Law and Human Behavior
2012, Vol. 36, No. 4, 275–282
© 2011 American Psychological Association
0147-7307/11/$12.00 DOI: 10.1037/h0093972
You Have the Right to Understand: The Deleterious Effect of Stress on
Suspects’ Ability to Comprehend Miranda
Kyle C. Scherr and Stephanie Madon
Iowa State University
Miranda v. Arizona (384 U.S. 436, 1966) required that suspects be explicitly warned of the right to avoid
self-incrimination and the right to legal representation. This research was designed to examine whether
stress, induced via an accusation of wrong-doing, undermined or enhanced suspects’ ability to comprehend their Miranda rights. Participants were randomly assigned to either be accused (n ⫽ 15) or not
accused (n ⫽ 15) of having cheated on an experimental task in a two-cell between-subjects experimental
design. Results supported the hypothesis that stress undermines suspects’ ability to comprehend their
Miranda rights. Participants who were accused of cheating exhibited significantly lower levels of
Miranda comprehension than participants who were not accused of cheating. The theoretical processes
responsible for these effects and the implications of the findings for police interrogation are discussed.
Keywords: Miranda rights, comprehension, stress, accusation
The Supreme Court set a landmark criminal procedure precedent in the Miranda v. Arizona (1966) ruling. The Miranda ruling,
based on the Fifth Amendment right against compelled selfincrimination and the Sixth Amendment right to counsel, was
designed as a prophylactic safeguard against police intimidation
during custodial questioning. However, debate over Miranda continues even 40 years after the ruling (Leo & Thomas, 1998; White,
2001). Central to this debate is concern that Miranda may not
afford suspects the protections that were initially intended by the
Supreme Court. For example, whereas many researchers have
documented that adults of average intelligence have a satisfactory
understanding of Miranda, these findings have been reported with
the caveat that assessments of Miranda comprehension have most
often been examined in low stress situations (e.g., Grisso, 1998).
This fact has led social scientists and legal scholars to hypothesize
that stress caused by police accusation (Gudjonsson, 2003; Irving,
1980) may undermine suspects’ ability to fully understand their
constitutional rights (Kassin et al., 2007; Oberlander & Goldstein,
2001; Rogers, Gillard, Wooley, & Fiduccia, 2010).
The potentially deleterious effect of stress on Miranda comprehension has profound implications for the judicial system because
it raises the possibility that Miranda may not be fully affording
suspects the prophylactic safeguards it was designed to provide. In
the current research, therefore, we experimentally examined how
stress—induced via an accusation of wrong-doing—affected Miranda comprehension. We begin by discussing research relevant to
Miranda comprehension after which we present the theoretical and
empirical basis underlying our research.
Miranda Comprehension Research
Several key factors have been empirically demonstrated to influence Miranda comprehension. One such factor is reading difficulty. There are considerable differences in the reading difficulty
of Miranda warnings across jurisdictions and these differences
influence suspects’ comprehension (Rogers, 2008a). For example,
recently detained adults showed significantly better comprehension of Miranda warnings when the warnings were written at 6th
grade level than when they were written between an 8th and 12th
grade level (Rogers, 2008b). The method by which Miranda
warnings are administered also influences suspects’ comprehension. Suspects demonstrate better comprehension when Miranda
warnings are administered in a written fashion than an oral fashion
(Rogers, 2008b). It is also clear that cognitive deficits associated
with mental retardation and substance use (Hazelwood, 2009)
substantially undermine suspects’ ability to comprehend Miranda
(e.g., Clare & Gudjonsson, 1991; Everington & Fulero, 1999;
O’Connell, Garmoe, & Goldstein, 2005). Moreover, suspects with
mental health conditions have been found to have a reduced ability
to comprehend Miranda compared to individuals without mental
health conditions (Cooper & Zapf, 2008).
An additional factor that has recently been demonstrated to
affect Miranda comprehension is stress. Rogers and colleagues
(2010) examined the influence of situational factors on individuals’ ability to comprehend Miranda, including stress evoked by an
accusation of having committed a mock crime. In the study,
participants either were or were not accused of having stolen a
watch. Although participants were informed ahead of time that
they might have to pretend to steal the watch, be apprehended, be
read their Miranda rights, and have their understanding of these
rights assessed, participants who were accused of stealing the
watch nonetheless reported feeling more stress and demonstrated
worse recall and reasoning on Miranda comprehension instruments than did participants who were not accused of stealing the
watch.
This article was published Online First April 26, 2011.
Kyle C. Scherr and Stephanie Madon, Department of Psychology, Iowa
State University, Iowa.
Correspondence concerning this article should be addressed to Kyle C.
Scherr, Department of Psychology, Iowa State University, W112 Lagomarcino Hall, Ames, IA 50011-3180, USA. E-mail: [email protected]
275
276
The findings of this study are important because they provide
initial evidence in support of the idea that stress can have a
disadvantageous influence on suspects’ ability to understand their
Miranda rights. However, because these findings emerged with
respect to an accusation that pertained to a mock crime, additional
evidence arising from research that involves an accusation of
actual wrong-doing is needed to strengthen the claim that stress
undermines suspects’ Miranda comprehension. This is particularly
important given that an accusation of actual wrong-doing would
likely elicit more stress than an accusation of having committed a
mock crime and stress has previously been shown to have a
complex relation to cognitive functioning, sometimes undermining
it and other times enhancing it.
Stress and Cognitive Functioning
Several lines of research support the idea that stress can undermine cognitive functioning. According to processing efficiency
theory (Eysenck, 1982, 1992, 1997), for instance, stress expends
valuable cognitive resources that would otherwise be used for
efficient cognitive functioning, and by so doing, compromises the
efficiency of individuals’ working memory system (e.g., Derakshan & Eysenck, 1998; MacLeod & Donnellan, 1993). Empirical
findings strongly support this prediction. A meta-analysis found
that high levels of stress negatively impacted both the accuracy of
eyewitness identifications and the recall accuracy of crime-related
details (Deffenbacher, Bornstein, Penrod, & McGorty, 2004).
