To Plead or Not to Plead: A Comparison of Juvenile and Adult True and False Plea Decisions

Abstract

In a criminal justice system in which almost every adjudicated defendant, regardless of age, pleads guilty, it becomes important to understand the decision-making process underlying this choice. In the present research, we examined how age (juvenile vs. young adult), guilt versus innocence, and plea comprehension influenced the decision to plead guilty and the underlying plea rationale. We found that whereas age did not affect willingness to plead guilty when participants were asked to assume guilt in a hypothetical scenario, juveniles were more than twice as likely as young adults to plead guilty when asked to assume innocence. In addition, consistent with past research and developmental theory, juveniles were significantly less likely than adults to consider the short- and long-term consequences of the decision, and to understand and appreciate plea-related information. We also found that legal knowledge, after controlling for age, was positively (albeit weakly) related to plea decisions, but only for guilty participants. Implications for juveniles and adults involved in the criminal justice system, as well as wrongful convictions, are discussed.

Full text

To Plead or Not to Plead: A Comparison of Juvenile and Adult True and
False Plea Decisions
Allison D. Redlich
George Mason University
Reveka V. Shteynberg
University at Albany, State University of New York
In a criminal justice system in which almost every adjudicated defendant, regardless of age, pleads guilty,
it becomes important to understand the decision-making process underlying this choice. In the present
research, we examined how age (juvenile vs. young adult), guilt versus innocence, and plea compre-
hension influenced the decision to plead guilty and the underlying plea rationale. We found that whereas
age did not affect willingness to plead guilty when participants were asked to assume guilt in a
hypothetical scenario, juveniles were more than twice as likely as young adults to plead guilty when
asked to assume innocence. In addition, consistent with past research and developmental theory, juveniles
were significantly less likely than adults to consider the short- and long-term consequences of the
decision, and to understand and appreciate plea-related information. We also found that legal knowledge,
after controlling for age, was positively (albeit weakly) related to plea decisions, but only for guilty
participants. Implications for juveniles and adults involved in the criminal justice system, as well as
wrongful convictions, are discussed.
Keywords: guilty pleas, defendant decision-making, juvenile and adult plea comprehension
As the 20th century grew to a close, two patterns in the criminal
justice system emerged. First was that the number of defendants
pleading guilty increased to all-time high levels, a trend which has
continued through today (Oppel, 2011). In 1980, for example, 19%
of federal criminal cases went to trial; by 2010, less than 3% went
to trial (Rakoff, 2014). In state courts as well, about 95% or more
of all convictions—in juvenile and criminal courts—are the result
of guilty pleas (Bureau of Justice Statistics, 2013). Almost all
convicted defendants plead guilty, thereby waiving many of their
constitutional rights and trial safeguards in exchange for what most
often is a reduction in sentences or charges (Redlich, 2010a). The
second pattern to emerge was an increase in the number of iden-
tified wrongful convictions, in large part due to the application of
DNA testing to forensic settings (see generally Cutler, 2011). As
of April 2016, the National Registry of Exonerations recognizes
1,777 wrongful convictions since 1989. False admissions, which
include false confessions and false guilty pleas, are a leading
contributing factor, accounting for up to 27% of exonerations (see
http://www.innocenceproject.org/causes-wrongful-conviction).
An examination of the wrongful conviction cases involving
false (police-induced) confessions reveals that juveniles are at
especial risk (Kassin et al., 2010). For example, 33% of Drizin and
Leo’s (2004) sample of 125 proven false confessors were juve-
niles. A relatively unexplored question is whether youthful status
is also a risk factor for false guilty pleas—that is, pleading guilty
to crimes when factually innocent. Indeed, the question of whether
juveniles are more likely than adults to plead guilty when factually
guilty has not been sufficiently addressed. In the present study, we
examine, first, whether juveniles are more or less likely to plead
guilty when asked to assume they are guilty and innocent, and
second, the rationales underlying plea decision making for juve-
niles and adults. We also assess comprehension of plea-relevant
materials to determine whether this influences willingness to plead
guilty.
Plea Decision Making
Every day, thousands of juvenile and adult defendants must
decide whether to plead guilty (Lynch, 2003). Classic decision-
making theories (e.g., rational choice, expected utility) predict that
the decision to plead is driven by a desire to maximize benefits and
reduce costs (see Landes, 1971;Weimer, 1978). To this end, there
is some evidence that the plea decision depends on the perceived
strength of evidence, the probability of conviction at trial (Landes,
1971;Rhodes, 1979;Smith, 1986), and the value of the plea offer
(the distance between the sentence if convicted at trial and the plea
sentence).
Defendants facing the plea decision often have a Hobson’s
choice: plead guilty and get out of jail, or risk a harsher fate at trial
(Gross, Jacoby, Matheson, Montgomery, & Patil, 2005;Redlich,
This article was published Online First June 27, 2016.
Allison D. Redlich, Department of Criminology, Law and Society,
George Mason University; Reveka V. Shteynberg, School of Criminal
Justice, University at Albany, State University of New York.
The research was generously supported by National Science Foundation
Award 1025925. The opinions, findings, and conclusions expressed in this
publication are those of the authors and do not necessarily reflect the views
of the National Science Foundation. We thank Catherine Walker-Resnick,
Alyssa Clark, Chelsea Dubner, Jennifer Rounds-Weintraub, and Krista
Wallace for their assistance with the research.
Correspondence concerning this article should be addressed to Allison
D. Redlich, Department of Criminology, Law, and Society, George Mason
University, 4400 University Drive, Fairfax, VA 22030. E-mail: aredlich@
gmu.edu
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Law and Human Behavior © 2016 American Psychological Association
2016, Vol. 40, No. 6, 611–625 0147-7307/16/$12.00 http://dx.doi.org/10.1037/lhb0000205
611

2010a). As acknowledged, most choose to plead guilty, likely
because of the leniency attached to reduced sentences and/or
charges. In laboratory research, Bordens (1984) demonstrated that
plea offers involving any amount of jail time (whether it was six
months, a year, or three years) were significantly less likely to be
accepted than plea offers involving probation only. And, in a
survey of juvenile offenders, the most common rationales for true
and false guilty pleas were instrumental, which included reasons
related to getting a reduced sentence or getting out of jail (Malloy,
Shulman, & Cauffman, 2014). For innocents, even though falsely
acknowledging guilt is required, pleading guilty may be the ratio-
nal option. As argued by Bowers (2008), innocent defendants
should also be able to enjoy the benefits of pleading guilty,
including reducing or altogether avoiding custodial (jail or prison)
time. In the present study, we further investigate how the presence
of incarceration in plea offers influences willingness to accept
pleas by juveniles and adults, when asked to assume guilt or
innocence.
1
There is also evidence demonstrating that defendants are not
rational plea decision-makers (see Redlich, Bibas, Edkins, & Ma-
don, in press). Indeed, Bordens (1984) found that defendants were
more likely to make satisficing (a hybrid term of satisfactory and
sufficient) decisions, rather than optimizing ones. Simply seeking
the most benefit is unlikely to explain variations in plea decision
making. Additional factors that may influence plea decision mak-
ing and decision justifications are defendant’s age, understanding,
and whether one is guilty or innocent; these factors and the
relations between them are our focus here.
Age and Plea Decisions
The U.S. legal system has maintained segregated systems for
juveniles and adults for more than 100 years. In large part, the
reasoning behind this segregation is the recognition that juveniles
are developmentally less mature (cognitively, socially, emotion-
ally, and neurologically) than adults, which mitigates their culpa-
bility, and affects their ability to understand and appreciate legally
relevant information and meaningfully participate in their own
defense (e.g., Cauffman & Steinberg, 2012). Research has consis-
tently demonstrated that juveniles are less active participants in
legal cases (e.g., Tobey, Grisso, & Schwartz, 2000) and possess
inadequate legal knowledge and understanding (Grisso et al.,
2003;Steinberg & Scott, 2003). In comparison to adults, juveniles
aged 15 and younger have deficits in their legal understanding,
knowledge, and decision-making capabilities. These deficits
have been shown for Miranda rights’ understanding and other
interrogation-related matters (Grisso, 1981;Redlich & Good-
man, 2003;Viljoen, Klaver & Roesch, 2005), adjudicative com-
petence (e.g., Grisso et al., 2003), legal terminology comprehen-
sion (Kaban & Quinlan, 2004;Saywitz, Jaenicke, & Camparo,
1990), and other legal decision making (Grisso et al., 2003;
Schmidt, Reppucci, & Woolard, 2003;Singleton, 2007).
These age- and competence-related differences can be expected
to influence plea decision making and comprehension. More spe-
cifically, a valid plea decision requires an understanding of the
plea, the rights one is waiving, and the collateral consequences,
weighed against the alternatives (Redlich, in press). Concrete
thinking deficits and immature development limit these abilities in
some juveniles (Cauffman & Steinberg, 2012). Moreover, tender-
of-plea forms and judicial plea colloquies—the mechanisms used
to inform defendants of their rights and consequences of guilty
pleas—may not sufficiently address the rights, consequences, and
alternatives associated with guilty pleas because they are incom-
plete and/or incomprehensible. The forms use complex legal ter-
minology and are written, on average, at reading levels three to six
grades higher than the capabilities of most offenders (Redlich &
Bonventre, 2015). One study found that juvenile court judges and
attorneys believed that fewer than half of juvenile defendants
understood most or all of the plea colloquy (Sanborn, 1992).
Deficiencies in legal understanding and appreciation can be
theorized to affect plea decisions, and certainly the rationales
underlying the decision. Specifically, defendants may be more
willing to accept plea offers because they do not have the relevant
information and/or do not understand the information provided
(Bibas, 2011). In interviewing juvenile defendants preplea, Viljoen
and colleagues (2005) found that those with less legal understand-
ing were more likely to state they were unsure as to how they
would plead than those with more understanding. In the present
research, we examine how deficiencies in general legal knowledge
and specific plea knowledge relate to plea decision making, and
whether patterns change when individuals assume guilt or inno-
cence. Overall, we expect juveniles’ plea comprehension to be
lower than adults’ and that decreased comprehension will result in
a higher willingness to plead guilty (particularly when assuming
innocence; see Norris & Redlich, 2011).
Though research examining age differences in plea decision
making is sparse, some studies have begun to explore this issue.
For example, Grisso et al. (2003) examined the willingness of
1,400 community- and justice-involved juveniles and adults
(youngest aged 11–13; oldest aged 18 –24) to accept a hypothetical
plea offer (when guilty). They found that with increasing age,
willingness decreased. About 70 –74% of the 11–15 year-olds
accepted the plea offer in comparison to 50% of the adults, which
was attributed in part to compliance with authority. These age-
related findings remained significant after considering gender,
ethnicity, detained/community status, intelligence, and socioeco-
nomic status. Grisso and his colleagues (see also Grisso, 2007)
further found that in their plea decision making, youths focused on
the length of time associated with the plea (two vs. six years),
whereas adults’ reasoning reflected attempts to weigh the odds
(two years vs. six years vs. the possibility of zero years). In
comparison to older youths and young adults, younger youths were
less able to recognize risks and identify longer-term future conse-
quences in their decision making.
However, within a smaller sample of juvenile detainees, Viljoen
and colleagues (2005), did not find age itself (11- to 14- vs. 15- to
17 year-olds) to influence future decisions to plead guilty. They
did find that strength of evidence predicted plea decisions of older
juveniles, but not younger juveniles, indicating that the reasoning
behind plea decision making (regardless of the decision itself) can
be affected by age. Similarly, Peterson-Badali and Abramovitch
(1993) found indications that younger individuals were less likely
1
The present study uses a hypothetical situation and asks participants to
assume either guilt or innocence (as part of the guilt manipulation). For
simplicity, when discussing results, we refer to participants assuming guilt
as “guilty” and those assuming innocence as “innocent” participants.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
612 REDLICH AND SHTEYNBERG

