Police Reports of Mock Suspect Interrogations: A Test of Accuracy and Perception

Abstract

A 2-phased experiment assessed the accuracy and completeness of police reports on mock interrogations and their effects on people's perceptions. In Phase 1, 16 experienced officers investigated a mock crime scene, interrogated 2 innocent suspects-1 described by the experimenter as more suspicious than the other-and filed an incident report. All 32 sessions were covertly recorded; the recordings were later used to assess the reports. In Phase 2, 96 lay participants were presented with a brief summary of the case and then either read 1 police report, read 1 verbatim interrogation transcript, or listened to an audiotape of a session. Results showed that (a) Police and suspects diverged in their perceptions of the interrogations; (b) Police committed frequent errors of omission in their reports, understating their use of confrontation, maximization, leniency, and false evidence; and (c) Phase 2 participants who read a police report, compared to those who read a verbatim transcript, perceived the process as less pressure-filled and were more likely to misjudge suspects as guilty. These findings are limited by the brevity and low-stakes nature of the task and by the fact that no significant effects were obtained for our suspicion manipulation, suggesting a need for more research. Limitations notwithstanding, this study adds to a growing empirical literature indicating the need for a requirement that all suspect interrogations be electronically recorded. To provide a more objective and accurate account of what transpired, this study also suggests the benefit of producing verbatim transcripts. (PsycINFO Database Record

Full text

Police Reports of Mock Suspect Interrogations: A Test of
Accuracy and Perception
Saul M. Kassin
John Jay College of Criminal Justice
Jeff Kukucka
Towson University
Victoria Z. Lawson
Institute for State and Local Governance of the City University
of New York
John DeCarlo
University of New Haven
A 2-phased experiment assessed the accuracy and completeness of police reports on mock interrogations
and their effects on people’s perceptions. In Phase 1, 16 experienced officers investigated a mock crime
scene, interrogated 2 innocent suspects—1 described by the experimenter as more suspicious than the
other—and filed an incident report. All 32 sessions were covertly recorded; the recordings were later used
to assess the reports. In Phase 2, 96 lay participants were presented with a brief summary of the case and
then either read 1 police report, read 1 verbatim interrogation transcript, or listened to an audiotape of a
session. Results showed that (a) Police and suspects diverged in their perceptions of the interrogations;
(b) Police committed frequent errors of omission in their reports, understating their use of confrontation,
maximization, leniency, and false evidence; and (c) Phase 2 participants who read a police report,
compared to those who read a verbatim transcript, perceived the process as less pressure-filled and were
more likely to misjudge suspects as guilty. These findings are limited by the brevity and low-stakes
nature of the task and by the fact that no significant effects were obtained for our suspicion manipulation,
suggesting a need for more research. Limitations notwithstanding, this study adds to a growing empirical
literature indicating the need for a requirement that all suspect interrogations be electronically recorded.
To provide a more objective and accurate account of what transpired, this study also suggests the benefit
of producing verbatim transcripts.
Keywords: interrogations, suspects, police reports, accuracy
Supplemental materials: http://dx.doi.org/10.1037/lhb0000225.supp
Over the years, research on police interrogations, confes-
sions, and their role in known cases of wrongful conviction has
animated calls for reform (for reviews, see Gudjonsson, 2003;
Kassin, 1997,2005,2012;Kassin & Gudjonsson, 2004;Lassiter
& Meissner, 2010; for an official white paper, see Kassin et al.,
2010). Many such reform efforts have been aimed at protecting
highly vulnerable populations (e.g., juveniles, people with in-
tellectual or mental health impairments) and at curtailing the
use of coercive interrogation practices (e.g., presentations of
false evidence, minimization tactics that imply leniency). Per-
haps the most significant proposed safeguard is to require the
electronic recording of interrogations—the entire process, not
just the confession. As stated in the AP-LS white paper: “With-
out equivocation, our most essential recommendation is to lift
the veil of secrecy from the interrogation process in favor of the
principle of transparency” (Kassin et al., 2010, p. 25).