Likewise, high levels of anxiety undermined eyewitness memory
performance (e.g., Bothwell, Brigham, & Pigott, 1987; Deffenbacher, 1994; Peters, 1988) and, as mentioned above, stress, induced via a mock crime, reduced individuals’ ability to comprehend their Miranda rights (Rogers et al., 2010). Stress has also
been shown to impair memory (Eichenbaum, Otto, & Cohen,
1992; LeDoux, 1995; Newcomer, Craft, Hershey, Askins, &
Bardgett, 1994), compromise the processing of inaccessible information (Bargh & Thein, 1985; Pratto & Oliver, 1991), undermine
performance (Lupien et al., 2005; MacKenzie, Smith, Hasher,
Leach, & Behl, 2007; Sand-strom, Rhodin, Lundberg, Olsson, &
Nyberg, 2005) and lead to a loss of distance cues in perception
(Callaway & Thompson, 1953). Overall, therefore, there is myriad
evidence demonstrating that stress can impair cognitive functioning, thus generating the hypothesis that stress caused by police
accusation compromises suspects’ ability to comprehend Miranda.
However, it is also possible that stress may enhance Miranda
comprehension. Theory and research dating back to the 1950s
indicates that negative emotions—including stress and anxiety—
cause people to narrow their attentional focus to central and
relevant information. For example, stress causes people to focus
more on cues that are central to the visual field than to cues that are
peripheral to it (Chagut & Algom, 2003; Easterbrook, 1959).
People experiencing anxiety better remember strongly related constructs than weakly related constructs (Derryberry & Tucker,
1994). And, anxiety, caused by threat-relevant words, leads people
to focus more on threat-relevant information than threat-irrelevant
information (Watts, McKenna, Sharrock, & Trezise, 1986). Accordingly, it is possible that the stress and anxiety that suspects
experience as a result of police accusation could have an advantageous influence on their ability to comprehend their Miranda
rights. Suspects could, for example, focus acutely on the content of
Miranda warnings, thereby enhancing their understanding of them.
To the extent that stress improves cognitive functioning, then the
alternative hypothesis—that stress enhances Miranda comprehension—is generated.
Research Overview
The current research examined two competing predictions regarding the effect of stress on Miranda comprehension. Consistent
with theory and research indicating that stress reduces cognitive
functioning (e.g., Grisso, 1998; Kassin et al., 2007; Oberlander &
Goldstein, 2001), we hypothesized that stress undermines Miranda
comprehension. Alternatively, based on theory and research indicating that stress can improve cognitive functioning by virtue of
causing individuals to narrow their attentional focus (e.g., Chagut
& Algom, 2003; Derryberry & Tucker, 1994; Easterbrook, 1959),
we hypothesized that stress enhances Miranda comprehension. We
tested these competing hypotheses with an experiment in which we
manipulated participants’ stress with an accusation of actual
wrong-doing. The procedures we employed were adapted from a
paradigm developed by Russano, Meissner, Narchet, and Kassin
(2005). In the Russano et al. (2005) paradigm, participants are
paired with a confederate with whom they complete logic problems. The experimenter instructs the pair to work independently on
some problems and together on others. The confederate asks some
participants (but not others) for help on a problem the pair was
instructed to work on alone. Shortly after, the pair is separated and
the participant is accused of cheating. In the current research, we
modified this paradigm such that all participants were asked for
help on a problem, but only half of the participants were accused
of cheating. The manipulation of the accusation was intended to
influence participants’ stress levels such that participants who
were accused of cheating would feel more stress than those who
were not accused of cheating.
Method
Participants
Thirty-two college students (16 women, 16 men) participated in
exchange for course credit. Participants ranged in age from 18 to
24 (M ⫽ 19.32). The sample included 29 Caucasians, 1 AfricanAmerican, and 2 participants who were multi-ethnic. All participants were native English speakers. Two participants were removed from the sample due to suspicion causing the final sample
to consist of 30 participants, 15 in each condition.
Experimental Design
Participants were randomly assigned to either be accused (n ⫽
15) or not accused (n ⫽ 15) of having cheated on an experimental
task in a two-cell between-subjects experimental design.
Measures
Miranda comprehension was assessed with Grisso’s (1998)
Instruments for Assessing Understanding and Appreciation of Miranda Rights, which includes four instruments. The four instruments are the Comprehension of Miranda Rights (CMR), Com-
277
prehension of Miranda Rights-Recognition (CMR-R),
Comprehension of Miranda Vocabulary (CMV), and Function of
Rights in Interrogation (FRI). Each of these instruments is described in more detail next.
Comprehension of Miranda Rights.
The CMR assesses
participants’ ability to understand the basic meaning of each of the
Miranda warning statements. The CMR requires that participants
paraphrase each of the four Miranda warnings (e.g., “You have the
right to remain silent”). If participants give initial answers that are
unsatisfactory, follow-up questions can be asked to give participants the opportunity to display an adequate understanding of the
statement. Responses for each of the four statements can be adequate, intermediate, or inadequate, with scores of 2, 1, or 0,
respectively. Thus, total scores on the CMR range from 0 to 8.
Comprehension of Miranda Rights-Recognition.