than older ones to consider evidence in their (hypothetical) plea
decision making. For example, within one of the scenarios pro-
vided, when the evidence was weak, 50% of 5th graders pled
guilty, compared with 17%, 13%, and 8% of 7th graders, 9th
graders, and adults, respectively. When the evidence was strong,
however, plea rates approached ceiling levels for all age groups
(and across differing scenarios).
Finally, within a sample of adult defendants who had pled
guilty, Bordens and Bassett (1985) found that younger defendants
were more likely than older ones to plead guilty to minimize
suffering (in contrast to other reasons, including avoidance of
felony convictions or lawyer’s advice). Indeed, age was “the only
variable that showed any relationship at all with reason for [plea]
acceptance” (Bordens and Bassett, 1985, p. 102). Factors found to
not influence plea decision making included crime severity (rang-
ing from thefts to DWIs to murders), education level, presence of
a prior record, attorney type, and knowledge concerning the like-
lihood of conviction and sentences associated with plea and trial
convictions.
Overall, the extant research is inconsistent regarding the rela-
tionship between age and plea decisions. However, with few
exceptions, much of the research has been within a framework of
presumed guilt. That is, the studies tend to employ methodologies
asking mock or actual juvenile defendants what they would do
when guilty, but not when innocent.
Innocence and Plea Decisions
Pleading guilty to crimes you did not commit— or false guilty
pleas— undoubtedly exist (Garrett, 2008;Gross et al., 2005). Of
the few studies that have been conducted with regard to culpabil-
ity, it is clear that mock innocent defendants are much less likely
to plead guilty than mock guilty defendants (Bordens, 1984;Greg-
ory, Mowen, & Linder, 1978), but still as many as 56% of
innocents in the laboratory have been induced to plead guilty
(Dervan & Edkins, 2013). This comparative lack of innocents’
willingness to falsely plead guilty has been construed as a system-
atic and overly optimistic bias toward the odds of acquittal as well
as a perceived unfairness of being wrongly accused (see Tor,
Gazal-Ayal, & Garcia, 2010). Across four studies, Tor and col-
leagues (2010) found that participants who were asked to imagine
they were innocent of a mock crime demonstrated increased risk-
seeking (in that they were willing to take their chances at trial and
not accept pleas) than those asked to imagine they were guilty. Tor
et al. (2010) reasoned that innocents prefer risking an unfavorable
disposition at trial because they perceive the choice between plea
and trial (two negative outcomes) to be substantively unfair.
False guilty pleas are essentially false confessions, though there
are important differences (Redlich, 2010a). As stated by Tepfer,
Nirider, and Tricarico (2010), “The similarities between the two
are unmistakable, as a false guilty plea can be thought of as
nothing more than a specific type of false confession” (p. 913).
With regard to age and false guilty pleas, to our knowledge, there
have been three investigations addressing this relationship. All
three revealed nonsignificant differences between juveniles and
adults in rates of false guilty pleas. First, among the exonerations
in the National Registry of Exonerations, the correlation between
age at time of conviction and false guilty pleas is not significant
(see www.exonerationregistry.org/). Second, Tepfer et al. (2010)
examined a sample of 103 youths aged 20 years and younger who
were wrongly convicted and later exonerated. They found that
6.8% had allegedly falsely pled guilty, whereas a comparison
sample of adult exonerees had a false guilty plea rate of 7.9%, a
nonappreciable difference. Third, Malloy and her colleagues
(2014) did not find age (within a limited age-range sample of 14-
to 17-year-old juvenile offenders) to influence self-reported false
(or true) guilty plea rates.
In hypothesizing relations between age group (juveniles vs.
adults) and pleading guilty, there are three possibilities (see also
Redlich, 2010b). First, juveniles could be predicted to be more
likely to plead guilty—truly and falsely—than adults. As reviewed
above, there is some evidence indicating that juveniles are more
likely than adults to plead guilty when guilty (Grisso et al., 2003).
Further, a leading risk factor for false confessions is youth (Gross
et al., 2005;Owen-Kostelnik, Reppucci, & Meyers, 2006). As
mentioned, juveniles are overrepresented in proven false confes-
sion cases (Drizin & Leo, 2004;Kassin et al., 2010). And in the
laboratory, preteens and teens have been found to be more likely
to falsely take responsibility for a mock crime in comparison to
adults (Redlich & Goodman, 2003). Thus, because of an increased
susceptibility to true guilty pleas and to false confessions, one
could hypothesize that age and likelihood of guilty pleas are
negatively related (see also Drizin & Luloff, 2007).
Another reason why age and likelihood of pleading guilty may
be expected to be negatively associated relates to the rationales
underlying the decision and the concept of comparative fairness;
in this instance, “evaluations of plea offers vis-a
`-vis offers made in
similar cases” (Tor et al., 2010, p. 107). To examine this notion in
relation to plea decisions, Tor and colleagues (2010) manipulated
whether the sentence associated with a plea was described as
shorter than, similar to, or longer than sentences typically offered
by the prosecutor. They found that, despite the value of the
sentence being fixed, adults who were offered the comparatively
“worse” and “similar” deals were significantly less likely to plead
guilty than those offered the “better” deal, a pattern that emerged
for both guilty and innocent participants. Because juveniles may be
particularly likely to base decisions on comparative evaluations, in
the present research we investigate whether manipulations of com-
parative fairness explain potentially differential plea rates among
juveniles and young adults. Social influence and conformity to
norms become very important in adolescents’ decision making
(Steinberg, 2005), and thus, juveniles are expected to be impacted
more than adults by manipulations of comparative sentences.
The second hypothesis concerning age and plea decision making
is that juveniles will be significantly less likely to plead guilty than
adults. The same developmental limitations that place juveniles at
risk for not being competent (e.g., in assisting in their own de-
fense) may also place them at risk for not recognizing, and there-
fore not acting in, their own best interest. As discussed, going to
trial is a riskier option than accepting a plea; trial acquittals
account for only 1% of all criminal justice outcomes and 25% of
trial outcomes (Cohen & Reaves, 2006). Given that juveniles as a
group are characterized as risk-seekers (e.g., Steinberg, 2005), it is
possible that juveniles will be less likely to accept plea offers
compared with adults. This sentiment is expressed in an article
titled I Ain’t Taking No Plea: The Challenges in Counseling Young
People Facing Serious Time (Smith, 2007).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
613
AGE AND PLEA DECISION MAKING

Regarding false guilty pleas specifically, another reason to sus-
pect that juveniles will be significantly less likely than adults to
plead guilty when innocent relates to the rationales underlying the
decision. Peterson-Badali and Abramovitch (1993) examined chil-
dren’s and adult’s reasons behind guilty plea decisions. They
found that children in grades five, seven, and sometimes nine
(mean ages 10.7, 13.0, and 14.7 years, respectively) were more
likely than adults to refer to a person’s actual guilt as a reason for
pleading guilty. The authors suggested that the younger partici-
pants had “not made the distinction between legal and moral
domains of reasoning” (p. 549). Thus, for false guilty pleas, but not
true guilty pleas, younger (and innocent) persons may rely on a
presumption of innocence as a defense and refuse to plead guilty
based on moral reasons; that is, only the guilty should plead guilty.
Finally, a third hypothesis is that the likelihood of pleading
guilty will not be affected by age group status, as found by the
research on age and guilty pleas reviewed above (i.e., Malloy et al.,
2014;Tepfer et al., 2010;Viljoen et al., 2005). In the present
study, we examine which of these three alternate hypotheses
between age and guilty pleas (positive, negative, or no significant
relations) will find support.
Method
Participants
Participants included 189 individuals, split into two age groups:
89 juveniles (aged 13–17 years), and 100 young adults (aged
18 –24 years). As shown in Table 1, juveniles and young adults
were similar in most respects, including gender, percent ever
arrested, average grades in school, level of mother’s education, and
self-reported impulsivity. (Father’s education was also asked
about, but a significant minority [20%] did not know their father’s
education.) However, a higher proportion of the juveniles were
minorities than the young adults (see Table 1).
Measures
Hypothetical scenario. The hypothetical scenario concerned
a robbery in a jewelry store (see Appendix), involving three
randomly assigned manipulations. First, for the guilt–innocence
manipulation, half of the participants were asked to imagine they
were innocent and half that they were guilty. Second, for the plea
deal manipulation, participants were informed that the plea offer
either included jail (i.e., 5 years’ probation and two years in jail or
juvenile detention) or did not include jail (i.e., 5 years’ probation).
Third, for the plea label manipulation, participants were either told
that the deal is “really good” or that the deal is “fair but not great”
“compared with what others charged with the same crime are
typically offered.” All participants were told that their chance of
being found guilty at trial was high (75%) and that their sentence
would be 5–7 years in prison if convicted at trial.
Demographics. Participants were asked demographic ques-
tions, including items about age, gender, race/ethnicity, educa-
tional attainment of participant/parent, school grades, and previous
legal experience. For race, participants were presented with a
laminated card listing six races (e.g., Caucasian/White; African
American/Black), plus an “other” race to specify. Participants
were read the race options aloud and asked to indicate all which
applied to them. For ethnicity, participants were asked whether or
not they were Hispanic/Latino(a). For analysis, race/ethnicity was
dichotomized into majority (White and non-Hispanic) versus mi-
nority (non-White and/or Hispanic). For educational attainment,
participants were asked questions regarding the highest grade they
completed in school and the highest level of education attained by
their mother and father, respectively (with nine options ranging
from 0 –11th grade to postgraduate degree). They were also asked
whether they mostly get/got A’s, B’s, C’s, D’s, or F’s in school.
Plea decision, tender of plea, and judicial oral plea colloquy.
Participants entered both a written and an oral plea of Guilty or
Not Guilty by either (a) waiving their rights and pleading guilty or
(b) invoking their right to a trial and pleading not guilty. After the
hypothetical scenario, participants read a tender of plea form,
which discussed the rights and consequences associated with ac-
cepting a plea offer, and entered a written plea decision. The oral
plea colloquy was verbatim the same as the tender of plea form
except for minor changes made so it could be read aloud. The
form/colloquy had a Flesch-Kinkaid reading level of 9th grade,
was 2.5 pages, and 970 words (copies can be obtained upon
request to the first author). There were three main headings: (a)
“Waiver of Constitutional Rights” (ending with the statement: “I
know and understand that by pleading guilty I would be waiving
and giving up all the constitutional rights as explained above”); (b)
“Consequences of Entering a Guilty Plea”; and (c) “Defendant’s
Certification of Voluntariness.” The form/colloquy ended with the
statement, “With knowledge of the above facts and considerations,
I wish to 1) Waive my rights and plead guilty; or 2) Invoke my
Table 1
Demographic Characteristics of Participants by Age Group
Characteristic Juveniles (n⫽89)
Young adults
(n⫽100) Significance test Effect size
% male 51.7% 51.0% ␹
2
(1) ⫽0.01 ⌽⫽.01
% minority 73.0% 59.0% ␹
2
(1) ⫽4.11
ⴱ
⌽⫽.15
% ever arrested 28.1% 32.0% ␹
2
(1) ⫽0.34 ⌽⫽.04
Average grades in school (SD) 1.90 (0.71) 1.89 (0.76) F(1) ⫽0.001 d⫽.01, 95% CI [⫺0.27, 0.30]
Average level of mother’s education (SD) 4.38 (2.43) 4.54 (2.48) F(1) ⫽0.19 d⫽⫺.06, 95% CI [⫺0.35, 0.66]
Average impulsivity (SD) 22.51 (4.54) 21.48 (4.69) F(1) ⫽2.33 d⫽.22, 95% CI [⫺0.06, 0.51]
Note. Average grades in school: 1 (Mostly As)to5(Mostly Fs). Average level of mother’s education: 1 (less than high school)to9(postgraduate).
Impulsivity: higher scores indicate more impulsivity. CI ⫽confidence interval.
ⴱ
p⬍.05.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
614 REDLICH AND SHTEYNBERG