There is a perennial debate concerning the recording of suspect
interviews and interrogations (for an overview, see Drizin &
Reich, 2004). As an historical matter, the practice has drawn strong
resistance from many federal, state, and local police professionals
(e.g., Boetig, Vinson, & Weidel, 2006)— especially those trained
by John Reid & Associates, which, until recently, had steadfastly
opposed the recording of interrogations (Inbau, Reid, Buckley, &
Jayne, 2001). The bases of opposition have varied. Some have
opposed recording on pragmatic and logistical grounds— citing the
scope of such a requirement; financial costs; the evidentiary con-
sequences of a failure to comply, for example, due to equipment
malfunction; and issues of consent, especially in two-party consent
states. Others have expressed concern over how recording might
alter the behavior of both police and suspects during interroga-
tion and the subsequent decision-making of judges and juries
This article was published Online First December 12, 2016.
Saul M. Kassin, Department of Psychology, John Jay College of Crim-
inal Justice; Jeff Kukucka, Department of Psychology, Towson University;
Victoria Z. Lawson, Institute for State and Local Governance of the City
University of New York; John DeCarlo, Department of Criminal Justice,
University of New Haven.
This work was supported in full by a grant of the National Science
Foundation, Division of Law and Social Sciences, to Saul M. Kassin (SES
1021422). We are also indebted to Kimberly Schantz and Stephanie Spie-
gel for their invaluable assistance in coding of the interrogations.
Correspondence concerning this article should be addressed to Saul M.
Kassin, Department of Psychology, John Jay college of Criminal Justice,
524 West 59th Street, New York, NY 10019. E-mail: skassin@jjay.cuny
.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
2017, Vol. 41, No. 3, 230–243 0147-7307/17/$12.00 http://dx.doi.org/10.1037/lhb0000225
230

(for an overview of these arguments, see Sullivan, 2008;Thur-
low, 2005).
At the same time, a policy of mandatory recording has received
support from various organizations (e.g., American Bar Associa-
tion, 2004;American Psychological Association, 2014;Buckley &
Jayne, 2005;The Justice Project, 2007), as well as from surveys of
police investigators across the United States and Canada (Geller,
1993;Kassin et al., 2007). Sullivan (2004) interviewed police from
hundreds of departments that recorded custodial interrogations and
consistently found that they fully embraced the practice. Respon-
dents cited numerous benefits of recording, namely that it allowed
detectives to focus on the suspect rather than on concurrent note-
taking; that it allowed them to later review the suspect’s answers
to questions for any incriminating comments that had initially gone
unnoticed; that it lessened the need for detectives to defend their
interrogation practices in court; and that it enhanced public trust in
law enforcement (also see Sullivan, Vail, & Anderson, 2008).
Importantly, the U.S. Department of Justice recently reversed its
longstanding opposition to recording by establishing the presump-
tive requirement that federal law enforcement agencies (including
the FBI) videotape the custodial interrogations of felony suspects
(Schmidt, 2014).
That police who record interrogations report high levels of
satisfaction is a vital data point in efforts to reform majority
practice. But what are the actual effects? There are two primary
sets of reasons for the proposed recording requirement. The first is
the expectation that the practice of recording will increase account-
ability and discourage the use of coercive interrogation tactics,
thus reducing the risk to all suspects. To test this hypothesis,
Kassin, Kukucka, Lawson, and DeCarlo (2014) conducted an
experiment in a police station, in which 61 experienced investiga-
tors interrogated a male suspect who was either guilty or innocent
of a mock theft. Before each interrogation, the investigator either
was or was not informed that their session would be surreptitiously
recorded. These recordings were later coded for the use of various
high-pressure tactics designed to elicit a confession. As predicted,
camera-informed interrogators were less likely than their unin-
formed counterparts to use both maximization and minimization
tactics; they were also judged by suspects—who were not told of
the camera’s presence—as trying less hard to obtain a confession.
These findings suggest that recording can affect the process of
interrogation—namely, by inhibiting the use of coercive tactics.
A second purported benefit of recording interrogations is to
provide an accurate factual record of the interrogation behavior of
police and suspects. Perhaps the most frequently invoked argument
is that it is the most effective way to memorialize the process by
which a statement was taken and, hence, increase the fact-finding
accuracy of prosecutors (who decide whether to charge a suspect),
judges (who rule on whether a confession was voluntary or co-
erced), and juries (who determine whether a confession is credible
and hence whether the confessor is guilty or innocent). In current
practice, whereby detectives take contemporaneous or retrospec-
tive notes, disputes often arise as to whether Miranda rights were
administered and waived in a timely manner; whether the suspect
was cooperative or evasive; whether police made or implied prom-
ises or threats, or lied about evidence; and, importantly, whether
the details contained within a confession originated from the suspect.