The
CMR-R assesses participants’ comprehension of their Miranda
rights with minimal need to construct a verbal response. The
CMR-R requires that participants identify a differently worded
statement as being the same or not the same as an original statement. For example, one of the four Miranda statements is read and
is followed by three other statements that either have the same
meaning or have a different meaning than the first statement. Thus,
for each of the four Miranda statements, there are three subsequent
statements that the participant has to respond to by saying the
subsequent statement either has a similar or a not similar meaning
than the first Miranda warning statement. Participants’ responses
(i.e., same or different) are then recorded. For each response,
participants can get a score of 1 or 0. Thus, for all four statements
and the three accompanying statements under each, scores can
range from 0 to 12.
Comprehension of Miranda Vocabulary.
The CMV assesses participants’ ability to correctly define key terms within
Miranda (e.g., entitled, attorney, etc.). Participants are read six
words and are asked to correctly define each one. Responses are
scored as being adequate, intermediate, or inadequate with scores
of 2, 1, or 0, respectively. Thus, total scores on the CMV range
from 0 to 12.
Function of Rights in interrogation. The FRI assesses various beliefs about the consequences of Miranda during the interrogation process and a subsequent trial. The FRI is composed of
three subscales. The Nature of Interrogation subscale assesses
participants’ understanding of the antagonistic relationship between interrogators and suspects. The Right to Counsel subscale
assesses the supportive relationship between an attorney and a
client. The Right to Silence subscale assesses participants’ ability
to appreciate the fact that, regardless of the authority, individuals
have a right to remain silent. Participants are shown four pictures
of hypothetical situations and asked a total of 15 questions regarding each picture— each question pertains to only one of the subscales. Participants’ responses are scored as being adequate, intermediate, or inadequate with scores of 2, 1, or 0, respectively.
Scores for each subscale range from 0 to 10 and scores on the total
FRI instrument range from 0 to 30.
American College Test. Prior investigations have demonstrated that intelligence can have a significant influence on individuals’ comprehension of Miranda (e.g., Clare & Gudjonsson,
1991; O’Connell, Garmoe, & Goldstein, 2005). To account for this
potential influence among our participants, we assessed American
College Test (ACT) scores, which have been shown to correlate
significantly with standard measures of intelligence (e.g., California test, Otis-Lennon, Lorge-Thorndike, Henmon-Nelson; Koenig,
Frey, & Detterman, 2008). Participants indicated their ACT scores
in response to an open-ended item that asked them to report their
scores as accurately as possible.
Demographic information.
Participants provided demographic information pertaining to their age, gender, and race.
Manipulation check. We assessed the effectiveness of the
accusation manipulation with pilot participants (N ⫽ 40). Following the same procedures of the main experiment, pilot participants
either were (n ⫽ 18) or were not (n ⫽ 22) accused of cheating. In
addition, shortly after the accusation manipulation, pilot participants reported how stressed, worried, anxious, nervous, concerned,
tense, and scared they felt on rating scales with anchors 1 (Not at
all) to 7 (Extremely). Pilot participants’ responses to these seven
questions were averaged to create one variable per participant with
higher values indicating greater stress, ␣ ⫽ .97.
Procedures
Participants were run individually. Upon arrival, a male
experimenter greeted the participant and asked her/him to be
seated at a counter with a confederate posing as another participant
in the study. The pair was then given a consent document to read
and sign. The experimenter then explained to the pair that the study
was designed to examine the influence of individual differences on
people’s ability to make decisions alone and with a partner. The
participant then reported their demographic information and ACT
scores individually on a computer while the confederate appeared
to do so as well in a separate room. After providing this information, the pair was brought into a different room and given 3 min to
get acquainted with one another. This was done to increase the
likelihood that the participant would provide an answer when
asked for help by the confederate. After the pair had a chance to
become acquainted, the experimenter told them that the next phase
of the study involved the pair completing some logic problems.
The pair was given three logic problem packets, two individual
packets and one team packet obtained from Russano et al. (2005).
The pair was told that they needed to work alone on the individual
packets, but that they needed to discuss their answers and strategies when completing the team packet. The experimenter then left
the room and closed the door.
While the pair was completing the logic problems, the confederate asked the participant for help on the final individual logic
problem after which the pair completed the final team logic problem. After the pair had completed all of the logic problems, they
informed the experimenter that they were done. The experimenter
then gave the pair a filler survey to complete while he ostensibly
corrected the logic problem packets. After the pair had completed
the survey, the experimental manipulation occurred.
In the accusation condition, the experimenter returned to the
pair’s room, looking irritated, and told the pair that he had to check
on something. After a minute, the experimenter informed the pair
that there might be a problem and that he would need to talk to
each of them separately. The experimenter then escorted the confederate out of the room and told the participant he would be back
in a moment to talk to her/him. After 5 min had passed, the
experimenter returned to the participant and explained that he had
talked with the other participant and there was indeed a problem.
278
The experimenter went on to explain that he was suspicious that
the two had shared answers on one of the individual logic problems because they both had the same wrong answer. The experimenter said that he was not sure how he should handle this
situation, so he called his advisor and he could tell that his advisor
was annoyed. The experimenter further explained that he was not
sure what his advisor would do or how he would handle the
situation and that the advisor may even consider this a case of
cheating. The experimenter said that the advisor wanted to talk to
the participant about this situation himself, but that he was in a
meeting and would come to the lab in about 10 min. In the
meantime, the participant was supposed to finish up the study. In
the no accusation condition, the experimenter simply told the
confederate to go back to their original room and that he would be
with her/him shortly. Following the experimental manipulation, all
participants were informed that the last part of the study was
designed to assess verbal and creative individual differences. Participants were then individually administered Instruments for Assessing Understanding and Appreciation of Miranda Rights
(Grisso, 1998). Their responses were audio recorded for scoring
purposes.