right to a trial and plead not guilty.” For the written plea, partic-
ipants were prompted to circle one of the two plea options, and
then sign and date the written form. For the oral plea, participants
provided an answer aloud. Later, in the interview phase, partici-
pants’ perceived voluntariness and confidence of their decision
were measured on 20-point scales (1 ⫽not at all voluntary/
confident to 20 ⫽extremely voluntary/confident).
Plea rationale. Participants were first given two opportunities
to explain in their own words why they did or did not plead guilty,
which was then followed up with more specific closed-ended
questions about their plea reasoning. Participants were asked to
rate the degree to which several statements reflected their plea
decision making on a scale of 1 (not true at all)to5(extremely
true). There were seven closed-ended statements, which differed
based on whether they chose to plead guilty or not guilty, but were
matched for consistency and content. For example, “I thought the
plea offer was a good deal” and “I was thinking about the long-
term benefits of pleading guilty” were statements for those who
pled guilty, in contrast to “I did not think the plea offer was that
good of a deal” and “I was thinking about the long-term conse-
quences of pleading guilty” for those who pled not guilty.
Plea knowledge. Plea knowledge consisted of separate mea-
sures, which included assessments of (a) vocabulary words used in
the tender-of-plea form and oral plea colloquy, (b) plea compre-
hension, and (c) adjudicative competence. The vocabulary mea-
sure was a list of 19 words directly from the tender-of-plea form/
colloquy that participants were read and asked to define. Example
words included liable, impartial, sentence, self-incrimination, and
right. Participants were instructed to define words as they “relate
to court procedures, the legal system, or pleading guilty.” Partic-
ipants were also told that they could say “I don’t know” if they did
not know what the word meant. Definitions were scored as “2” for
correct, “1” for partially correct, or “0” for incorrect responses. “I
don’t know” answers were scored as incorrect. To determine
whether definitions were correct, we used Black’s Law Dictionary
(Garner, 2014), TheFreeDictionary online legal dictionary, and
The Living Word Vocabulary: A 44,000 Word Vocabulary Inven-
tory (Dale & O’Rourke, 1981), which defines words and phrases
as well as assigns grade-level and difficulty ratings to each. This
method for defining and coding legal terms and phrases has been
used in past research (Kaban & Quinlan, 2004). An example of an
incorrect definition of liable would be “to rely on,” a partially
correct definition would be “whether it’s your fault,” and a correct
definition would be “held accountable” or “legally responsible.” A
total vocabulary score was created by summing up scores; possible
range is 0 to 38. The Cronbach’s alpha for this summary score
is .90.
Plea comprehension was measured with 16-True–False-I-Do-
Not-Know statements assessing participants’ accuracy in under-
standing the plea process, requirements, and consequences (e.g.,
“guilty pleas do not have to be the voluntary choice of the defen-
dant”). These questions were a subset of those used by Redlich and
Summers (2012), which have demonstrated sufficient internal con-
sistency. Answers to the questions could be found in the tender-
of-plea form/colloquy. Answers were scored as incorrect (0) or
correct (1); “I don’t know” answers were scored as incorrect. A
summary score was created by summing the 16 correct–incorrect
scores; scores in the present study ranged from 4 to 16.
The MacArthur Competence Assessment Tool–Criminal Adju-
dication (MacCAT-CA; Poythress et al., 1999) was used to mea-
sure adjudicative competence. The MacCAT-CA evaluates (a)
understanding, (b) reasoning, and (c) appreciation as it pertains to
the Dusky v. US (1960) standard of adjudicative competence (see
Bonnie, 1992). The MacCAT-CA was developed as a two-part
research instrument for use with adult defendants with and without
mental health problems. In the first section, participants are read a
brief story about two men who get into a fight at a bar and are then
asked questions pertaining to their Understanding (e.g., what are
the roles of the judge, jury, and attorneys) and Reasoning (e.g.,
what information defendant should tell his lawyer) based on this
story. The second section asks Appreciation questions related to
the defendant’s own pending criminal case. The questions in the
latter section were not appropriate for the present study because
participants had already made their (hypothetical) plea decision, so
participants in this sample were not asked these questions. The
MacCAT-CA, which has been used with juveniles (e.g., Grisso et
al., 2003), has well-established psychometric properties. Under-
standing and Reasoning have strong interrater reliability correla-
tions, ranging from very good to excellent (i.e., from .85 to .90),
and internal consistency, with alphas of .85 and .81, respectively
(see Otto et al., 1998).
Impulsivity. To measure impulsivity, participants rated 10
impulsivity-related statements on a scale of 1 to 4, where 1 ⫽
Rarely or Never,2⫽Occasionally,3⫽Often, and 4 ⫽Almost
Always or Always. Redundancy was built into these statements to
assess the same construct (e.g., “I act ‘on impulse’,” “I act on the
spur of the moment,” “I find it hard to sit for long periods of
time”). This impulsivity measure was taken from the MacArthur
Foundation’s Clinically Useful Actuarial Measure of Risk devel-
oped for a study on Violence Risk Assessment (Monahan et al.,
2000), which they adapted from the 30-item Barratt Impulsiveness
Scale (BIS-11; Barratt, 1994;Patton, Stanford, & Barratt, 1995).
Manipulation checks. Participants were asked four True–
False-I Do-Not-Know manipulation check questions about the hypo-
thetical robbery scenario. Participants were asked questions regarding
their guilt–innocence (“I was asked to pretend I was innocent of the
crime”), plea deal (“the plea deal I was offered was 5 years on
probation and 2 years in jail”), and comparative fairness (“My lawyer
told me that in comparison to others charged with this crime, my plea
offer was really good”). “I don’t know” answers were scored as
incorrect.
Procedures
The University human subject review board approved all study
procedures and instruments. Juvenile and adult participants were
recruited from the nearby community of the authors’ university.
Thirteen community organizations for at-risk youth and young
adults served as places of recruitment. Efforts involved posting
flyers, handing out information sheets, and making 3-min speeches
about the study and how to contact researchers if interested. A
minority of participants were also recruited from Craigslist.
Interested participants came into the lab for a one-time session.
Informed consent/assent was administered and obtained from par-
ticipants and parents; there was no deception and all were told that
the study was examining how teens and young adults make deci-
sions whether to go to trial or plead guilty. Participants were
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
615
AGE AND PLEA DECISION MAKING

randomly assigned into one of eight conditions by varying guilt–
innocence, the plea deal (some jail–no jail), and plea label (really-
good–fair-plea offer), with roughly equal numbers assigned across
age group and gender.
First, participants were asked demographic questions. Second,
the interviewer read the hypothetical case aloud to the participant
while they read along (see Appendix). Third, participants read the
written tender-of-plea form, were asked if they had any questions
after reading the form, and were then asked to enter a written plea
of Guilty or Not Guilty. Then, a judge (a female confederate
wearing a black robe and holding a gavel) entered the room and
conducted the oral plea colloquy. Participants were asked to rise
and answer the judge’s questions, including the final question
about whether they wanted to waive their rights and plead guilty or
invoke their rights and plead not guilty. At three separate times,
judges asked participants if they had any questions about their
rights or consequences of pleading guilty (in addition to specific
points of understanding; e.g., “Do you understand that you will
only be allowed to withdraw your plea if you show that it was not
knowingly and voluntarily made?”).
Next, participants were interviewed. They were asked about the
rationales underlying their plea decision and about their perceived
voluntariness of, and confidence in, their plea decision. Partici-
pants were also asked the manipulation check questions, inter-
viewed about their plea understanding via the plea knowledge
measures described above, and assessed on their adjudicative
competence using the MacCAT-CA. Upon completion, partici-
pants were debriefed and paid $15 for their participation. The
entire session lasted 73 min (SD ⫽13), on average.
Results
As noted above, four manipulation check questions were asked.
In regard to the guilt–innocence manipulation, 79% of the sample
responded correctly. For the comparative fairness manipulation,
85% of the sample was correct. For the jail–no jail manipulation,
two questions were asked, and 84% to 91% of respondents were
correct. The main analyses are conducted only with participants
who correctly answered all four manipulation questions (n⫽115).
However, accuracy on the manipulation checks was significantly
correlated with self-reported grades, r⫽⫺.34, p⬍.0001 and with
all of the knowledge based measures (plea comprehension, vocab-
ulary, and MacCAT-CA scores), rsⱖ.43, psⱕ.0001. Thus, so as
not to exclude participants with potentially lower intellect (and
perhaps those who generalize the most to actual offender samples),
analyses were also conducted with the entire sample. When find-
ings diverge from those conducted with the subsample who accu-
rately answered all manipulation checks, differences are noted.
For the entire sample, juveniles and young adults differed sig-
nificantly by minority status (see Table 1); in the subsample of
those who correctly answered the manipulation checks, minority
status did not differ by age group, ␹
2
(1, N⫽115) ⫽2.63, p⫽.11,
⌽⫽.15. (The other demographic characteristics listed in Table 1
remained nonsignificant by age group in the subsample of those
who answered all manipulation check questions correctly.) None-
theless, we examined the relationship between minority status and
willingness to plea, the main dependent measure. Minority status
did not influence the dichotomous plea decision, ␹
2
(1, N⫽115) ⫽
0.00, p⫽1.00, ⌽⫽.00; 40% of minorities and nonminorities each
pled guilty. Thus, minority status is not considered further. In
addition, we examined the relationship between impulsivity scores
and willingness to plea and found it to be nonsignificant, Spear-
man’s r⫽⫺.01, p⫽.94 (when guilty, r⫽⫺.10, p⫽.34; when
innocent, r⫽.09, p⫽.38).
Participants entered both a written (on the tender-of-plea form)
and an oral (to the judge) plea decision. In the full sample, only 11
participants (5.8%; seven of whom were adults) changed their
decision (in the analysis subsample of 115 participants, four of
these 11 participants remained). Eight participants who pled guilty
on the written form changed their plea to not guilty when in front
of the judge, whereas three people did the opposite. Because so
few people changed their decision and because the oral plea was
the final decision, the oral plea was used in the analyses below.
Overall, 47.8% of participants (55 of 115) pled guilty when in
front of the judge. However, there were several factors that served
to increase or decrease this overall rate.
Plea Decision and Age Group
We first conducted a chi-square analysis between age group and
plea decision. The difference between juveniles and young adults
was not significant (see Table 2). Our next step was to examine
plea decision differences by age group and the manipulated con-
ditions. We note here that the some jail–no jail (plea offer) and the
really-good–fair (plea label) manipulations did not directly influ-
ence the plea decision, ␹
2
ⱕ2.40, psⱖ.12, ⌽sⱕ.14 (i.e., when
age group and guilt–innocence were not considered). However, the
guilt–innocence manipulation strongly influenced willingness to
plea, ␹
2
(1, N⫽115) ⫽28.87, p⬍.0001, ⌽⫽.50. Whereas
71.0% of participants in the guilty condition pled guilty, 20.8% of
participants in the innocent condition did so. With the full sample,
results remained the same with a threefold difference between plea
rates for the guilty and the innocent.
Guilt–innocence conditions. When the guilt–innocence ma-
nipulation was added as a level to the chi-square analysis with age
group, an interaction emerged (see Table 2). Specifically, when
participants were asked to assume guilt, the difference in decisions
to plead guilty remained nonsignificant between juveniles and
young adults, ␹
2
(1) ⫽0.07, p⫽.79. When asked to assume
innocence, however, juveniles were 2.47 times more likely to plead
Table 2
Plea Acceptance Rates by Age Group and Condition
Accurate on all
manipulation checks Entire sample
Juveniles
(n⫽48)
Young adults
(n⫽67)
Juveniles
(n⫽89)
Young adults
(n⫽100)
Overall plea rate 52.1% 44.8% 46.1% 39.0%
Guilty 69.2% 72.2% 62.8% 66.0%
Innocent 31.8% 12.9% 30.4% 15.1%
Plea offer: No jail 46.7% 59.5% 45.0% 48.1%
Plea offer: Some jail 54.5% 26.7% 46.9% 28.3%
Plea label: Fair deal 47.6% 43.8% 38.1% 37.0%
Plea label: Really
good deal 55.6% 45.7% 53.2% 40.7%
Note. Significant differences (pⱕ.10) by age group are bolded.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
616 REDLICH AND SHTEYNBERG

