Content uploaded by Christian A. Meissner
Author content
All content in this area was uploaded by Christian A. Meissner on Jul 22, 2015
Content may be subject to copyright.
Trevaskes S (2010) The shifting sands of punishment in the
era of ‘harmonious society’. Law Policy 32:332–361
Wang ZF (1989) (ed) Theory and practice of comprehen-
sive management of public order in China. Masses
Press, Beijing (in Chinese)
Welsh B, Hoshi A (2002) Communities and crime preven-
tion. In: Sherman L, Farrington DP, Welsh BC,
MacKenzie DL (eds) Evidence-based crime preven-
tion. Routledge, London
Wilson AM, Greenblatt SL, Wilson RW (eds)
(1977) Deviance and social control in Chinese society.
Praeger, New York
Yang YM (2010) Analyzing the fourth wave of the strike
hard policy. Gov Leg Syst 36:22–23 (in Chinese)
Zhang L, Messner SF, Liu J (2007a) A multilevel analysis
of the risk of household burglary in the city of Tianjin,
China. Brit J Criminol 47:918–937
Zhang L, Messner S, Liu J (2007b) Criminological research
in contemporary China: challenges and lessons learned
from a large-scale criminal victimization survey. Int
J Offender Ther Comp Criminol 51:110–121
Zhao J (2005) Conflicts between advantages and disadvan-
tages in restorative justice. Law 5:113–115 (in Chinese)
Zhao R, Liu J (2011) A system’s approach to crime pre-
vention: the case of Macao. Asian J Criminol 6:207–227
Zhong L (2008) Communities, crime and social capital in
contemporary China. Willan, Cullompton
Zhuo Y (2012) Social capital and satisfaction with crime
control in urban China. Asian J Criminol 7:121–136
Intervention Research
▶Bullying Prevention: Assessing Existing Meta-
Evaluations
Interview and Interrogation
Methods Effects on Confession
Accuracy
Christopher E. Kelly
1
, Allison D. Redlich
1
,
Jacqueline R. Evans
2
and Christian A. Meissner
3
1
University at Albany, State University of
New York, Albany, NY, USA
2
The University of Texas at Tyler, Tyler,
TX, USA
3
The University of Texas at El Paso, El Paso,
TX, USA
Synonyms
Confessions;Criminal interviewing;Interroga-
tion;Meta-analysis
Overview
The interviewing and interrogation of suspects is
important to securing convictions against the
guilty and freeing the wrongly accused. There
are two general methods of questioning suspects:
information gathering and accusatorial.
The information-gathering approach, used in the
United Kingdom, New Zealand, Australia,
and elsewhere, is characterized by rapport
building, truth seeking, and listening. The
accusatorial approach, used primarily in the
United States and Canada, is characterized
by accusation, confrontation, psychological
manipulation, and the disallowing of denials.
Academics and practitioners hotly debate which
method is more effective, particularly in light of
increased awareness of the problems with
false confessions. Two separate but related
meta-analyses were conducted to address this
question. The first relied upon data from five
observational field studies and the second from
12 experimental, laboratory-based studies. The
primary outcome measures were true and
false confessions. Results revealed that
both information-gathering and accusatory
methods were associated with the
production of confessions in field studies,
though the experimental data indicated that the
information-gathering method was more
diagnostic in that it increased the likelihood of
true confessions while reducing the likelihood
of false confessions. Although continued
research is needed to better understand
the methods that produce true and false
confessions, the results of these meta-analyses
suggest that the information-gathering
approach may be more effective in comparison
to the accusatorial approach.
Introduction
The elicitation of false confessions is an interna-
tional problem that has been documented in
almost every continent (Kassin et al. 2010;
Lassiter and Meissner 2010). Two general
factors have been linked to the incidence of
Interview and Interrogation Methods Effects on Confession Accuracy 2673 I
I
false confessions: personal (psychological)
vulnerabilities of the individual and the
use of accusatorial (psychologically based)
interrogative methods. Accusatorial methods
are common practice in the United States,
Canada, and many Asian nations, and the use
of these methods has been linked to false
confessions (Kassin et al. 2010; Lassiter and
Meissner 2010). In response to the increased
awareness of the prevalence of false confession,
the United Kingdom, Norway, New Zealand,
Australia, and several other countries have
amended their interrogation practices to employ
information-gathering methods of interrogation
(Bull and Soukara 2010). This entry summarizes
the research evidence on the relative
diagnosticity of the accusatorial interrogation
and information-gathering interview approaches.