After the experimenter finished assessing participants’ comprehension of Miranda, participants were probed for suspicion. Participants were then fully debriefed, with extra care being taken to
assure accused participants that they were not in any trouble. After
all questions regarding the study were answered, participants were
thanked and dismissed.
correlations between participants’ ACT scores and their scores on
the Miranda comprehension instruments. These correlations indicated that participants’ ACT scores evidenced marginally significant correlations of moderate magnitudes with three of the four
comprehension instruments, i.e., the CMR (r ⫽ .31, p ⫽ .09), the
CMR-R (r ⫽ .27, p ⫽ .16), and the CMV (r ⫽ .31, p ⫽ .10).1
Accordingly, we included participants’ ACT scores as a covariate
in all of the main analyses.
Gender. To test for gender differences in participants’ Miranda comprehension scores, we performed a multivariate analysis
of covariance (MANCOVA). Gender was included as the subject
variable and participants’ ACT scores were included as a covariate. The dependent variables were participants’ scores on the four
instruments and the three subscales of the FRI used to assess
Miranda comprehension. The results indicated that there were no
significant differences between men’s and women’s responses, F
(7, 21) ⫽ 1.05, p ⫽ .426. Therefore, we did not include participants’ gender as a variable in any of the subsequent analyses.
Manipulation check.
An independent samples t-test was
used to examine the effectiveness of the accusation manipulation.
We examined whether pilot participants’ self-reported levels of
stress differed across the accusation and no accusation conditions.
The results indicated that pilot participants who were accused of
cheating reported significantly higher levels of stress (M ⫽ 3.75)
than pilot participants who were not accused of cheating (M ⫽
2.48), t(38) ⫽ 3.04, p ⫽ .004, 95% CI: .43; 2.12, d ⫽ .99. This
result provides support for the effectiveness of the accusation
manipulation.
Scoring of Miranda Comprehension
Using standardized scoring procedures (Grisso, 1998), three
coders, blind to condition, independently coded participants’ audio
recorded responses to each of the Instruments for Assessing Understanding and Appreciation of Miranda Rights except for the
CMR-R which is scored in a purely objective manner.
Results
Preliminary Analyses
Interrater agreement for Miranda comprehension instruments. Interrater agreement among the three coders with respect
to the Miranda comprehension instruments was examined with an
intraclass correlation (Suen & Ary, 1989). The results indicated
that the three coders had a good level of agreement, ICC ⫽ .71
(Cicchetti, 1994). Therefore, the coded responses were averaged to
yield one Miranda comprehension score for each of the instruments and subscales used in the main analyses.
ACT scores. A one-way analysis of variance (ANOVA) was
used to examine if there were differences in participants’ ACT
scores across the two accusation conditions. The independent
variable for this analysis was the accusation manipulation. The
dependent variable was participants’ ACT scores. Results indicated that participants’ ACT scores did not differ across the
accusation conditions (MAccused ⫽ 24.50; MNot Accused ⫽ 24.87),
F(1, 29) ⫽ .08, p ⫽ .78. Furthermore, an independent samples
t-test indicated that there were no significant differences between
men’s (M ⫽ 24.46) and women’s (M ⫽ 24.85) ACT scores,
t(28) ⫽ .31, p ⫽ .76. We also conducted several zero-order
Main Analyses
A MANCOVA, followed by separate analyses of covariance
(ANCOVAs), were conducted to test the competing hypotheses
regarding the effect of stress on Miranda comprehension. The
independent variable for these analyses was the accusation manipulation. Participants’ ACT scores were included as a covariate. For
the MANCOVA, the dependent variables were participants’ coded
responses to the four individual instruments and the three subscales of the FRI instrument of the Instruments for Assessing
Understanding and Appreciation of Miranda Rights (Grisso,
1998). Results of the MANCOVA analysis indicated that participants in the accusation condition had significantly lower Miranda
comprehension scores than participants in the no accusation condition, F(7, 21) ⫽ 3.27, p ⫽ .016.
Next, therefore, we conducted seven separate ANCO VAs to
examine the effect that the accusation had on participants’ ability
to comprehend Miranda as assessed by each individual instrument
and the three subscales of the FRI. Results indicated that the
accusation of wrong-doing undermined Miranda comprehension.
In comparison to participants in the no accusation condition,
participants in the accusation condition scored significantly lower
on the CMR, F(1, 27) ⫽ 8.75, p ⫽ .006, 95% CI: ⫺2.86; ⫺.52,
1
We also conducted the seven individual main analyses without participants’ ACT scores using seven ANOVAs. Results remained the same in
terms of trends and significance with two exceptions. The CMR-R and the
RC subscale of the FRI instrument that were significant when ACT scores
were included as a covariate (ps ⫽ .04) became marginally significant
when ACT scores were excluded from the analyses, ps ⫽ .07.
279
d ⫽ 1.14 (MAccused ⫽ 4.72; MNot Accused ⫽ 6.41), the CMR-R,
F(1, 27) ⫽ 4.74, p ⫽ .038, 95% CI: ⫺1.94; ⫺.06, d ⫽ .84
(MAccused ⫽ 9.27; MNot Accused ⫽ 10.27), the CMV, F(1, 27) ⫽
8.31, p ⫽ .008, 95% CI: ⫺1.59; ⫺.27, d ⫽ 1.11 (MAccused ⫽ 9.43;
MNot Accused ⫽ 10.35), and the FRI, F (1, 27) ⫽ 14.44, p ⫽ .001,
95% CI: ⫺3.85; ⫺1.15, d ⫽ 1.46 (M Accused ⫽ 22.34;
MNot Accused ⫽ 24.84). Furthermore, results indicated that the
accusation manipulation significantly influenced participants’ ability to comprehend two out of the three subscales of the FRI
instrument. Specifically, participants in the accusation condition
scored significantly lower than participants in the no accusation
condition on the RC, F(1, 27) ⫽ 4.42, p ⫽ .04, 95% CI: ⫺1.48;
⫺.02, d ⫽ .81 and RS subscales, Fs(1, 27) ⫽ 12.41, p ⫽ .002, 95%
CI: ⫺2.30; ⫺.61, d ⫽ 1.36 (Table 1; Figure 1). Although the NI
subscale failed to achieve significance, the results were in the
expected direction (MAccused ⫽ 8.67; MNot Accused ⫽ 9.07). Results
did not support the competing hypothesis that stress enhances
suspects’ Miranda comprehension.