guilty than adults, ␹
2
(1) ⫽2.80, p⫽.09, ⌽⫽.23. Results remained
the same when rerun with the entire sample (see Table 2).
Some jail–no jail conditions. We next added the plea deal
manipulation as a level of the chi-square analysis with age group.
In the no-jail condition, juveniles and young adults did not signif-
icantly differ in their likelihood to plead guilty, ␹
2
(1) ⫽0.71, p⫽
.40, ⌽⫽.12. In the some-jail condition, however, juveniles were
significantly more likely to plead guilty than adults, ␹
2
(1) ⫽5.04,
p⫽.025, ⌽⫽.28 (see Table 2); an effect which remained
significant when all participants were included.
We also conducted a chi-square analysis with the guilt and the
jail manipulations and plea decision, revealing that willingness to
plead guilty was affected by these two manipulations. Specifically,
when guilty, participants were more likely to plead guilty when the
plea offer did not include jail time (86.2%) than when the plea
offer included jail time (57.6%), ␹
2
(1) ⫽6.14, p⫽.013, ⌽⫽.32.
However, when innocent, participants were not significantly more
or less likely to plead guilty both when there was jail time (23.3%)
and was not jail time (17.4%), ␹
2
(1) ⫽0.28, p⫽.60, ⌽⫽.07.
Because age group interacted with both the plea offer and the
guilt–innocence conditions to influence plea rates, we conducted
chi-squares of the plea offer and guilt manipulations separately for
juveniles and young adults. For young adults, the same pattern was
found. Specifically, when guilty (but not when innocent), young
adults were significantly more likely to plead guilty when the plea
offer did not include jail time (90.9%) than when the offer did
include jail time (42.9%), ␹
2
(1, N⫽67) ⫽9.85, p⫽.002, ⌽⫽
.52. For juveniles, the amount of jail time, when either guilty or
innocent, did not significantly influence plea rates, ␹
2
’s (1, n⫽
48) ⱕ0.27, psⱖ.60. Results remained the same when the entire
sample was analyzed.
Really-good–fair-plea offer manipulation. Next, we con-
ducted chi-square analyses with age group and plea decision,
adding the plea label manipulation as a level. No significant
effects of age group were found for this manipulation (see Table
2), ␹
2
s (1) ⱕ0.59, psⱖ.44. Plea decisions were also examined by
comparative fairness for guilty and innocent subjects separately,
and no significant results emerged, ␹
2
(1) ⱕ0.89, psⱖ.34. Results
did not change when the entire sample was included.
Plea decision confidence and voluntariness. Participants
were asked to rate the confidence and voluntariness of their plea
decision on a scale of 1 (not at all confident/voluntary)to20
(extremely confident/voluntary). Separate 2 (age group) ⫻2
(guilt–innocence) ⫻2 (plea offer) ⫻2 (plea label) analyses of
variance (ANOVAs) were conducted, using these two scales as
dependent measures.
For the confidence in decision analysis, the only significant
effect to emerge was a three-way interaction between age group,
guilt manipulation, and plea label manipulation, F(1, 99) ⫽6.14,
p⫽.015, partial ␩
2
⫽.06. In examining these interactions via
ANOVAs isolating age group within conditions, only one signif-
icant difference emerged by age group. Specifically, adults, M⫽
16.76 (SD ⫽3.09), were significantly more confident than juve-
niles, M⫽14.20 (SD ⫽4.04), F(1, 30) ⫽4.12, p⫽.05, d⫽.72
(95% confidence interval CI [0.02, 1.42]) when innocent and the
plea offer was labeled as “really good.” When guilty, regardless of
the plea label, juvenile and young adults did not differ in their
confidence levels. Similarly, when innocent and the plea was
described as a fair offer, confidence was unaffected by age group
(all psⱖ.12). When this ANOVA concerning confidence scores
was rerun using the entire sample, the only significant effect to
emerge was a main effect of age group. Juveniles, M⫽14.76
(SD ⫽4.43) were significantly less confident in their plea deci-
sions than adults, M⫽16.40 (SD ⫽3.33), F(1, 173) ⫽8.55, p⫽
.004, d⫽.42 (95% CI [0.13, 0.71]).
For the voluntariness analysis, again, only one significant effect
emerged. Specifically, the interaction between age group and the
jail manipulation was significant, F(1, 99) ⫽5.43, p⫽.022,
partial ␩
2
⫽.05. In the some-jail condition, adults, M⫽18.17
(SD ⫽4.15) perceived their plea decision as more voluntary than
juveniles, M⫽14.73 (SD ⫽5.46), F(1, 61) ⫽7.81, p⫽.007, d⫽
.70 (95% CI [0.19, 1.21]). In the no-jail condition, perceived
voluntariness did not significantly differ between juveniles, M⫽
17.27 (SD ⫽3.35) and adults, M⫽16.92 (SD ⫽4.57), F(1, 50) ⫽
0.07, p⫽.79, d⫽.08 (95% CI [⫺0.52, 0.68]). When the ANOVA
was rerun with the entire sample, this interaction effect remained
significant (p⫽.022), in addition to the main effect of age group
reaching significance (juveniles M⫽15.89 [SD ⫽4.83]; adults
M⫽17.37 [SD ⫽4.22], d⫽.33 (95% CI [0.04, 0.61]). There
were no other significant main or interaction effects.
Plea Knowledge
There were a total of four measures relevant to knowledge:
vocabulary, plea comprehension, and the MacCAT-CA Under-
standing and Reasoning scores. These four measures were signif-
icantly correlated, rs ranged from .22 to .61, psⱕ.02. Thus, they
were entered in a multivariate analysis of variance (MANOVA)
with age group and the three manipulations as independent factors.
Significant multivariate effects were found for age group only,
F(4, 96) ⫽3.94, p⫽.005, partial ␩
2
⫽.14. No other significant
multivariate main or interaction effects emerged.
Three of the four univariate effects of age group were signifi-
cant. Means, standard deviations, and effect sizes are presented in
Table 3. With the exception of MacCAT Reasoning scores, juve-
niles scored significantly lower on plea and legal knowledge than
adults. When the same MANOVA was rerun with the entire
sample, again only the multivariate effect of age group was sig-
nificant. However, all univariate effects were significant, including
the effect of age group on MacCAT Reasoning scores: juveniles,
M⫽12.50 (SD ⫽2.89), young adults, M⫽13.41 (SD ⫽2.12)
F(1, 172) ⫽5.62, p⫽.019, d⫽.36 (95% CI [0.07, 0.65]).
It was also of interest to determine if knowledge was related to
the plea decision itself. Because there were clear age differences in
knowledge, partial correlations, controlling for participant age,
were conducted to examine relations between the four knowledge
scores and the plea decision. None of the correlations were sig-
nificant, rsⱕ.14, psⱖ.13. As shown in Figure 1, when the same
correlations were recomputed for guilty and innocent participants
separately, divergent patterns emerged. None, however, reached
significance, as the sample sizes decreased when separate by the
guilt–innocence manipulation. When the entire sample was in-
cluded, however, higher scores on the MacCAT Understanding,
r⫽.30, p⫽005 and Reasoning scores, r⫽.32, p⫽.002, and
vocabulary scores, r⫽.24, p⫽.02 were significantly correlated
with the decision to plead guilty, but only when participants were
asked to assume guilt. When asked to assume innocence, all
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
617
AGE AND PLEA DECISION MAKING

knowledge scores were negatively correlated with the plea deci-
sion, but none reached significance (rsⱕ⫺.09, psⱖ.36).
Plea Rationales
As described above, participants who pled guilty and who pled
not guilty were asked separately about seven different possible rea-
sons for their decision. Because the primary focus of this paper is on
differences between juveniles and adults, and guilt–innocence, we
conducted a series of separate 2 (age group) ⫻2 (guilt–innocence
manipulation) ANOVAS. It was not possible to include the jail and
fairness manipulations because of the split samples (i.e., those who
pled guilty vs. not guilty). Responses to the seven reasons were
generally not significantly correlated with one another (e.g.,
among 21 possible correlations for reasons for pleading not guilty,
only two significant correlations were found, rsⱕ.40). As a result,
MANOVAs were not appropriate.
Reasons for pleading not guilty. Among the seven reasons
supplied for pleading not guilty, ANOVAs revealed significant
main effects for three. Means and standard deviations for juveniles
and adults, and the guilt-innocent manipulation are shown in Table
4. There were no significant interaction effects. Results generally
remained the same when the entire sample was analyzed (also
shown in Table 4).
For age group, there was only one significant main effect (for
pleading not guilty), pertaining to the rationale, “I was thinking
about the long-term consequences of pleading guilty,” F(1, 56) ⫽
6.38, p⫽.014, d⫽.73 (95% CI [0.19, 1.26]). In comparison to
juveniles, young adults were more likely to endorse pleading not
guilty because of this reason. There was also a nonsignificant trend
(i.e., pⱕ.10) for the rationale, “I thought if I did not accept the
offer, the DA would offer a better deal”; juveniles tended to be
more likely than adults to endorse this rationale, d⫽.52, 95% CI
[⫺0.02, 1.04]. In addition, when the entire sample was included
and the analyses rerun, juveniles tended to be less likely than
adults to endorse these two rationales: “I was thinking about what
would happen in the near future” (d⫽.48, 95% CI [0.10, 0.86])
and “I did not think the plea offer was that good of a deal” (d⫽
.29, 95% CI [⫺0.09, 0.67]; see Table 4).
For the guilt-innocent manipulation, two significant main effects
emerged (see Table 4). Guilty participants, in comparison to in-
nocent ones, were more likely to endorse pleading not guilty
because “I thought if I did not accept the plea offer, the DA would
offer a better deal,” F(1, 56) ⫽8.00, p⫽.006, d⫽.84 (95% CI
[0.25, 1.40]). In contrast, innocent participants were more likely
than guilty participants to think “the truth would come out at trial,”
F(1, 56) ⫽27.02, p⬍.0001, d⫽1.52 (95% CI [0.88, 2.11]).
These findings remained significant when rerun with the entire
sample.
Reasons for pleading guilty. Among the seven reasons sup-
plied for pleading guilty, when only the subsample of those who
correctly answered the manipulations were included, ANOVAs
revealed no significant main effects of age group (Table 4,
nonbolded rows). However, there was a significant interaction
between age group and the guilt manipulation, as well as one
significant main effect of the guilt-innocent manipulation. Par-
ticipants in the guilty condition were more likely to endorse the
rationale, “It was the right thing to do” than those in the
innocent condition, F(1, 51) ⫽9.95, p⫽.003, d⫽1.14 (95%
CI [0.43, 1.81]).
The significant interaction concerned “I did not think my public
defender lawyer would do a good job,” F(1, 51) ⫽4.84, p⫽.032,
partial ␩
2
⫽.07. When innocent, the difference between juveniles,
M⫽3.00 (SD ⫽1.41) and adults, M⫽2.00 (SD ⫽0.00) was not
significant, F(1, 11) ⫽1.91, p⫽.20. When guilty, adults, M⫽
3.08 (SD ⫽1.57) were significantly more likely than juveniles,
M⫽1.89 (SD ⫽1.32) to plead guilty because they thought their
lawyer would not do a good job at trial, F(1, 42) ⫽6.88, p⫽.012,
d⫽.81 (95% CI [0.17, 1.42]).
When the ANOVAs were rerun with the entire sample, several
differences were noted (see Table 4), particularly for age group.
Specifically, four significant main effects of age group emerged
Table 3
Plea Knowledge Score Comparisons by Age Group
Score
Juveniles
(n⫽48)
Young adults
(n⫽67) F(1, 158); d, 95% CI
Vocabulary 13.77 (8.99) 21.31 (8.32) 13.69
ⴱⴱⴱ
; .94, [0.64, 1.24]
Plea comprehension 10.23 (2.38) 11.76 (2.05) 10.08
ⴱⴱ
; .69, [0.39, 0.98]
MacCAT Understanding 12.90 (2.09) 13.94 (1.83) 4.34
ⴱ
; .53, [0.24, 0.82]
MacCAT Reasoning 13.46 (2.14) 13.91 (1.71) 0.10; .36, [0.07, 0.65]
Note. MacCAT ⫽MacArthur Competence Assessment Tool–Criminal Adjudication; CI ⫽confidence inter-
val.
ⴱ
p⬍.05.
ⴱⴱ
p⬍.01.
ⴱⴱⴱ
p⬍.001.
0.08
-0.14
-0.03
0.14
-0.02
0.05
0.15
-0.1
-0.5 -0.3 -0.1 0.1 0.3 0.5
Plea Vocabulary
Plea Comprehension
MacCAT Understanding
MacCAT Reasoning
Innocent
Guilty
Figure 1. Correlations between plea knowledge measures and plea deci-
sion by guilt–innocence manipulation (controlling for age; n⫽115). 0 ⫽
not guilty plea; 1 ⫽guilty plea. See the online article for the color version
of this figure.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
618 REDLICH AND SHTEYNBERG