Disputes over this latter issue can prove devastating. In a descriptive
analysis of 38 false confessions from the Innocence Project, Garrett
(2010) found that 36 were “contaminated,” containing accurate crime
details allegedly known only to the perpetrator— details not in the
public domain, but known to police, that the innocent confessor could
not have produced without exposure to secondhand information. The
result: An increased likelihood of conviction, as the presence of
details in a confession enhances perceptions of its credibility
(Appleby, Hasel, & Kassin, 2013).
In lieu of electronic recordings, the Federal Rules of Evidence
(2015) provide that a police witness may use a personally prepared
report concerning an interview or interrogation, including what the
defendant said, to refresh his or her recollection while testifying
(FRE 612). Yet the accuracy of these reports, which is often in
dispute, has never been tested in the context of a suspect interro-
gation. The question we posed in the present research is: What
does recording reveal about the memorial accuracy of police
reports of interrogations, and with what effect on prospective fact
finders?
Basic research on memory for conversation content suggests
that this process may be fraught with bias and error. Neisser (1981)
highlighted this problem in a case study in which he analyzed John
Dean’s high-stakes memory of Watergate-related conversations
with President Nixon. Dean testified as to his recollections with
specificity and confidence. Yet when tapes of oval office conver-
sations were discovered, Neisser discovered that although Dean
was generally correct about what happened (e.g., that there was a
cover-up), his memory of specifics was often distorted (e.g., over-
estimating his own role). This case stands in contrast to research
showing that active involvement in a conversation tends to en-
hance memory for content; for example, people recall what they
said better than what they read or heard (MacLeod, Gopie, Hou-
rihan, Neary, & Ozubko, 2010).
Apart from the potential for intrusion and omission errors in
memory for conversations, research shows that memories of con-
tent and context are stored independently, thereby increasing the
risk of source monitoring confusion (Johnson, Hashtroudi, & Lind-
say, 1993). Moreover, such errors often reflect the operation of
cognitive confirmation biases (Kleider, Pezdek, Goldinger, &
Kirk, 2008;Schacter, 2001). In a study that demonstrated the
problem in a forensic context, Lamb, Orbach, Sternberg, Hersh-
kowitz, and Horowitz (2000) took a sample of 20 interviews of
alleged child sex abuse victims and compared interviewers’ con-
temporaneous notes against audiotapes of these sessions. The
interviewers’ notes proved inadequate, as they failed to mention
57% of their own utterances and 25% of details that the children
provided. Moreover, the notes often contained serious source
attribution errors, such as citing the children rather than their own
prompting questions as the source of details that were disclosed.
Lamb et al. (2000) concluded: “Even when they made contempo-
raneous verbatim notes, these investigators tended to understate
their role in eliciting the information” (p. 705; see also Bruck,
Ceci, & Francoeur, 1999).
The consequences of source attribution errors were realized by
former D.C. Detective James Trainum (2007) who—in an article
entitled “I took a false confession so don’t tell me it doesn’t
happen!”— described a case in which a former suspect who had
confessed was later exonerated:
Years later, during a review of the videotapes, we discovered our
mistake. We had fallen into a classic trap. We believed so much in our
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.
231
POLICE REPORTS OF MOCK INTERROGATIONS

suspect’s guilt that we ignored all evidence to the contrary. To
demonstrate the strength of our case, we showed the suspect our
evidence, and unintentionally fed her details that she was able to
parrot back to us at a later time. It was a classic false confession case
and without the video we would never have known.
To further complicate matters of recollection, there is reason to
believe that suspects likewise cannot be trusted to provide accurate
accounts. Unlike most social interactions, police interrogations are
exceptionally stressful events for the accused, and this level of
stress can produce deleterious effects on memory retrieval (Def-
fenbacher, Bornstein, Penrod, & McGorty, 2004). In a study that
illustrates this point, Morgan et al. (2004) randomly assigned
trainees undergoing military survival training to endure a realistic
high-stress or low-stress mock interrogation. One day later, many
of those in the high-stress condition could not even identify their
interrogator from a lineup.