Discussion
The Miranda ruling was designed to provide a safeguard for
suspects against police intimidation. However, social scientists and
legal scholars have long hypothesized that stress, caused by police
accusation, may undermine suspects’ ability to fully understand
their constitutional rights as described in Miranda warnings. Our
research provided empirical support for this important hypothesis.
Consistent with a large body of research showing that stress
adversely affects cognitive functioning, our results indicated that
stress, induced through an accusation of wrong-doing, significantly undermined participants’ comprehension of Miranda. Specifically, participants who were accused of wrong-doing demonstrated significantly lower Miranda comprehension scores than
participants who were not accused of wrong-doing.
Table 1
Means, Standard Deviations in Parentheses, F Values, and
Effect Sizes of Instruments for Assessing Understanding and
Appreciation of Miranda Rights for Both Conditions
Instrument
Accused
(n ⫽ 15)
Not accused
(n ⫽ 15)
F
d
CMR
CMR-R
CMV
FRI
NI
RC
RS
4.72 (1.63)
9.27 (1.29)
9.43 (0.86)
22.34 (2.20)
8.67 (0.93)
8.36 (1.09)
5.34 (1.23)
6.41 (1.42)
10.27 (1.26)
10.35 (0.85)
24.84 (1.17)
9.07 (0.92)
9.11 (0.86)
6.80 (1.00)
8.74ⴱⴱ
4.74ⴱ
8.31ⴱⴱ
14.44ⴱⴱⴱ
1.5
4.42ⴱⴱ
12.41ⴱⴱⴱ
1.14
.84
1.11
1.46
.47
.80
1.36
Note. Values under each condition reflect the average comprehension
score for each instrument or subscale with higher scores indicating better
comprehension. The following abbreviations are followed by their corresponding instrument or subscale: CMR ⫽ Comprehension of Miranda
Rights instrument; CMR-R ⫽ Comprehension of Miranda RightsRecognition instrument; CMV ⫽ Comprehension of Miranda Vocabulary
instrument; FRI ⫽ Function of Rights in Interrogation instrument; NI ⫽
Nature of Interrogation subscale; RC ⫽ Right to Counsel subscale; RS ⫽
Right to Silence subscale.
ⴱ
p ⬍ .05. ⴱⴱ p ⬍ . 01. ⴱⴱⴱ p ⬍ .001.
Magnitude of Effects
By using an experimental design, the current research was able
to infer a causal relationship between the accusation manipulation
and Miranda comprehension. Moreover, because the pilot study
demonstrated that accused participants experienced more stress
than participants who were not accused, there is reason to believe
that the accusation manipulation influenced participants’ Miranda
comprehension scores, at least in part, because it affected their
stress levels. However, because this research was conducted as a
lab experiment, it suffers from the potential lack of generalizability. Indeed, it would have been neither ethical nor feasible to
expose our participants to the same degree of stress that accompanies police accusation. Nonetheless, it is important to understand
what this means with respect to our findings. Because the degree
of stress experienced by our participants was very likely less than
that experienced by suspects accused of serious crimes by police,
the observed results are likely conservative. This possibility is
particularly noteworthy given that for six of the seven analyses we
performed, the accusation manipulation had a large effect on our
participants’ ability to comprehend Miranda, with effect sizes of d
C .80 (Cohen, 1988). Therefore, it seems likely that the effects that
were observed in these data are conservative estimates of effect
that stress has on Miranda comprehension during police accusation.
Furthermore, the effect sizes observed in the current research
were larger than those observed in previous research that induced
stress via a mock crime (Rogers et al., 2010). This is important to
note because it provides evidence that an accusation of actual
wrong-doing likely elicits more stress and more greatly undermines Miranda comprehension than does an accusation of having
committed a mock crime. Moreover, if actual accusations of
wrong-doing do elicit a higher degree of stress, as our results
suggest, then it would seem reasonable to conclude that the more
stress suspects experience, the less able they will be to fully
comprehend their Miranda rights.
Another way to appreciate the effect that the accusation had on
our participants’ ability to comprehend Miranda is to compare
their comprehension scores to the comprehension scores of other
samples reported in the literature. Such a comparison reveals that
participants who were accused of wrong-doing in our research
demonstrated an average comprehension score that was relatively
equivalent to the average comprehension scores of juveniles
(Grisso, 1998), adults who were diagnosed as psychotic, and adults
who were patients at a psychiatric ward (Cooper & Zapf, 2008).
These comparisons are particularly noteworthy given that we
relied exclusively on college students as participants and college
students are typically assumed to have a higher level of intelligence than the general population. Thus, the detrimental effect that
the accusation had on our participants’ Miranda comprehension
scores was not only large in a statistical sense, but also meaningful
in a practical sense.