Table 4
Means and Standard Deviations by Age Group and Guilty–Innocent Manipulation for Plea Rationales for Subsample and Entire Sample
Plea rationale
Answered all manipulation checks correctly Entire sample
Juveniles
(n⫽23)
Adults
(n⫽37)
Guilty
(n⫽18)
Innocent
(n⫽42)
Juveniles
(n⫽49)
Adults
(n⫽61)
Guilty
(n⫽33)
Innocent
(n⫽77)
I chose to plead Not Guilty because....
I thought my chances to be found guilty at trial were not good 3.48 (1.12) 3.76 (0.86) 3.67 (0.97) 3.64 (0.98) 3.55 (1.10) 3.83 (0.96) 3.73 (1.04) 3.69 (1.05)
I thought the truth would come out at trial 3.87 (1.22) 3.70 (1.39) 2.61 (1.58) 4.26 (0.80) 3.81 (1.09) 3.77 (1.33) 2.94 (1.66) 4.01 (1.01)
I was thinking about what would happen in the near future 2.74 (1.14) 3.35 (1.38) 3.00 (1.28) 3.17 (1.34) 2.92 (1.37) 3.57 (1.34) 3.24 (1.35) 3.30 (1.41)
I did not think the plea offer was that good of a deal 3.09 (1.35) 3.19 (1.49) 3.44 (1.42) 3.02 (1.42) 2.76 (1.38) 3.18 (1.48) 3.27 (1.57) 2.87 (1.38)
I thought if I did not accept the plea offer, the DA would offer a better deal 2.09 (1.16) 1.57 (0.90) 2.33 (1.14) 1.52 (0.89) 2.39 (1.20) 1.85 (1.18) 2.67 (1.34) 1.84 (1.08)
I was thinking about the long-term consequences of pleading guilty 3.65 (1.19) 4.41 (0.93) 3.83 (1.15) 4.24 (1.06) 3.53 (1.46) 4.16 (1.24) 3.613 (1.37) 4.00 (1.37)
I did not want to admit guilt 3.13 (1.55) 3.43 (1.83) 2.84 (1.80) 3.48 (1.69) 2.82 (1.67) 3.13 (1.85) 2.76 (1.71) 3.09 (1.79)
Juveniles
(n⫽25)
Adults
(n⫽30)
Guilty
(n⫽44)
Innocent
(n⫽11)
Juveniles
(n⫽41)
Adults
(n⫽39)
Guilty
(n⫽58)
Innocent
(n⫽22)
I chose to plead Guilty because....
I did not want to risk losing at trial 4.00 (1.29) 4.40 (1.16) 4.23 (1.26) 4.18 (1.17) 3.95 (1.32) 4.38 (1.14) 4.22 (1.27) 4.00 (1.20)
I thought the plea offer was a good deal 4.04 (1.14) 4.20 (0.86) 4.23 (0.96) 3.73 (1.01) 4.05 (1.06) 4.18 (0.85) 4.23 (0.95) 3.82 (0.96)
I was thinking about what would happen in the near future 3.56 (1.45) 4.10 (0.89) 3.86 (1.17) 3.82 (1.33) 3.60 (1.37) 4.15 (0.84) 3.91 (1.12) 3.77 (1.31)
It was the right thing to do 2.56 (1.53) 3.30 (1.37) 3.27 (1.44) 1.73 (0.91) 2.85 (1.55) 3.44 (1.46) 3.44 (1.46) 2.36 (1.43)
I did not think my (public defender) lawyer would do a good job at trial 2.20 (1.41) 2.93 (1.51) 2.59 (1.58) 2.64 (1.21) 2.00 (1.24) 3.08 (1.49) 2.49 (1.57) 2.64 (1.18)
I was thinking about the long-term benefits of pleading guilty 3.72 (1.28) 4.43 (0.82) 4.09 (1.14) 4.18 (0.98) 3.70 (1.16) 4.49 (0.79) 4.11 (1.10) 4.05 (1.00)
I thought the judge would go easier on me 2.40 (1.44) 2.70 (1.42) 2.66 (1.45) 2.18 (1.17) 2.22 (1.33) 2.77 (1.40) 2.54 (1.45) 2.36 (1.22)
Note. All questions used scale of 1 (not true at all)to5(extremely true). The means that are bolded within rows indicate a significant difference by age group or the guilt–innocent manipulation
at p⬍.05. Italicized means indicate a difference at p⬍.10. Statistics, with effect sizes, are presented in the text.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
619
AGE AND PLEA DECISION MAKING

for the plea rationales, “I was thinking about what would happen
in the near future,” F(1, 75) ⫽7.91, p⫽.006, d⫽.48 (95% CI
[0.03, 0.92]), “I was thinking about the long-term benefits of
pleading guilty,” F(1, 75) ⫽12.01, p⫽.001, d⫽.79 (95% CI
[0.33, 1.24]), “I thought that the judge would go easier on me,”
F(1, 75) ⫽4.42, p⫽.039, d⫽.40 (95% CI [0.04, 0.84]), and “I
did not think my public defender lawyer would do a good job at
trial,” F(1, 75) ⫽6.35, p⫽.014, d⫽.79 (95% CI [0.33, 1.24]).
For this latter rationale, the interaction between age group and the
guilt manipulation was not significant (p⫽.15) as it was in the
subsample analyses. For all four significant main effects, young
adults were more likely to endorse the rationales than juveniles.
Discussion
In the present study, our main foci were determining how age
(juvenile vs. young adult) and guilt versus innocence influenced
decisions to plead guilty and the rationales underlying these deci-
sions. We were also interested in examining factors that interacted
with age and guilt status, including aspects of the plea offer and
plea label (i.e., comparative fairness), as well as general and
specific (plea) legal knowledge.
Age and Plea Decision Making
Based on developmental aspects of youth and previous research
on legal decision making and understanding, we posited three
alternate hypotheses about relations between age and the willing-
ness to plead guilty (see also Redlich, 2010b). First, it was possible
that juveniles would be more likely to plead guilty than adults.
Some prior research has shown that juveniles have higher admis-
sion rates than adults when guilty (Grisso et al., 2003), and when
innocent (e.g., Drizin & Leo, 2004;Redlich & Goodman, 2003).
Second, it was also possible that the opposite trend would present
itself (i.e., adults would have higher plea rates than juveniles)
because of known youthful characteristics related to risk-taking,
impulsivity, and deficient legal knowledge. Finally, a third possi-
bility was that no significant differences would be found by age, a
result found in previous research (Malloy et al., 2014;Tepfer et al.,
2010;Viljoen et al., 2005). Support for two of the three hypotheses
was found but depended on guilt status. Specifically, when juve-
niles and young adults in the present study were asked to imagine
they were guilty of the crime, willingness to plead guilty did not
significantly differ (in support of the third [null] hypothesis).
When asked to imagine innocence, however, juveniles were about
2.5 times more likely to plead guilty than adults (in support of the
first hypothesis). Although this effect was only significant at p⬍
.10, the effect size of .23 is considered small to medium (i.e., .10
is small, .30 is medium, Cohen, 1988). Thus, as it has been shown
for false confessions in the interrogation room (see Drizin & Leo,
2004;Redlich & Goodman, 2003), this research preliminarily
supports that youth are also more likely than adults to make false
admissions in the form of guilty pleas.
Our results lead to the question, “Why are juveniles more
susceptible to false guilty pleas than adults?” As discussed above,
juveniles are less mature— cognitively, socially, neurologically,
and emotionally—than adults (Cauffman & Steinberg, 2012;
Steinberg, 2005). This immaturity is known to associate with poor
and sometimes rash decision making, which can be construed as
pleading guilty when innocent (but see Bowers, 2008). Although
we did not find a significant relationship between impulsivity and
the plea decision (when either guilty or innocent), it may be that
our measure of this construct was not valid. To wit, we did not find
a significant difference by age group on this measure in contrast to
most other studies examining impulsivity in adolescents and adults
(e.g., see Steinberg et al., 2008).
Further, juveniles’ higher false guilty plea rates may be due to
legal knowledge, which juveniles in this study (and many others;
e.g., Grisso et al., 2003) were found to be lower on than adults.
Specifically, juveniles scored lower than young adults when tested
on plea comprehension, competence, and vocabulary. As discussed
by Drizin and Luloff (2007), false guilty pleas in juvenile court
may occur because “there is a good chance that children will
simply not understand that they are waiving their right to trial and
admitting guilt” (p. 293). Adjudicative competence, in particular,
relates directly to one’s ability to understand the conditions and
consequences of the plea decision, and to participate meaningfully
in one’s defense. Insofar as the ability to differentiate between
viable defenses is linked to legal knowledge, it may be that
juveniles are less likely than young adults to identify potentially
viable legal defenses stemming from their innocence (e.g., forensic
evidence, witnesses, and alibis), thereby increasing the likelihood
for youth to falsely plead guilty. Among those asked to assume
innocence and who pled guilty, juveniles and adults did not differ
in their endorsement of the rationale, “I did not think my public
defender would do a good job.” When asked to assume guilt,
however, adults were more likely than juveniles to endorse this as
a reason to accept the plea. Although this interaction effect (which
was not significant when rerun with the entire sample) does not
explain why juveniles were found to be more likely to falsely plead
guilty in comparison to adults, it does serve to demonstrate the age
group influences the rationales underlying the plea decision, when
guilty and when innocent. Our analyses concerning plea/legal
knowledge and the plea decision (which controlled for age) are
discussed in greater depth below.
Juveniles in the present study were also more likely to plead
guilty than adults when the plea offer included jail time, a pattern
which did not emerge when the plea offer did not include jail time.
More pointedly, whereas adults’ guilty plea rates went from 27%
when the plea offer included jail to 60% when it did not, juveniles’
plea rates remained somewhat steady across these conditions (55%
to 47%). This pattern of results suggests that youth are less likely
than adults to consider the consequences of their plea decisions,
which is consistent with findings from our plea rationale results
(see Table 4). Specifically, in comparison to the young adults,
youth were less likely to consider both short- and long-term
consequences/benefits, regardless of the plea decision itself (that
is, whether they pled not guilty or guilty). This is not to say,
however, that youth are not thinking about the future at all. For
example, Grisso and colleagues (2003) found that whereas young
adults in their sample considered the plea sentence length and the
chance of acquittal, youth tended to only consider the sentence
length, but nonetheless still weighed this factor. Further, in inter-
viewing juveniles transferred to criminal court, Daftary-Kapur and
Zottoli (2014) found that juveniles’ plea acceptances were “overly
influenced by short-term outcomes” (p. 333). In the present re-
search, juveniles were also significantly less likely than adults to
see their plea decision as voluntary when jail time was present (a
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
620 REDLICH AND SHTEYNBERG

