavior in Psychopathic and
Nonpsychopathic Offenders, 87(5) J. of Abnormal Psychology 514.

130

Saks et al.

Silver, H., & C. Goodman (2007) Impairment in Error Monitoring Predicts Poor Executive Function
in Schizophrenia Patients, 94 Schizophrenia Research 156.
Society of Nuclear Medicine Brain Imaging Council (1996) Ethical Clinical Practice of Functional
Brain Imaging, 37 J. of Nuclear Medicine 1256.
Tiihonen, J., R. Rossic, M. Laakso, S. Hodgins, C. Testa, J. Perez, E. Repo-Tiihonen, O. Vaurio, H.
Soininen, H. Aronen, M. Knene, P. Thompson, & G. Frisoni. (2000) Amygdaloid Volume
Loss in Psychopathy, Society for Neuroscience Abstracts 2017.
Trout, J. D. (2008) Seduction Without Cause: Uncovering Explanatory Neurophilia, 12 Trends in
Cognitive Sciences 281.
Turken, A. U., P. Vuilleumier, D. H. Mathalon, D. Swick, & J. M. Ford (2003) Are Impairments of
Action Monitoring and Executive Control True Dissociative Dysfunctions in Patients with
Schizophrenia? 160 American J. of Psychiatry 1881.
Uttal, W. R. (2003). The New Phrenology: The Limits of Localizing Cognitive Processes in the Brain. Cambridge,
MA: MIT Press.
(2008). Neuroscience in the Courtroom: What Every Lawyer Should Know About the Mind and the Brain.
Danvers, MA: Lawyers & Judges Publishing.
van Honk, J., E. J. Hermans, P. Putman, B. Montagne, & D. J. L. G. Schutter (2002) Defective Somatic
Markers in Sub-Clinical Psychopathy, 13(8) Cognitive Neuroscience & Neuropsychology 1025.
Viding, E., & U. Frith (2006) Genes for Susceptibility to Violence Lurk in the Brain, 103(16)
Proceedings of the National Academic of Sciences 6085.
Viljoen, J. L., E. A. M. MacDougall, N. C. Gagnon, & K. S. Douglas (2010) Psychopathy Evidence in
Legal Proceedings Involving Adolescent Offenders, 16(3) Psychology, Public Policy, & Law 254.
Vul, E., C. Harris, P. Winkielman, & H. Pashler (2009) Puzzlingly High Correlations in fMRI Studies
of Emotion, Personality, and Social Cognition, 4 Perspectives on Psychological Science 274.
Vul, E., & N. Kanwisher (2010). Begging the Question: The Non-Independence Error in fMRI Data
Analysis, in S. J. Hanson & M. Bunzl, eds., Foundational Issues in Human Brain Mapping.
Cambridge, MA: MIT Press.
Wallace, C., P. E. Mullen, & P. Burgess (2004) Criminal Offending in Schizophrenia Over a 25-Year
Period Marked by Deinstitutionalization and Increasing Prevalence of Comorbid Substance Use
Disorders, 161 American J. of Psychiatry 716.
Weisberg, D. S., F. C. Keil, J. Goodstein, E. Rawson, & J. R. Gray (2008) The Seductive Allure of
Neuroscience Explanations, 20(3) J. of Cognitive Neuroscience 470.
Weiss, E. M., C. G. Kohler, K. A. Nolan, P. Czobor, J. Volavka, M. M. Platt, C. Brensinger, J. Loughead,
M. Delazer, R. E. Gur, & R. C. Gur (2006) The Relationship Between History of Violent and
Criminal Behavior and Recognition of Facial Expression of Emotions in Men with Schizophrenia and Schizoaffective Disorder, 32 Aggressive Behavior 187.
Wessely, S. (1998) The Camberwell Study of Crime and Schizophrenia, 33 Social Psychiatry &
Psychiatric Epidemiology S24.
Wright, I. C., S. Rabe-Hesketh, P. W. Woodruff, A. S. David, R. M. Murray, & E. T. Bullmore (2000)
Meta-Analysis of Regional Brain Volumes in Schizophrenia, 157 American J. of Psychiatry 16.

Appendix: Sample Case Materials (Neurological

Evidence of Psychopathy)
The defendants parents testified that Mr. Haggard was a difficult child from an early age.
He seemed unable to learn from their guidance or discipline. At the age of 5, he was
removed from school for fighting, showed no emotional response to punishment, and he
had what was described by doctors as an emotional abnormality. At the age of 11, he
started engaging in criminal activities beginning with shoplifting, drug use, and simple

The Impact of Neuroimages on Sentencing

131

assault, and in adolescence, escalating to check fraud, unlawful confinement, and sexual
assault. His parents described him as unruly and making no effort to maintain relationships
with others. They made numerous attempts to provide him with psychological counseling,
but he would not communicate with the therapist.
He was held in juvenile detention at age 15, where he was diagnosed with severe
Conduct Disorder. Reports at the juvenile detention center described his temperament as
superficial, manipulative, withdrawn, impulsive, and aggressive.
During his late teens and early adulthood, mental health evaluations indicated that
Mr. Haggard was regarded as having the following personality characteristics: impulsive and
withdrawn; makes up false stories; shows reckless behavior toward friends, work, school, and
family; makes irresponsible decisions in finances, relationships, and work; frequently
engages in drugs, alcohol, and antisocial behavior, and is emotionally disturbed, showing
no concern with the impact his behavior has on others. He seemed to lack any guilt or
genuine remorse.
The defense presented the expert testimony of two independent neuroscientists.
These experts examined the defendant by conducting fMRI (functional magnetic resonance imaging) brain scans on the defendant, and they both concluded that the defendant
meets the necessary criteria to be considered as having the personality disorder of psychopathy. Scientists have recently shown that individuals with psychopathy develop abnormally, especially emotionally, from an early age. These emotional abnormalities lead to
problems forming normal relationships with family and friends, and are often associated
with severe impulsivity. All combined, the traits seem to cause impairments in the afflicted
individual in attaining normal educational outcomes, maintaining jobs or romantic relationships, and there is a strong increased tendency towards criminal behavior.
In sum, the prosecutor argued that the nature of the defendants crimes presented
more than sufficient evidence of aggravating factors to outweigh any mitigating evidence
and justify sentencing him to death. The defense counsel argued that evidence of Mr.
Haggards childhood and adolescent behavior shows that he suffered from some kind of
emotional dysfunction which made it impossible for him to conform his behavior to
societys requirements. In addition, the expert testimony showed that the defendant suffers
from an illness or dysfunction which renders him incapable of controlling his criminal
behavior. Because this condition is not his fault, the defense argues that justifies sparing
him from death and sentencing him to life in prison.

Copyright of Journal of Empirical Legal Studies is the property of Wiley-Blackwell and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.