Of course, this does not necessarily mean that the comprehension scores of these participants were unacceptably low from a
legal standpoint. It is entirely possible that the degree to which the
accusation undermined these participants’ comprehension scores
was not so large as to consider their comprehension of Miranda to
be unacceptable during trial. But, as we discussed earlier, the
effects that were observed in these data are likely conservative
280
Figure 1. (N ⫽ 30). Average comprehension scores for each condition across the four instruments and the
subscales with higher values reflecting better comprehension of Miranda. The following abbreviations are
followed by their corresponding instrument or subscale and range: CMR ⫽ Comprehension of Miranda Rights
instrument (0 –10); CMR-R ⫽ Comprehension of Miranda Rights-Recognition instrument (0 –12); CMV ⫽
Comprehension of Miranda Vocabulary instrument (0 –12); FRI ⫽ Function of Rights in Interrogation instrument (0 –30); NI ⫽ Nature of Interrogation subscale (0 –10); RC ⫽ Right to Counsel subscale(0 –10); and RS ⫽
Right to Silence subscale (0 –10).
estimates of the effect that stress may have on suspects’ ability to
comprehend Miranda during actual police accusations. In the
naturalistic setting, then, the stress that suspects will likely experience in response to police accusation could reduce their Miranda
comprehension to such a degree that their understanding of Miranda would be considered unacceptable, even from a legal standpoint.
Theoretical Reasons for the Observed Effects
Existing theory suggests that the tendency for stress to have
undermined our participants’ comprehension of Miranda may reflect a compromised working memory system. According to the
processing efficiency theory (Eysenck, 1982, 1983, 1997), individuals under stress make more errors, require more processing
time, and rely on cognitive shortcuts when engaged in problem
solving and decision making. Cognitive efficiency is especially
likely to be compromised when individuals are engaged in a
difficult task (Eysenck, 1997), such as when suspects attempt to
comprehend the legal vernacular of Miranda warnings. Thus, the
reduced comprehension exhibited by our participants may have
occurred because they relied on less effortful and more wellrehearsed cognitive processing. For example, because lawyers are
often characterized in our society as being too expensive for most
citizens (American Bar Association, 2002), suspects may rely on
this heuristic and not realize that Miranda states that suspects have
access to free counsel. Indeed, right to free counsel is one of the
worst comprehended components of Miranda (Rogers, 2008b).
Implications for the Legal System
The effect that the accusation manipulation had on our participants’ comprehension of Miranda has important implications for
the integrity of the criminal justice system as well as for the civil
liberties of suspects. Most notably, it suggests a need for standardization in the administration of Miranda warnings. Indeed, until
very recently Miranda warnings were thought to be relatively
uniform. However, recent investigations of different jurisdictions
across the United States have shown that warnings differ in length,
reading difficulty, and method of administration (Green-field,
Dougherty, Jackson, Podboy, & Zimmermann, 2001; Rogers, Harrison, Shuman, Sewell, & Hazelwood, 2007). Thus, because warnings are not administered in a standardized fashion, it seems
plausible that stress, caused by police accusation, could interact
with various characteristics of Miranda warnings. For example,
stress may undermine Miranda comprehension more strongly
when the warnings are longer versus shorter, presented at a more
difficult versus a less difficult reading level, and presented in an
oral versus a written format. Empirical evidence showing that
stress undermines suspects’ comprehension of the warnings under
these conditions would support reforms to standardize Miranda
warnings in terms of length, reading difficulty, and method of
administration.
Our findings also have implications for Miranda waivers. If it is
the case that stress undermines individuals’ ability to comprehend
Miranda in naturalistic settings, then it is possible that the effects
we observed might also influence Miranda waivers. For example,
people have a tendency to engage in informational social influence
processes whereby they look to others as a way to determine
correct, appropriate, or socially desirable behavior (Cialdini &
Griskevicius, 2010). Applying this tendency to a Miranda administration situation raises the possibility that suspects will be especially susceptible to police tactics used to attain Miranda waivers.
That is, because stress is compromising suspects’ ability to appreciate the significance of Miranda, suspects may decide whether or
not to waive their Miranda rights based on cues given by the
police. If police minimize the importance of Miranda warnings
(Leo, 1996; Simon, 1991), then suspects may assume that waiving
their rights is in their long-term interests even though it is not. This
could be a main reason why four out of every five suspects waive
their Miranda rights (Leo, 1996; Schulhofer, 1996).
281
Limitations and Future Directions
There are two limitations of the current study that warrant
discussion. First, the current research manipulated stress via an
accusation of wrong-doing. Although pilot data demonstrated that
the accusation manipulation significantly influenced participants’
stress, this does not rule out the possibility that the accusation
manipulation had additional effects on participants’ Miranda comprehension scores for reasons that were unrelated to stress. In other
words, it is possible that some characteristic of the accusation that
was not associated with stress was partly responsible for the
differences that we observed in participants’ Miranda comprehension scores. Although our data cannot rule out this possibility,
current theoretical perspectives specifically hypothesize that stress
caused by police accusation undermines suspects’ comprehension
of Miranda rights (Grisso, 1998; Kassin et al., 2007; Oberlander &
Goldstein, 2001), thereby supporting our interpretation that the
accusation manipulation at least partly affected participants’ Miranda comprehension scores because of its effect on their stress
levels.
A second potential limitation of our research pertains to the
perceived relevancy of the Miranda measures. In our study, participants were led to believe that the Miranda comprehension
instruments were a measure of individual differences. Thus, it is
not likely that participants who were accused of wrong-doing
perceived the Miranda comprehension instruments to be relevant
to the accusation. This is an important issue because it raises the
possibility that participants might have demonstrated better comprehension of Miranda had they perceived the assessment instruments as relevant to their current circumstances. Although the
current research cannot directly address this issue, it is worthwhile
to point out that police often attempt to reduce suspects’ perceptions of the relevancy of Miranda warnings. For example, legal
scholars have observed that police will employ various tactics that
make it appear as though the warnings are not particularly relevant
to suspects’ long-term outcomes by minimizing and trivializing the
importance of the warnings and by administering the warnings in
a nonchalant fashion (Leo, 1996; Simon, 1991). Accordingly, the
scenario used in this research closely resembles a situation that
would be observed outside the lab. Nonetheless, the issue of
perceived relevancy needs to be addressed with future research to
draw more definitive conclusions about the effect that perceived
relevancy has on suspects’ ability to comprehend Miranda.