Diagnosticity is the ratio of true to false confes-
sions, with an effective interview/interrogation
method maximizing true confessions while
minimizing false confessions.
This entry is based on the systematic
review conducted for and published by
the Campbell Collaboration (http://www.
campbellcollaboration.org) (Meissner et al.
2011). As will be shown below, the existing
research evidence suggests that although both
accusatorial and information-gathering
approaches increase the likelihood of obtaining
a confession in the field context, experimental
laboratory studies demonstrated that the informa-
tion-gathering approach yielded more diagnostic
confessions (by increasing true confessions and
reducing false confessions) when compared with
the accusatorial method.
Key Issues and Controversies
Generally speaking, two distinct methods of
suspect interview and interrogation approaches
have emerged: information gathering and
accusatorial. The information-gathering
approach to interviewing and interrogation
is based on establishing rapport between the
interviewer and suspect and seeks to elicit
information rather than obtain confessions. The
interviewer uses positive confrontation and
an open-ended questioning approach to elicit
information from the suspect; psychological
manipulation is anathema to the goals of
the interview. This approach is exemplified and
codified in Great Britain’s adoption of the Police
and Criminal Evidence (PACE) Act of 1984 and
later modified as the PEACE model in 1993.
The PACE Act explicitly prohibited the use of
psychologically manipulative techniques and
required the recording of all custodial interroga-
tions. Furthermore, the PEACE model focuses on
developing rapport, explaining the allegation and
the seriousness of the offense, emphasizing the
importance of honesty and truth gathering,
and requesting the suspect’s version of events.
Suspects are permitted to explain the situation
without interruption, and questioners are
encouraged to actively listen. Finally, the
information-gathering approach relies on
cognitive-based strategies to detect deception.
In contrast, the accusatorial method is typified
by the United States model and is best illustrated
by the “Reid Technique” (Inbau et al. 2001).
In this method, the interviewer establishes
control over and confronts the suspect, presumes
guilt, and ultimately aims to obtain a confession.
As distinct from the information-gathering
approach, accusatorial methods use closed-ended
and confirmatory questions and employ psycho-
logical manipulation which generally involve
three components: (a) custody and isolation,in
which the suspect is detained in a small room
and left to experience the anxiety, insecurity,
and uncertainty associated with police interroga-
tion; (b) confrontation, in which the suspect is
presumed guilty and told (sometimes falsely)
about the evidence against him/her, is warned of
the consequences associated with his/her
guilt, and is prevented from denying his/her
involvement in the crime; and finally (c) minimi-
zation, in which a now sympathetic interrogator
attempts to gain the suspect’s trust, offers the
suspect face-saving excuses or justifications for
the crime, and implies more lenient consequences
should the suspect provide a confession.
Lastly, this approach relies on anxiety-based
cues (e.g., verbal, nonverbal, and paralinguistic)
I2674 Interview and Interrogation Methods Effects on Confession Accuracy
in order to detect when suspects are being
deceptive. Table 1 summarizes the differences
between the information-gathering and
accusatorial approaches along four dimensions.
Meta-Analysis Methods
Researchers have relied on two primary methods
of examining the influence of questioning
technique on confessions: field studies of actual
interrogations and experimental laboratory
studies with mock crimes and transgressions.
Whereas the field studies offer greater external
validity, it is not possible to determine the
“ground truth” or the veracity of confessions
obtained in this context. In contrast, experimental
studies allow for the manipulation of guilt
and innocence but are limited in their ability
to precisely model the context of a criminal
interrogation. As such, two systematic reviews
and meta-analyses of the accusatorial and
information-gathering literature were conducted,
one examining field studies and the other
examining experimental studies.