pattern that did not emerge when jail time was absent), suggesting
that the consequences of the plea decision are affecting juveniles’
perceptions, if not the decision itself.
Our hypothesis concerning age and comparative fairness was
not supported. The comparative fairness manipulation, however,
did influence the perceived confidence of the plea decision by age
group. Juveniles, compared with adults, were significantly less
confident about their plea decision but only under certain condi-
tions (i.e., when innocent and the plea offer was described as really
good).
2
Additionally, regardless of age, the comparative fairness
manipulation did not influence plea acceptance rates, a finding at
odds with Tor et al. (2010). Tor and colleagues found plea accep-
tance rates of 59%, 44%, and 27% when the offers were described
as better than, similar to, or worse than typical offers, respectively
(Study 5, guilty condition). Why the present study did not replicate
this effect of comparative fairness needs further exploration. It
may be that the labels we used (i.e., “fair but not great”/“really
good” in comparison to what others get) were not salient or
meaningful enough to our participants. The labels are also ambig-
uous, in that, fair can mean “just” as well as “so-so,” which is a
limitation that should be avoided in future examinations.
Guilt–Innocence Status and Plea Decision Making
Research on what has been dubbed the “innocence effect”
(Gazal-Ayal & Tor, 2012) suggests that innocents may be more
confident and optimistic about their chance of a favorable outcome
rather than being wrongfully convicted, thereby leading defendants
to maintain their innocence because of a strong belief that the truth
of their innocence will come out at trial (see also, Kassin, 2005).
The present study finds support for this effect among both juve-
niles and adults, further supporting the notion that innocence can
influence defendant’s risk-perceptions of a trial conviction. Spe-
cifically, as anticipated, guilty participants were much more will-
ing (2.2 more times for juveniles, and 5.6 times for adults) to plead
guilty than innocent ones. This is a pattern consistently found in
previous research examining true and false guilty pleas (e.g.,
Bordens, 1984;Dervan & Edkins, 2013;Gregory et al., 1978).
We also presume that the pattern of guilty people pleading
guilty more than innocent people plays out in the real world,
though base rates of guilt and innocence for actual defendants is
unknown and perhaps unknowable. Covey (2013) compared
groups of guilty pleas in the scandalized Rampart, CA, case, which
is often referred to as a “mass exoneration case” (see Gross et al.,
2005). Within this case, Covey identified three groups: (a) inno-
cent (n⫽38); (b) guilty (n⫽27); and (c) “may be innocent” (n⫽
22). Guilty plea rates were 77%, 88%, and 89%, respectively for
the innocent, guilty, and “may be innocent” groups. Covey con-
cludes,
It thus appears from the data that actual innocence does induce some
defendants to refuse a guilty plea and hold out for trial, but that the
incentive has only a marginal effect, leading the innocent to contest
their cases at trial at an approximately 10% greater rate than those
who are actually guilty. (p. 1174)
Thus, although preliminary and in need of more research, there is
amassing evidence from social science and the real world that the
innocent plead guilty, and likely more often than is presently
known, but not as often as the guilty.
We had also expected that the magnitude of the difference
between guilty and innocent participants’ plea rates would be
especially large when the plea offer did not include jail time. When
the plea offer did not include jail time, plea acceptance rates
between guilty and innocent participants bridged a 68.8 percentage
point difference. When the plea offer did include jail, the differ-
ential remained significant, but was only a 34.4 percentage point
difference. More telling, however, was that the manipulation of jail
time influenced guilty participants’ plea decisions (though only
adults’ and not juveniles’), but did not influence innocent partic-
ipants’ decisions—a finding which supports the notion that inno-
cent and guilty defendants use different rationales for their deci-
sion making. Thus, unlike previous research, we found an
interaction between guilt–innocence status and plea offer. For
example, Gregory and colleagues (1978), in one of the first studies
examining false guilty pleas, did not find guilt–innocence to in-
teract with manipulations concerning the number of the charges or
the severity of the sentence if convicted at trial (see also, Dervan
& Edkins, 2013 for a similar null interaction effect). Bordens
(1984) found a marginally significant interaction between guilt
status and probability of conviction. In his study, guilty subjects
were just as willing to plead guilty if the probability of conviction
at trial was 50% or 90%, but significantly less willing when it was
10%. In contrast, innocent subjects showed an opposite pattern;
when the probability of conviction was 10% or 50%, innocents
were equally likely to refuse the plea offer, but when the proba-
bility was 90%, plea acceptance rates significantly increased. In
the present study, probability of conviction was not manipulated
but held constant at 75%. Whether the patterns found by Borden
would hold when the plea takers are juveniles is an open question,
as is whether the pattern of results found here would differ if
probability of conviction was manipulated.
Guilty and innocent participants also demonstrated differences
in their rationales underlying their decision to either plead guilty or
not guilty. Among those who pled not guilty, innocent participants
were more likely than guilty ones to do so because they thought the
truth would come out at trial. As established by Tor and colleagues
(2010), innocent participants are more apt to risk the uncertain
outcomes of a trial because the choice to plead guilty or risk trial
is seen as substantively unfair. Additionally, innocents often be-
lieve that their innocence will be readily apparent to others, and
thus have been found more likely (than the guilty) to waive their
Miranda rights (Kassin & Norwick, 2004) and even falsely confess
in interrogations (Kassin, 2005). Guilty participants who pled not
guilty were also more likely than innocent ones to think the plea
offer was not that good and to think that the DA would offer a
better deal (upon rejecting the first plea offer).
Among those who pled guilty, guilty participants were signifi-
cantly more likely to do so than innocent ones because “it was the
right thing to do.” This line of reasoning could be construed as
comparable to the moral-based criteria noted by Peterson-Badali
and Abramovitch (1993). Although these authors did not investi-
gate plea decisions when innocent, they found that the use of
guilt-based (moral) justifications was not influenced by the
2
When the entire sample was analyzed, juveniles were found to be
significantly less confident about their plea decision than adults regardless
of other conditions (i.e., no significant interaction effects emerged).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
621
AGE AND PLEA DECISION MAKING

strength of the evidence. They also found that juveniles were more
likely than adults to mention guilt as a basis for plea decisions. In
the present study, however, we did not find a significant main
effect of age or an age-guilt interaction effect for this rationale.
This lack of an effect may be because “the right thing to do” is
open to multiple interpretations, including a moral justification, as
well as one meaning the “accurate” or best option.
There were also some instances in which the rationales of guilty
and innocent individuals did not differ that are noteworthy as well.
For example, guilty and innocent participants who pled not guilty
did not differ in their endorsement of “I did not want to admit
guilt.” Thus, among those who chose to plead not guilty, there was
a reticence against admitting guilt. From a cost-benefit standpoint,
it makes sense that neither the innocent nor the guilty would want
to admit guilt, as an admission (plea) leads to a conviction and a
variety of other consequences. There was also a nonsignificant
difference by guilt-status among those who did opt for pleading
guilty for “I did not want to risk losing at trial.” That is, among
those who chose not to take their chances at trial, both the innocent
and guilty did so because they were risk-averse. Thus, again,
although innocent participants in general showed more risk pro-
clivity (in that they more often chose the trial option), innocents
who did not choose the trial option (i.e., pled guilty) demonstrated
risk aversion similar to the guilty participants (Gazal-Ayal & Tor,
2012).
Plea Comprehension
Another factor examined in the present study was general legal
and specific (to plea) understanding and appreciation. A consistent
and robust finding in the psycho-legal literature is that juveniles, in
comparison to adults, have deficiencies in legal knowledge (e.g.,
Grisso et al., 2003;Viljoen et al., 2005). Thus, we hypothesized—
and found—similar patterns. First, adults were better able to define
vocabulary words used in the tender-of-plea and judge’s colloquy
than juveniles. The tender-of-plea and judge’s colloquy required a
9th grade reading level, which has been found to be the average for
these forms (Redlich & Bonventre, 2015), and thus, because some
of the juveniles had not yet reached this grade, these findings are
perhaps unsurprising. However, it should be noted that both juve-
niles and adults demonstrated a lack of understanding. A perfect
score on the vocabulary measure was a 38; the average for adults
was about 21. It is important to note that all words were used in the
tender-of-plea form and the colloquy and participants were asked
multiple times if they understood or had questions. Words/phrases
that proved particularly difficult (more than 50% incorrect) for
juveniles and adults included: concession, restitution, motion, im-
partial, burden of proof, reasonable doubt, and unlawful influence
(results not shown). Similar to what Kaban and Quinlan (2004)
found in a sample of justice-involved juveniles, participants in our
sample were also unable to define key words and phrases neces-
sary to understand and appreciate what it means to plead guilty.
Further, Kaban and Quinlan found that even after receiving in-
struction about plea concepts and words, juveniles’ understanding
and knowledge remained poor.
Second, adults were also found to possess a better understanding
of general and plea-related concepts than juveniles. Regarding plea
comprehension, the assessment questions generally followed from
the tender-of-plea and the judges’ colloquy. On average, juveniles
were correct 64% of the time, and adults 74%. Although our
participants did not have an opportunity to interact with an attor-
ney, they were given repeated opportunities to ask questions and
receive clarification on aspects they did not understand; most
often, participants claimed to have no questions. Do the low scores
on this measure (as well as the vocabulary measure) indicate that
written tender-of-plea forms and oral colloquies are unreliable
assessments of whether plea decision making is knowing and
voluntary? Per Boykin v. Alabama (1969), judicial colloquies are
intended to ensure valid plea decision making, but there has been
little research to assess this intention (Redlich, in press). In con-
trast, most writings describe plea colloquies as “boilerplate” for-
malities (Bibas, 2011) that take “less time than it takes to get a
hamburger from a McDonald’s drive-through window” (Bo-
ruchowitz, Brink, & Dimino, 2009, p. 32). Findings from the
present research suggest that juvenile and adult defendants alike
are unlikely to appreciate much of what judges explain to them in
plea colloquies (see also Redlich et al., in press).
Finally, we also examined whether general and plea-specific
knowledge was associated with the plea decision. When collapsed
across guilt–innocence, knowledge was unrelated to decision mak-
ing. After computing the correlations separately for guilty and
innocent participants (as well as controlling for age), however,
divergent patterns emerged for relationships between plea knowl-
edge and the plea decision. When the entire sample was included
(which included those with potentially low intellect), significant,
positive correlations were found between competence understand-
ing and reasoning scores, vocabulary scores, and willingness to
plead, but only among subjects asked to assume guilt. Thus, we
found preliminary support indicating that plea knowledge and the
plea decision are related, though only under certain circumstances.
Implications for Practice and Research
Almost all adjudicated adult and juvenile defendants plead
guilty. Over the past 15 years, the National Juvenile Defender
Center (NJDC) has conducted more than 20 assessments of state
juvenile courts (see http://njdc.info/our-work/juvenile-indigent-
defense-assessments/). In doing so, they examined the plea pro-
cess, noting common problems across sites, such as high rates of
pleas, pleas occurring early in the process (leaving no time for
investigation, trust-building, or education), unknowledgeable pleas,
developmentally inappropriate, incomplete, and even inaccurate plea
colloquies, and systemic setting problems around culture and judge-
attorney interactions. For example, as noted in the Illinois NJDC
report (Crawford, Dohrn, Geraghty, Moss, & Puritz, 2007), “When
children enter admissions during the initial hearing, it is almost
always true that the attorney and the child lacked the opportunity
to engage in a meaningful discussion about the case and the
consequences of pleading guilty” (p. 2). Further, observers noted
that “many of the children looked bewildered or disengaged during
the plea colloquy” (p. 43), and that some youth left the courthouse
without awareness that they had just pled guilty. Both youth and
probation officers commonly reported intense pressure to plead,
which they attributed to insufficient time attorneys spent with
youth, attorneys’ failure to explore and/or understand the youths’
wishes, and lack of investigation into the case. These trends noted
by the NJDC assessments, in combination with results from the
present research, suggest that innocent juvenile defendants pro-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
622 REDLICH AND SHTEYNBERG