In light of the research literature on memory for conversations,
the acceptance in court of note taking in lieu of electronic record-
ing, and the significance of the issue for policy and practice, the
present research was designed to assess the memorial accuracy of
police reports of suspect interrogations. In an experiment con-
ducted at a large Northeastern police station, a sample of experi-
enced investigators examined a mock crime scene, interrogated
two innocent male suspects, and submitted a report on their inter-
rogations. For each investigator, expectations were varied from
one session to the next, with one of the two suspects presented via
demeanor cues as having behaved suspiciously. Unbeknownst to
these investigators, all sessions were covertly audio recorded. By
later comparing police reports of the sessions with the actual tapes,
we addressed two questions. First, what does audio recording
reveal about substantive accuracy of police accounts, as measured
by intrusion errors, omission errors, and source monitoring con-
fusion, and what does it reveal about the extent to which suspicion
elicits confirmation bias in police reports of interrogations? Sec-
ond, with regard to the fact finder, how might access to the
unabridged verbatim transcripts influence others’ perceptions of
the interrogations and suspects compared to reading only the
police reports?
Phase 1 Method
Participants
Phase 1 participants were 18 investigators of various ranks from
a large Northeastern police department and 36 male community
members recruited via Craigslist to serve as mock suspects. Police
participants were recruited during daily roll calls and later via
word-of-mouth once the study was ongoing. As noted in the
consent form, the stated purpose of the research was “to study the
way police investigators conduct suspect interviews and interro-
gations, and write informative incident reports on their work.”
Data from one police participant were excluded after he expressed
suspicion that his session had been recorded; a second police
participant failed to submit the required written report following
his session. Hence, the final sample consisted of N⫽16 police
participants (eight officers, two detectives, and six sergeants) who
conducted and reported on a total of 32 suspect interrogations. All
sessions were conducted in a vacant office at the police station.
Police participants, half of whom were female, ranged in age
from 26 to 55 (M⫽43.94, SD ⫽7.46) and had an average of
16.33 years of law enforcement experience (SD ⫽6.00). Half had
received formal training in suspect interviewing and interrogation;
62.50% estimated that they had conducted over 100 suspect inter-
views during their careers (if a range was provided, we converted
the estimate to the midpoint of that range; the overall median was
250). All suspects were male and ranged in age from 18 to 62
(M⫽34.91, SD ⫽13.86). A total of 46.88% had previously been
arrested; 25% had been suspect-interviewed by police; 28.13% had
been convicted of a crime.
Design
Within each session, one police participant investigated a staged
crime scene involving a theft and then interrogated two male
suspects, both of whom were factually innocent. After looking at
the crime scene but prior to interrogating the first suspect, they
were told that the first or second of the two suspects (in counter-
balanced order) had acted suspiciously upon learning that he
would be questioned about the theft. Phase 1 thus employed a
two-group (Suspicion: Present vs. Absent) within-subjects design.
Procedure
Two experimenters were involved in each session— one to
meet and instruct the police participant, and the other to sepa-
rately meet and instruct the two suspects. Each session began
with a preexperiment questionnaire, at which time a theft was
said to have occurred. This was followed by a police crime
scene investigation; two sequential suspect interrogations; and
postinterrogation questionnaires. After each session, the police
participant was asked to produce and submit a written report
within 48 hours that detailed his or her crime scene investiga-
tion and interrogations.
Preexperiment questionnaires and mock theft. Upon ar-
rival, the police participant was escorted to an interview room; the
two suspects were escorted to a separate waiting room. After
giving informed consent, the police participant indicated his or her
age and gender and answered several questions concerning back-
ground and training.
After suspect participants gave their informed consent, an
experimenter directed but did not accompany them, one-at-a-
time, to another room where they were to complete a preex-
periment questionnaire. The questionnaire asked suspects to
indicate their age and whether they had ever been interviewed
by police, arrested, or convicted of a crime. On their way to and
from this room, suspects passed an unattended briefcase in the
hallway that had been staged to look as though a theft had
occurred. The briefcase zipper was left ajar, and on the floor
next to it were an open lock (which had ostensibly been re-
moved from the briefcase) and a wallet with no cash inside.
Suspects were instructed to take notice of— but not to touch—
the briefcase. All suspects, therefore, had exposure to the scene
of the mock theft but were factually innocent.
Once both suspects had completed the questionnaire and
returned to the waiting room, the experimenter informed them
that they were the targets of an investigation into a theft
committed at the police station and that they would soon be
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.
232 KASSIN, KUKUCKA, LAWSON, AND DECARLO

interviewed by a detective. Suspects were paid $20 for their
participation prior to being interrogated and told that they
would receive a $15 bonus if they convinced the detective of
their innocence. If they did not succeed, they were told that they
would have to return for a second session to receive the addi-
tional payment (in actuality, all suspects were paid the full $35
after the session).