The first meta-analysis focused on observa-
tional field studies; the second, on experimental
laboratory-based studies. Although briefly sum-
marized here, readers interested in more details
regarding the methods are referred to Meissner
et al. (2011). To identify studies that might be
included in the analyses, more than a dozen
reference databases were searched (using over
20 search terms), the reference sections of several
comprehensive sources on interviewing and
interrogation (books, articles, and reports) were
reviewed, and researchers in the United States
and overseas were contacted for any unpublished
research. To be eligible for inclusion in the meta-
analyses, field and laboratory studies must have
met the following criteria:
•Intervention: Eligible field studies must
have coded or quantified the use of at
least one interview or interrogation
technique, which was then categorized
into information-gathering, accusatorial,
or general interrogation approaches.
Eligible laboratory studies must have
involved the experimental manipulation of
information-gathering and/or accusatorial
methods with one another or with a control
interview method (e.g., a simple request for
compliance).
•Outcomes: Eligible field studies must have
reported the analysis of confessions and
have had sufficient quantitative data to
calculate effect sizes, specifically including
the relationship between the use of
certain interview/interrogation methods and
elicitation of a confession. Laboratory studies
must have reported the number of true and/or
false confessions from “guilty” or “innocent”
participants, respectively, in a manner that
would permit calculation of an effect size.
•Population/samples: The population of
interest for both the field and experimental
studies was criminal (mock) suspects of any
age, nationality, or status who were accused of
committing a criminal act, a transgression, or
withholding important information.
The initial search identified over 2,000
studies, though the vast majority was screened
out. After this preliminary screening, 33 field
studies and 22 laboratory studies were coded to
determine final eligibility, of which five field
studies and 12 laboratory studies were ultimately
included in the meta-analyses. All included
and excluded field and experimental studies
(as well as the reasons for exclusion) can be
found in Meissner et al. (2011).
Interview and Interrogation Methods Effects on
Confession Accuracy, Table 1 Five distinguishing
dimensions of information-gathering and accusatorial
interrogation approaches
Dimension
Information
gathering Accusatorial
Interpersonal
dynamic
Rapport building Control oriented
Approach Direct, positive
confrontation
Psychological
manipulation
Question type Open-ended,
exploratory
Closed-ended
confirmatory
Primary goal Elicit information Obtain confession
Clues to
deception
Cognitive Anxiety
Interview and Interrogation Methods Effects on Confession Accuracy 2675 I
I
Results of the Meta-analyses
Field Studies
The five field studies were identified through the
systematic search, three of which were conducted
in the United Kingdom, one in Canada, and one in
the United States. These five studies represented
data from 608 interrogation sessions. Of
these sessions, 39.4 % were coded from audio
recordings, 36 % from video recordings, 20 %
were live interrogation sessions, and the
remaining 4.4 % were from transcriptions of the
sessions. The alleged crimes for which the
suspects were being interrogated varied from
social security benefit fraud to serious violent
crime such as murder and robbery. Between
39 % and 64 % of the suspects in the interrogation
sessions confessed or offered some admission of
guilt. The studies classified the interrogations as
accusatory, information-gathering, or combined
interrogative method, resulting in eight
possible pair-wise comparisons across the five
study samples.
The effect size was Hedges’ gbased on
dichotomous outcomes (i.e., confess vs. not
confess) using the Cox computation method
(see Lipsey and Wilson 2001;Sa
´nchez-Meca
et al. 2003). The relationship between
the use of certain interrogative methods
(accusatorial, information-gathering, or general
interrogative methods) and elicitation of
a confession was the focus of the analysis. All
analyses were based on a random-effects model,
and the results are presented in Table 2.
Three field studies assessed the relationship
between use of accusatorial methods and
elicitation of a confession in a real-world context
and showed that such a method was associated
with a large and significant increase in confession
rates. Furthermore, there was no significant
degree of variability across the studies suggesting
consistency in findings across studies.
The relationship between information-
gathering methods and elicitation of
a confession was assessed through two field
studies. These studies also found that the use of
such a method was associated with a large and
significant increase in confession rates. The
results varied meaningfully between these two
studies, however, with a substantially smaller
effect in one study (.38 vs. 1.04).