Conclusion
The landmark decision of the Supreme Court in the case of
Miranda v. Arizona (1966) afforded suspects prophylactic safeguards against police intimidation. The current investigation addressed the influence that stress has on suspects’ ability to comprehend and grasp the significance of these warnings. Consistent
with the hypothesis that stress worsens suspects’ ability to comprehend Miranda, participants in our research who experienced
stress in response to having been accused of wrong-doing, had
significantly worse levels of Miranda comprehension than participants who were not under stress as a result of not having been
accused of wrong-doing. This finding supports theoretical speculation that, although adults generally have a good level of understanding of Miranda normal settings, the psychological state of
individuals upon arrest is likely to be very different and have a
disadvantageous influence on their ability to effectively comprehend Miranda (Grisso, 1998; Oberlander & Goldstein, 2001; Rogers et al., 2010). In this way, the current findings raise the possibility that the protections Miranda was designed to afford suspects
may not be realized to the full extent initially intended and that
research needs to identify ways to improve suspects’ Miranda
comprehension to increase the extent to which suspects are afforded protections against police intimidation.
References
American Bar Association. (2002). Public perceptions of lawyers consumer
research findings. Section of Litigation (April). Retrieved April 4, 2011,
from http://apps.americanbar.org/litigation/lawyers/publicperceptions.pdf
Bargh, J., & Thein, R. (1985). Individual construct accessibility, person
memory, and the recall-judgment link: The case of information overload.
Journal of Personality and Social Psychology, 49, 1129 –1146. doi:
10.1037/0022-3514.49.5.1129
Bothwell, R. K., Brigham, J. C., & Pigott, M. A. (1987). An exploratory
study of personality differences in eyewitness memory. Journal of Social
Behavior and Personality, 2, 335–343.
Callaway, E., & Thompson, S. V. (1953). Sympathetic activity and perception: An approach to the relationships between autonomic activity
and personality. Psychosomatic Medicine, 15, 443– 455.
Chagut, E., & Algom, D. (2003). Selective attention improves under stress:
Implications for theories of social cognition. Journal of Personality and
Social Psychology, 85, 231–248. doi:10.1037/0022-3514.85.2.231
Cialdini, R. B., & Griskevicius, V. (2010). Social influence. In R. F.
Baumeister & E. J. Finkel (Eds.), Advanced social psychology: The state
of the science (pp. 384 – 417). New York: Oxford University Press.
Cicchetti, D. V. (1994). Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology.
Psychological Assessment, 6, 284 –290. doi:10.1037/1040-3590.6.4.284
Clare, I., & Gudjonsson, G. (1991). Recall and understanding of the
caution and rights in police detention among persons of average intellectual ability and persons with a mild mental handicap. Issues in
Criminological and Legal Psychology, 1, 31– 42.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences.
New York: Academic Press.
Cooper, V. G., & Zapf, P. A. (2008). Psychiatric patients’ comprehension
of Miranda rights. Law and Human Behavior, 32, 390 – 405. doi:
10.1007/s10979-007-9099-3
Deffenbacher, K. A. (1994). Effects of arousal on everyday memory.
Human Performance, 7, 141–161. doi:10.1207/s15327043hup 0702_3
Deffenbacher, K. A., Borstein, B. H., Penrod, S. D., & McGorty, E. K.
(2004). A meta-analytic review of the effects of high stress on eyewitness memory. Law and Human Behavior, 28, 687–706. doi:10.1007/
s10979-004-0565-x
Derakshan, N., & Eysenck, M. W. (1998). Working memory capacity in
high trait anxious individuals and repressors. Cognition and Emotion,
12, 697–713. doi:10.1080/026999398379501
Derryberry, D., & Tucker, D. M. (1994). Motivating the focus of attention.
In P. M. Niedenthal & S. Kitayama (Eds.), The heart’s eye: Emotional
influences in perception and attention (pp. 167–196). San Diego, US:
Academic.
Easterbrook, J. A. (1959). The effect of emotion on cue utilization and the
organization of behavior. Psychological Review, 74, 16 –28. doi:
10.1037/h0047707
Eichenbaum, H., Otto, T., & Cohen, N. J. (1992). The hippocampus: What
does it do? Behavioral and Neural Biology, 57, 2–36. doi:10.1016/01631047(92)90724-I
Everington, C., & Fulero, S. (1999). Competence to confess: Measuring
understanding and suggestibility of defendants with mental retarda-
282
tion. Mental Retardation, 37, 212–220. doi:10.1352/00476765(1999)037⬍0212:CTCMUA⬎2.0.CO;2
Eysenck, M. W. (1982). Attention and arousal: Cognition and performance. Berlin: Springer.
Eysenck, M. W. (1983). Anxiety and individual differences. In G. R. J.
Hockey (Ed.), Stress and fatigue in human performance. Chichester:
Wiley and Sons.
Eysenck, M. W. (1992). Anxiety: The cognitive perspective. Hove, England: Lawrence Erlbaum Associates.
Eysenck, M. W. (1997). Anxiety and cognition: A unified theory. Hove:
Lawrence Erlbaum Associates.
Greenfield, D. P., Dougherty, E. J., Jackson, R. M., Podboy, J. W., &
Zimmerman, M. L. (2001). Retrospective evaluation of Miranda reading
levels and waiver competency. American Journal of Forensic Psychology, 19, 75– 86.
Grisso, T. (1998). Instruments for assessing understanding and appreciation of Miranda rights. Sarasota, FL: Professional Resource Press.