cessed in juvenile court, may be especially likely to falsely plead
guilty (see also Drizin & Luloff, 2007).
The present findings also have implications for the growing
literature on juveniles’ capabilities and decision-making processes
in the legal system. Comprehensive studies have assessed juve-
niles’ abilities to understand and appreciate their roles as defen-
dants (e.g., Grisso et al., 2003) and their constitutional rights (e.g.,
Viljoen et al., 2005). In addition to replicating the findings that
juveniles have a depreciated understanding and appreciation of
legal concepts (both general and specific to pleas), our findings
also demonstrate that such a decreased understanding may influ-
ence decision making over and above chronological age. Further,
the present research did not only examine whether juveniles are at
increased risk for true and false guilty pleas, but also examined
why. Our plea rationale findings, on the one hand, indicate that
juveniles are less likely than adults to consider the short- and
long-term consequences in their decisions to plead guilty and not
guilty. On the other hand, we did not find age to influence
endorsement of other plea rationales, including ones relating to
morality and risk-taking.
Conclusions and Limitations
Some limitations of the present study must be acknowledged. First,
an oft-cited limitation of experimental lab studies is that asking
individuals to imagine they are suspects of a crime, where either
guilty or innocent, does not generalize to actually being accused.
Although it is certainly true that an experiment does not take into
account all of the real-world implications, it is important to note that
despite the low stakes situation encountered in this study, one out of
every five innocent individuals (or more depending on certain factors)
falsely pled guilty instead of going to court to affirm their innocence.
It was also the differences by condition that were noteworthy here, as
the goal was not to establish prevalence rates of true and false guilty
pleas. Further, we used an imagined situation in lieu of deceiving
subjects. Most experimental studies examining willingness to plead,
especially those involving juveniles, have used such methods (e.g.,
Grisso et al., 2003;Peterson-Badali & Abramovitch, 1993;Tor et al.,
2010). Ultimately, this study is a starting point and further research
should be conducted in the field by interviewing juveniles and adults
who recently entered a plea decision (see Daftary-Kapur & Zottoli,
2014).
A second limitation of this research is that the reduced sample
sizes may not have provided sufficient power to detect statistically
significant differences, particularly in the main analyses when only
those who answered all four manipulation check questions were
included (although many findings remained the same when the
entire sample was included). Significant effects may not have been
detected as a result of a smaller nwithin cells; this was especially
problematic with the closed-ended rationale questions as partici-
pants who pled guilty and not guilty were asked separate questions
(thereby further reducing the n). Third, the plea offer manipulation
was intended to examine plea decisions when jail time was and
was not included (with probation time as a constant). However,
this manipulation may have confounded sentence length. Specifi-
cally, the no-jail condition was 5 years’ probation, whereas the
some-jail condition was 5 years’ probation and two years in jail.
Thus, whether adults’ decreased willingness to plead guilty in the
some-jail condition (vs. the no-jail condition) was due to the
presence of custodial time or the sentence length itself is unclear.
Although previous research has demonstrated that any amount of
incarceration, regardless of length, increases willingness to plead
guilty in comparison to probation (Bordens, 1984), future research
should be conducted to address this potential confound.
Despite these limitations, the present study is one of few to inves-
tigate age-related differences in true and false guilty pleas, and the
only, to our knowledge, to do so experimentally. Youth status has
been a consistently identified risk factor for false confessions in the
interrogation room (Kassin et al., 2010;Owen-Kostelnik et al., 2006).
Findings from the present research indicate that youthful status may
also be a risk factor for false admissions in the form of false guilty
pleas. Given the extremely high rate of convictions via pleas in our
criminal justice system, and the increasing identification of wrongful
convictions, it is important for researchers and practitioners to better
understand the complexities of the guilty plea decision among juve-
nile and adult defendants.
References
Barratt, E. S. (1994). Impulsiveness and aggression. In J. Monahan & H. J.
Steadman (Eds.), Violence and mental disorder: Developments in risk
assessment (pp. 61–79). Chicago, IL: University of Chicago Press.
Bibas, S. (2011). The myth of the fully informed rational actor. Saint Louis
University Public Law Review, XXXI, 79 – 86. Retrieved from https://
www.law.upenn.edu/cf/faculty/sbibas/workingpapers/!31StLouis
ULRev79(2011).pdf
Bonnie, R. J. (1992). The competence of criminal defendants: A theoretical
reformulation. Behavioral Sciences & the Law, 10, 291–316. http://dx
.doi.org/10.1002/bsl.2370100303
Bordens, K. S. (1984). The effects of likelihood of conviction, threatened
punishment, and assumed role on mock plea bargaining decisions. Basic
and Applied Social Psychology, 5, 59 –74. http://dx.doi.org/10.1207/
s15324834basp0501_4
Bordens, K. S., & Bassett, J. (1985). The plea bargaining process from the
defendant’s perspective: A field investigation. Basic and Applied Social
Psychology, 6, 93–110. http://dx.doi.org/10.1207/s15324834basp
0602_1
Boruchowitz, R. C., Brink, M., & Dimino, M. (2009). Minor crimes,
massive waste: The terrible toll of America’s broken misdemeanor
courts. Retrieved from https://www.nacdl.ord/reports/misdemeanor/
Bowers, J. (2008). Punishing the innocent. University of Pennsylvania Law
Review, 156, 1117–1179. Retrieved from https://www.law.upenn.edu/
journals/lawreview/articles/volume156/issue5/Bowers156U.Pa.L.Rev
.1117(2008).pdf
Boykin v. Alabama, 395 U.S. 238 (1969). Retrieved from https://supreme
.justia.com/cases/federal/us/395/238/case.html
Bureau of Justice Statistics. (2013). Felony defendants in large urban
counties, 2009. Retrieved from http://www.bjs.gov/content/pub/pdf/
fdluc09.pdf
Cauffman, E., & Steinberg, L. (2012). Emerging findings from research on
adolescent development and juvenile justice. Victims & Offenders, 7,
428 – 449. http://dx.doi.org/10.1080/15564886.2012.713901
Cohen, J. (1988). Statistical power and analysis for the behavioral sciences
(2nd ed.). Hillsdale, NJ: Erlbaum.
Cohen, T. H., & Reaves, B. A. (2006). Felony defendants in large urban
counties, 2002. Washington, DC: U. S. Department of Justice, Office of
Justice Program, Bureau of Justice Statistics. Retrieved from http://
www.bjs.gov/content/pub/pdf/fdluc06.pdf
Covey, R. (2013). Police misconduct as a cause of wrongful convictions.
Washington University Law Review, 90, 1133–1189. Retrieved from
http://papers.ssrn.com/sol3/papers.cfm?abstract_id⫽2349872
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
623
AGE AND PLEA DECISION MAKING