A number of tactics observed in these studies
could reasonably be coded as a part of
accusatorial and information-gathering
approaches. Three field studies examined the
influence of these combined methods in eliciting
confessions in a real-world context as opposed to
those methods that might be exclusively linked
to either accusatorial or information-gathering
approaches, and showed no significant
relationship between the use of these general
methods and confession statements provided
by suspects. However, these results were
highly variable across studies, ranging from
a low of .40 to a high of .79.
Based on these field studies, it appears that the
use of accusatorial and information-gathering
methods of interrogation was significantly
associated with the elicitation of confession evi-
dence in a real-world context in contrast to those
characterized as not accusatorial or information
gathering, or even those techniques characterized
as both accusatorial and information gathering.
Although these results suggest that such methods
are effective tools for elicitation of confession
evidence, it is important to note that these
findings fail to distinguish the diagnostic value
of the interrogative evidence – field studies offer
little or no information to distinguish between
innocent and guilty suspects, and ground truth
in such contexts is nearly impossible to
determine. As such, researchers have assessed
the diagnostic value of interrogative methods
Interview and Interrogation Methods Effects on Con-
fession Accuracy, Table 2 Mean weighted effect sizes
for field studies
Interrogative
method kN g 95 % CI Q
Accusatorial 3 306 0.090
***
(0.38, 1.41) 4.89
Information
gathering
2 222 0.86
*
(0.04, 1.69) 5.54
*
General
interrogative
methods
3 422 0.19 (.069, 1.06) 25.35
***
*
p<.05;
**
p<.01;
***
p<.001
I2676 Interview and Interrogation Methods Effects on Confession Accuracy
by modeling the interrogative process in an
experimental, laboratory context.
Experimental Studies
Twelve experimental studies were identified
through the systematic search, and these studies
provided 30 contrasts between the different
interrogation methods. All but one of the studies
was conducted in the United States. Nine were
published in peer-reviewed journals and three
are currently unpublished.
Eleven of the 12 studies used variations of the
Kassin and Kiechel (1996) or the Russano et al.
(2005) paradigm. The Kassin and Kiechel
paradigm is one in which all participants are
“innocent” of the mock crime of crashing the
computer. The Russano et al. paradigm includes
participants randomly assigned to an innocent
or guilty condition of a known, intentional act
(i.e., cheating in an academic context). Eleven
of the studies used undergraduate students
as participants, with two studies including
participants from other age groups. The results
are presented in Table 3.
Accusatorial versus Control. The contrast
between an accusatorial interrogative method
and a control interview condition demonstrated
that accusatorial methods yielded a moderate
and significant increase in the frequency of true
confessions and large and significant increase
in false confessions. Typically, the control
technique was the absence of a specific
technique; for example, if the experimental
manipulation was the presentation of false
evidence, the control condition did not present
false evidence. The findings for false confessions
were highly variably across studies.
Information Gathering versus Control.
Two studies examined the influence of
information-gathering interrogative methods
versus a control condition in eliciting true
confessions and false confessions. These
studies demonstrated that information-gathering
methods yielded a greater frequency of true
confessions, but did not significantly influence
the likelihood of eliciting false confessions.
Accusatorial versus Information Gathering.
Three studies assessed the direct contrast
between accusatorial and information-gathering
interrogative methods in eliciting true
confessions and false confessions. The results
demonstrated that information-gathering
methods produced a significantly greater
frequency of true confessions, while significantly
reducing the frequency of false confessions,
when compared with accusatorial methods.
Neither analysis produced significant variability,
suggesting consistency in findings across studies
(Qs<4.43, ns).