Gudjonsson, G. H. (2003). The psychology of interrogations and confessions: A handbook. Chichester, England: John Wiley & Sons.
Hazelwood, L. (2009). Deficits in Miranda comprehension and reasoning:
The effects of substance use and attention deficits. Retrieved January 1,
2011, from Dissertations & Theses: A&I (Publication No. AAT
3399422).
Irving, B. (1980). Police interrogation. A case study of current practice.
Research Studies, No. 2. London: HMSO.
Kassin, S. M., Leo, R. A., Meissner, C. A., Richman, K. D., Colwell, L. H.,
Leach, A., et al. (2007). Police interviewing and interrogation: A selfreport survey of police practices and beliefs. Law and Human Behavior,
31, 381– 400. doi:10.1007/s10979-006-9073-5
Koenig, K. A., Frey, M. C., & Detterman, D. K. (2008). ACT and general
cognitive ability. Intelligence, 36, 153–160. doi:10.1016/j.intell
.2007.03.005
LeDoux, J. E. (1995). Emotion: Clues from the brain. Annual Review of
Psychology, 46, 209 –235. doi:10.1146/annurev.psych.46.1.209
Leo, R. A. (1996). Miranda’s revenge: Police interrogation as a confidence
game. Law and Society Review, 30, 259 –288. doi:10.2307/3053960
Leo, R. A., & Thomas, G. C. (1998). The Miranda debate: Law justice, and
policing. Boston: Northeastern University Press.
Lupien, S., Fiocco, A., Wan, N., Maheu, F., Lord, C., Schramek, T., et al.
(2005). Stress hormones and human memory function across the lifespan. Psychoneuroendocrinology, 30, 225–242. doi:10.1016/j.psyneuen
.2004.08.003
Mackenzie, C., Smith, M., Hasher, L., Leach, L., & Behl, P. (2007).
Cognitive functioning under stress: Evidence from informal caregivers
of palliative patients. Journal of Palliative Medicine, 10, 749 –758.
doi:10.1089/jpm.2006.0171
MacLeod, C., & Donnellan, A. M. (1993). Individual differences in anxiety
and the restriction of working memory capacity. Personality and Individual Differences, 15, 163–173. doi: 10.1016/0191-8869(93)90023-V
Miranda v. Arizona. (1966). 384 U.S. 436.
Newcomer, J. W., Craft, S., Hershey, T., Askins, K., & Bardgett, M. E.
(1994). Glucocorticoid induced impairment in declarative memory performance in adult humans. Journal of Neuroscience, 14, 2047–2053.
O’Connell, M. J., Garmoe, W., & Goldstein, N. E. S. (2005). Miranda
comprehension in adults with mental retardation and the effects of
feedback style on suggestibility. Law and Human Behavior, 29, 359 –
369. doi:10.1007/s10979-005-2965-y
Oberlander, L. B., & Goldstein, N. E. (2001). A review and update on the
practice of evaluating Miranda comprehension. Behavioral Sciences and
the Law, 19, 453– 471. doi:10.1002/bsl.453
Peters, D. P. (1988). Eyewitness memory in a natural setting. In M. M.
Gruneberg, P. E. Morris, & R. N. Sykes (Eds.), Practical aspects of
memory: Current research and issues. Vol. 1. Memory in everyday life
(pp. 89 –94). Chichester: Wiley.
Pratto, F., & Oliver, P. (1991). Automatic vigilance: The attentiongrabbing power of negative social information. Journal of Personality
and Social Psychology, 61, 380 –391. doi:10.1037/0022-3514.61.3.380
Rogers, R. (2008a). A little knowledge is a dangerous thing. . .Emerging
Miranda research and professional roles for psychologists. American
Psychologist, 63, 776 –787. doi:10.1037/0003-066X.63.8.776
Rogers, R. (2008b). Advances in the assessment of Miranda. Symposium
presented at the annual conference of the American Psychology-Law
Society, Jacksonville, FL.
Rogers, R., Gillard, N. D., Wooley, C. N., & Fiduccia, C. E. (2010).
Decrements in Miranda abilities: An investigation of situational effects
via a mock-crime paradigm. Law and Human Behavior. doi:10.1007/
s10979-010-9248-y
Rogers, R., Harrison, K. S., Shuman, D. W., Sewell, K. W., & Hazelwood,
L. L. (2007). An analysis of Miranda warnings and waivers: Comprehension and coverage. Law and Human Behavior, 31, 177–192. doi:
10.1007/s10979-006-9054-8
Russano, M. B., Meissner, C. A., Narchet, F. M., & Kassin, S. M. (2005).
Investigating true and false confessions within a novel experimental
paradigm. Psychological Science, 16, 481– 486. doi:10.1111/j.09567976.2005.01560.x
Sandstrom, A., Rhodin, I. N., Lundberg, M., Olsson, T., & Nyberg, L.
(2005). Impaired cognitive performance in patients with chronic burnout
syndrome. Biological Psychology, 69, 271–279. doi:10.1016/
j.biopsycho.2004.08.003
Schulhofer, S. (1996). Miranda by the data: Substantial benefits and
vanishingly small social costs. Northwestern University Law Review, 90,
500 –564.
Simon, D. (1991). Homicide: A year on the killing streets. New York: Holt
Publishing.
Suen, H. K., & Ary, D. (1989). Analyzing quantitative behavioral observation data. Hillsdale, NJ: Lawrence Erlbaum Associates.
Watts, F. N., McKenna, F. P., Sharrock, R., & Tresize, L. (1986). Colour
naming of phobia-related words. British Journal of Psychology, 77,
97–108.
White, W. S. (2001). Miranda’s waning protections: Police interrogation
practices after Dickerson. Ann Arbor: The University of Michigan
Press.