Crawford, C., Dohrn, B., Geraghty, T. F., Moss, M. B., & Puritz, P. (2007).
Illinois: An assessment of access to counsel and quality of representa-
tion in delinquency proceedings. Retrieved from http://www.njdc.info/
pdf/illinois_assessment.pdf
Cutler, B. (Ed.), (2011). Conviction of the innocent: Lessons from psycho-
logical research. Washington, DC: American Psychological Associa-
tion.
Daftary-Kapur, T., & Zottoli, T. (2014). A first look at the plea deal
experiences of juveniles tried in adult court. International Journal of
Forensic Mental Health, 13, 323–336. http://dx.doi.org/10.1080/
14999013.2014.960983
Dale, E., & O’Rourke, J. (1981). The living word vocabulary. Chicago, IL:
World Book/Childcraft International.
Dervan, L., & Edkins, V. (2013). The innocent defendant’s dilemma: An
innovative empirical study of plea bargaining’s innocence problem. The
Journal of Criminal Law & Criminology, 103, 1– 48. Retrieved from
http://scholarlycommons.law.northwestern.edu/jclc/vol103/iss1/1/
Drizin, S. A., & Leo, R. A. (2004). The problem of false confessions in the
post-DNA world. North Carolina Law Review, 82, 891–1008. Retrieved
from http://papers.ssrn.com/sol3/papers.cfm?abstract_id⫽1134094
Drizin, S. A., & Luloff, G. (2007). Are juvenile courts breeding grounds for
wrongful convictions? North Kentucky Law Review, 34, 257–322. Re-
trieved from http://chaselaw.nku.edu/academics/lawreview/archives
.html
Dusky v. U. S., 362 U.S. 402 (1960).
Garner, B. (Ed.). (2014). Black’s law dictionary: Deluxe (10th ed.). Eagan,
MN: Thomson West.
Garrett, B. L. (2008). Judging innocence. Columbia Law Review, 100,
101–190. Retrieved from http://connection.ebscohost.com/c/articles/
28626628/judging-innocence
Gazal-Ayal, O., & Tor, A. (2012). The innocence effect. Duke Law
Journal, 62, 339 – 401. Retrieved from http://dlj.law.duke.edu/article/
the-innocence-effect/
Gregory, W. L., Mowen, J. C., & Linder, D. E. (1978). Social psychology
and plea bargaining: Applications, methodology, and theory. Journal of
Personality and Social Psychology, 36, 1521–1530. http://dx.doi.org/10
.1037/0022-3514.36.12.1521
Grisso, T. (1981). Juvenile’s waiver of rights: Legal and psychological
competence. New York, NY: Plenum Press. http://dx.doi.org/10.1007/
978-1-4684-3815-4
Grisso, T. (2007). Adolescents’ decision-making: A developmental per-
spective on constitutional provisions in delinquency cases. New England
Journal on Criminal and Civil Confinement, 32, 3–14. Retrieved from
http://heinonline.org/HOL/LandingPage?handle⫽hein.journals/
nejccc32&div⫽3&id⫽&p
Grisso, T., Steinberg, L., Woolard, J., Cauffman, E., Scott, E., Graham, S.,
. . . Schwartz, R. (2003). Juveniles’ competence to stand trial: A
comparison of adolescents’ and adults’ capacities as trial defendants.
Law and Human Behavior, 27, 333–363. http://dx.doi.org/10.1023/A:
1024065015717
Gross, S. R., Jacoby, K., Matheson, D. J., Montgomery, N., & Patil, S.
(2005). Exonerations in the United States 1989 through 2003. The
Journal of Criminal Law & Criminology, 95, 523–560. Retrieved from
http://scholarlycommons.law.northwestern.edu/cgi/viewcontent
.cgi?article⫽7186&context⫽jclc
Kaban, B., & Quinlan, J. C. (2004). Trends and developments in the
juvenile court: Rethinking a “knowing, intelligent and voluntary waiver”
in Massachusetts’ juvenile courts. Journal of the Center for Families,
Children, and the Courts, 5, 35–55. Retrieved from http://www.courts
.ca.gov/documents/JournalVol5.pdf
Kassin, S. M. (2005). On the psychology of confessions: Does innocence
put innocents at risk? American Psychologist, 60, 215–228.
Kassin, S. M., Drizin, S. A., Grisso, T., Gudjonsson, G. H., Leo, R. A., &
Redlich, A. D. (2010). APLS-Approved White Paper: Police-induced
confessions: Risk factors and recommendations. Law and Human Be-
havior, 34, 3–38.
Kassin, S. M., & Norwick, R. J. (2004). Why people waive their Miranda
rights: The power of innocence. Law and Human Behavior, 28, 211–
221. http://dx.doi.org/10.1023/B:LAHU.0000022323.74584.f5
Landes, W. (1971). An economic analysis of the courts. Journal of Labor
Economics, 14, 61–107.
Lynch, T. (2003, Fall). The case against plea bargaining. Regulation,
24 –27. Retrieved from http://object.cato.org/sites/cato.org/files/serials/
files/regulation/2003/10/v26n3-7.pdf
Malloy, L. C., Shulman, E. P., & Cauffman, E. (2014). Interrogations,
confessions, and guilty pleas among serious adolescent offenders. Law
and Human Behavior, 38, 181–193. http://dx.doi.org/10.1037/lhb
0000065
Monahan, J., Steadman, H. J., Appelbaum, P. S., Robbins, P. C., Mulvey,
E. P., Silver, E.,...Grisso, T. (2000). Developing a clinically useful
actuarial tool for assessing violence risk. The British Journal of Psychi-
atry, 176, 312–319. http://dx.doi.org/10.1192/bjp.176.4.312
Norris, R., & Redlich, A. D. (2011). At-risk populations under investiga-
tion and at trial. In B. Cutler (Ed.), Conviction of the innocent: Lessons
from psychological research (pp. 13–32). Washington, DC: American
Psychological Association.
Oppel, R. A. (2011, September 25). Sentencing shift gives new leverage to
prosecutors. The New York Times. Retrieved from http://www.nytimes
.com/2011/09/26/us/tough-sentences-help-prosecutors-push-for-plea-
bargains.html?pagewanted⫽all&_r⫽2&
Otto, R. K., Poythress, N. G., Nicholson, R. A., Edens, J. F., Monahan, J.,
Bonnie, R. J.,...Eisenberg, M. (1998). Psychometric properties of the
MacArthur Competence Assessment Tool–Criminal Adjudication. Psy-
chological Assessment, 10, 435– 443. http://dx.doi.org/10.1037/1040-
3590.10.4.435
Owen-Kostelnik, J., Reppucci, N. D., & Meyer, J. R. (2006). Testimony
and interrogation of minors: Assumptions about maturity and morality.
American Psychologist, 61, 286 –304. http://dx.doi.org/10.1037/0003-
066X.61.4.286
Patton, J. H., Stanford, M. S., & Barratt, E. S. (1995). Factor structure of
the Barratt impulsiveness scale. Journal of Clinical Psychology, 51,
768 –774. http://dx.doi.org/10.1002/1097-4679(199511)51:6⬍768::
AID-JCLP2270510607⬎3.0.CO;2-1
Peterson-Badali, M., & Abramovitch, R. (1993). Grade related changes in
young people’s reasoning about plea decisions. Law and Human Behav-
ior, 17, 537–552.
Poythress, N. G., Nicholson, R., Otto, R. K., Edens, J. F., Bonnie, R. J.,
Monahan, J., & Hoge, S. K. (1999). The MacArthur Competence As-
sessment Tool–Criminal Adjudication. Odessa, FL: Psychological As-
sessment Resources.
Rakoff, J. S. (2014, November). Why innocent people plead guilty. The
New York Review of Books. Retrieved from http://www.nybooks.com/
articles/archives/2014/nov/20/why-innocent-people-plead-guilty/
Redlich, A. D. (2010a). False confessions, false guilty pleas: Similarities
and differences. In G. D. Lassiter & C. Meissner (Eds.), Interrogations
and confessions: Current research, practice, and policy (pp. 49 – 66).
Washington, DC: American Psychological Association. http://dx.doi
.org/10.1037/12085-003
Redlich, A. D. (2010b). The susceptibility of juveniles to false confessions
and false guilty pleas. Rutgers Law Review, 62, 943–957.
Redlich, A. D. (in press). The validity of pleading guilty. In M. K. Miller
& B. Bornstein (Eds.), Advances in psychology and law. New York, NY:
Springer.
Redlich, A. D., Bibas, S., Edkins, V., & Madon, S. (in press). The
psychology of defendant plea decision making. American Psychologist.
Redlich, A. D., & Bonventre, C. L. (2015). Content and comprehensibility
of juvenile and adult tender-of-plea forms: Implications for knowing,
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
624 REDLICH AND SHTEYNBERG

intelligent, and voluntary guilty pleas. Law and Human Behavior, 39,
162–176. http://dx.doi.org/10.1037/lhb0000118
Redlich, A. D., & Goodman, G. S. (2003). Taking responsibility for an act
not committed: The influence of age and suggestibility. Law and Human
Behavior, 27, 141–156. http://dx.doi.org/10.1023/A:1022543012851
Redlich, A. D., & Summers, A. (2012). Voluntary, knowing, and intelligent
pleas: Understanding plea inquiries. Psychology, Public Policy, and
Law, 18, 626 – 643. http://dx.doi.org/10.1037/a0026066
Rhodes, W. (1979). Plea bargaining: Its effect on sentencing and convic-
tions in the District of Columbia. The Journal of Criminal Law &
Criminology, 70, 360 –375. http://dx.doi.org/10.2307/1142579
Sanborn, J. B. (1992). Pleading guilty in juvenile court: Much ado about
something very important to young defendants. Justice Quarterly, 9,
127–150.
Saywitz, K., Jaenicke, C., & Camparo, L. (1990). Children’s knowledge of
legal terminology. Law and Human Behavior, 14, 523–535. http://dx.doi
.org/10.1007/BF01044879
Schmidt, M. G., Reppucci, N. D., & Woolard, J. L. (2003). Effectiveness
of participation as a defendant: The attorney–juvenile client relationship.
Behavioral Sciences & the Law, 21, 175–198. http://dx.doi.org/10.1002/
bsl.532
Singleton, L. C. (2007). Say “Pleas”: Juveniles’ competence to enter plea
agreements. Journal of Law and Family Studies, 9, 439 – 455. Retrieved
from https://litigation-essentials.lexisnexis.com/webcd/app?action⫽
DocumentDisplay&crawlid⫽1&srctype⫽smi&srcid⫽3B15&doctype⫽
cite&docid⫽9⫹J.⫹L.⫹Fam.⫹Stud.⫹439&key⫽dcdf9d9aa08458fff68
b50753ddcab60
Smith, A. (2007). I ain’t taking no plea: The challenge in counseling young
people facing serious time. Rutgers Law Review, 60, 11–31. Available
from http://scholarship.law.georgetown.edu/facpub/875
Smith, D. A. (1986). The plea bargaining controversy. The Journal of Criminal
Law & Criminology, 77, 949 –968. http://dx.doi.org/10.2307/1143445
Steinberg, L. (2005). Cognitive and affective development in adolescence.
Trends in Cognitive Sciences, 9, 69 –74. http://dx.doi.org/10.1016/j.tics
.2004.12.005
Steinberg, L., Albert, D., Cauffman, E., Banich, M., Graham, S., &
Woolard, J. (2008). Age differences in sensation seeking and impulsivity
as indexed by behavior and self-report: Evidence for a dual systems
model. Developmental Psychology, 44, 1764 –1778. http://dx.doi.org/10
.1037/a0012955
Steinberg, L., & Scott, E. (2003). Less guilty by reason of adolescence:
Developmental immaturity, diminished responsibility, and the juvenile
death penalty. American Psychologist, 58, 1009 –1018. http://dx.doi.org/
10.1037/0003-066X.58.12.1009
Tepfer, J. A., Nirider, L. H., & Tricarico, L. M. (2010). Arresting development:
Convictions of innocent youth. Rutgers Law Review, 62, 887–941. Re-
trieved from http://pegasus.rutgers.edu/~review/vol62n4/Tepfer.pdf
Tobey, A., Grisso, T., & Schwartz, R. (2000). Youths’ trial participation as
seen by youths and their attorneys: An exploration of competence-based
issues. In T. Grisso & R. G. Schwartz (Eds.), Youth on trial: A devel-
opmental perspective on juvenile justice (pp. 225–242). Chicago, IL:
University of Chicago Press.
Tor, A., Gazal-Ayal, O., & Garcia, S. M. (2010). Fairness and willingness
to accept plea bargain offers. Journal of Empirical Legal Studies, 7,
97–116. http://dx.doi.org/10.1111/j.1740-1461.2009.01171.x
Viljoen, J. L., Klaver, J., & Roesch, R. (2005). Legal decisions of
preadolescent and adolescent defendants: Predictors of confessions,
pleas, communication with attorneys, and appeals. Law and Human
Behavior, 29, 253–277. http://dx.doi.org/10.1007/s10979-005-
3613-2
Weimer, D. L. (1978). Plea bargaining and the decision to go to trial: The
application of rational choice model. Policy Sciences, 10, 1–24. http://
dx.doi.org/10.1007/BF00205890
Appendix
Hypothetical Scenario
Instructions
Please read the following, while I read it aloud:
Imagine you are a defendant in a criminal case. You have been
charged with Robbery in the 1st degree. A jewelry store in your
neighborhood was broken into and the employees there were
held at gunpoint. Glass cases were smashed and approximately
$5,000 worth of jewelry was taken. No one was hurt.
Imagine that you are guilty [innocent] of this crime and that you
did [not] rob the jewelry store. The jewelry store had a hidden
video camera which captured the crime on tape. You [The
perpetrator] wore a hooded sweatshirt and sunglasses, and thus
it’s hard to be sure who is on the tape. There is also an eyewit-
ness who has identified you as a person who was hanging around
outside the store right before the robbery occurred. However, the
police are certain that it is you, and it actually is you [but
actually it is not you. This is a case of mistaken identity].
You have a public defender lawyer who has told you that you
can go to trial or accept a plea offer from the District Attorney.
Your lawyer has told you that if you go to trial, there is about a
75% chance that the jury will find you guilty.
The District Attorney’s plea offer is to plead guilty to attempted
Robbery in the 3rd degree. If you accept this offer, you will have
to spend 5 years on probation [and 2 years in the county jail/
juvenile detention]. If you are found guilty at trial, your sentence
will be 5–7 years in state prison.
Your lawyer tells you that in comparison to others charged with
this crime, this plea offer is really good [fair but not great].
In order to make an informed decision about whether to accept
the plea offer or go to trial, you will now be handed a tender-of-
plea form. This form reviews your rights and the conditions and
consequences of accepting the plea offer. When you are done
reviewing the form, you will be asked to make a decision about
whether to accept the DA’s plea offer or go to trial.
Received June 22, 2015
Revision received May 23, 2016
Accepted May 23, 2016 䡲
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
625
AGE AND PLEA DECISION MAKING