This meta-analysis of the eligible experimen-
tal literature demonstrated several key findings
that may have implications for policy and
practice. First, while accusatorial methods signif-
icantly increased the likelihood of obtaining
a true confession (when compared with a no-tactic
control condition), these methods also
significantly increased the likelihood of obtaining
a false confession – a finding consistent with
many cases of wrongful conviction in the
United States (see Kassin et al. 2010;
Lassiter and Meissner 2010). In contrast to this,
information-gathering approaches significantly
increased true confession rates, but showed
no significant increase in the rate of false
Interview and Interrogation Methods Effects on Confession Accuracy, Table 3 Mean weighted effect sizes for
experimental studies
Interrogative contrast Outcome KN g 95 % CI Q
Accusatorial versus control True confession 6 272 0.46* (0.06, 0.86) 7.52
False confession 14 892 0.74*** (0.35, 1.12) 32.99**
Information gathering versus control True confession 2 110 0.67* (0.02, 1.32) 1.41
False confession 2 110 0.23 (0.98, 0.52) 0.11
Accusatorial versus information gathering True confession 3 215 0.64* (0.01, 1.28) 3.62
False confession 3 215 0.77* (1.46, 0.08) 4.43
*p<.05; **p<.01; ***p<.001
Interview and Interrogation Methods Effects on Confession Accuracy 2677 I
I
confessions when compared with a no-tactic
control condition. In fact, information-gathering
approaches appeared to show a numerical
decrease in the rate of false confessions obtained.
When compared directly against accusatorial
methods, information-gathering approaches
showed superior diagnosticity by significantly
increasing the elicitation of true confessions
and significantly reducing the incidence of
false confessions.
Future Directions
A number of variables were considered for
inclusion in a moderator analysis of the influence
of accusatorial methods in eliciting false
confessions. Unfortunately, studies varied little
in several key factors. For example, only two of
the 14 independent samples involved children
or adolescents, while the remainder involved
college students. In addition, none of the studies
manipulated race or ethnicity in participant
recruitment or analyses of the data. Similarly,
only one study was conducted outside of
the United States.
It is important here to note that field studies
fail to offer us important information regarding
the relative diagnostic value of the confession
that is elicited. That is, such studies lack “ground
truth” that would enable us to factually determine
the veracity of the statement provided by
a suspect and thereby preclude the ability to
assess the diagnostic value of the information
elicited and the effectiveness of such techniques
when employed in the field. One method often
used to assess veracity in field studies has been to
evaluate the “strength” of available evidence
against the defendant; however, none of the
studies took this approach to evaluating the likely
credibility of the confession evidence obtained as
a moderator of interrogative efficacy.
Additionally, each of the studies included
in the field study meta-analysis examined
the bivariate relationship between certain
interrogative methods and elicitation of
a confession. As indicated in the review,
a number of control variables could reasonably
be included in such analyses (e.g., factors
related to interrogator experience, crime type,
interrogator/suspect ethnic backgrounds,
geographic characteristics), and more complex
modeling approaches (such as multilevel
modeling or path analysis) could have been
pursued, albeit many (if not all) of these studies
may not have had a large enough sample size to
consider multiple factors simultaneously.
Researchers are strongly encouraged to initiate
more systematic, multilevel analyses of the
influence of interrogative methods. Further,
there is a great need for the use of
quasi-experimental methods in the field context
as the effects of certain interrogative methods are
currently unknown. Quasi-experimental methods
could include the examination of the influence of
certain factors in real-world interviews and inter-
rogations, such as the use of the cognitive inter-
view, whether suspects are told they are being
recorded, and many of the variables under con-
sideration here. Such quasi-experimental
methods are effective tools for assessing the pol-
icy implications of alternative approaches to
police interviewing and interrogation and should
be considered in the years ahead.
Conclusions
The meta-analyses summarized above suggest
that the information-gathering approach
introduced by Great Britain can be effective in
eliciting confession evidence but also
has the advantage of eliciting more
diagnostic information. In the experimental
meta-analysis, when the information-gathering
and accusatorial approaches were contrasted,
the information-gathering approach clearly
produced more advantageous outcomes
(although caution is warranted given the small
number of eligible studies). Specifically,
the information-gathering approach produced
significantly more true confessions, whereas the
accusatorial approach produced significantly
more false confessions. As such, the results
suggest that law enforcement, military, and
intelligence agencies should consider the use
I2678 Interview and Interrogation Methods Effects on Confession Accuracy