Crime scene investigation. At that point, the second experi-
menter informed the police participant that an unknown sum of
cash had been stolen from a briefcase in the hallway and that the
crime scene had been left exactly as it was found. In the context of
a lengthy instruction, police were asked to investigate the crime
scene, question two suspects who were apprehended nearby
around the time that the theft occurred, and solve the crime. The
experimenter explained that both suspects had walked past the
briefcase to complete paperwork in a nearby room and were out of
view when they did so. We provided no inculpatory evidence
against the suspects aside from the fact that the money was missing
and both were known to have been alone in the vicinity at the time.
Police were assured that no actual crime had taken place and that
interrogation must be terminated if a participant suspect wanted to
stop (the verbatim instructions are available as online supplemen-
tal material).
The police participant was then taken to the crime scene, pro-
vided with a notepad and digital camera, and given 5 minutes to
investigate the area. Afterward, the experimenter escorted the
police participant back to the interrogation room and inquired as to
whether he or she wanted hard copies of their crime scene photo-
graphs for use during the interrogation. If they did, these images
were printed, in color, on 4 ⫻6 in. photo paper.
Suspicion manipulation. While escorting the police partici-
pant back to the interrogation room, the experimenter reiterated
that he or she would question two theft suspects. In counterbal-
anced order, the experimenter added that:
The first [or second] guy you’re going to question seemed okay, but
the second [or first] guy was acting really strange. He looked nervous:
He kept pacing back and forth and wouldn’t make eye contact with me
while I was talking to him.
Given that a mock crime was said to have been committed, this
comment about demeanor was designed to selectively raise each
police participant’s relative a priori suspicion toward one of the
two suspects.
Interrogations and postinterrogation questionnaires. Police
were told that they would have 20 to 30 min to interrogate each
suspect, after which they were to return the suspect to the waiting
room. If an interrogation was still ongoing at 20 min, the experi-
menter knocked on the door to signal to the police participant that
it was time to wrap up the interrogation. Suspects were randomly
assigned to be interrogated either first or second. All interrogations
were surreptitiously audio recorded through a digital voice re-
corder hidden among office supplies on a desk. Neither police
participants nor suspects were preinformed of the fact that the
interrogations would be recorded.
After each interrogation, the suspect returned to the waiting
room, and both the police participant and suspect completed a
self-report questionnaire concerning their perceptions of the inter-
rogation experience. Police participants completed this question-
naire twice, once after each of their interrogations. After the
suspect completed his questionnaire, he was fully debriefed, paid,
and dismissed.
Police reports. After completing the second postinterrogation
questionnaire, police were instructed on how to prepare and submit
Incident Reports of their investigation, which were due within 48
hours. To ensure that investigators took the task seriously, these
instructions stipulated that reports should be approximately three
to five typed pages in length, single spaced, and should consist of
three sections: a summary of their crime scene analysis (one page),
an account of the first suspect interrogation (1–2 pages), and an
account of the second suspect interrogation (1–2 pages). The
instructions also specified the sorts of details that should be in-
cluded—including key questions asked during each interrogation,
suspects’ answers to these questions, descriptions of each suspect’s
demeanor, and any indications or impressions of each suspect’s
involvement. Upon receipt of these reports, police were debriefed
and paid $100 for their participation.
Dependent Measures
Postinterrogation questionnaires. Immediately after each
interrogation, police and suspect participants completed a self-
report questionnaire consisting of 12 parallel items that measured
their perceptions of the interrogation. First, police indicated
whether they believed the suspect to be guilty or innocent and
indicated their confidence in this impression on a scale ranging
from 1 (not at all)to10(very). These two items were later
combined to form a guilt-confidence composite score that could
range from ⫺10 (highly confident guilty judgment)to⫹10 (highly
confident innocent judgment). Police also gave continuous ratings
of how credible the suspect’s denials were, how knowledgeable the
suspect seemed to be about the crime, how cooperative the suspect
was, how hard they tried to get the suspect to confess, how friendly
they were toward the suspect, how confrontational they were
toward the suspect, how anxious the suspect was, and how stress-
ful they thought the interrogation was for the suspect. All ratings
were made on a scale that ranged from 1 (not at all)to10(very).
Finally, police gave two dichotomous yes/no judgments as to
whether the suspect had made any suspicious remarks and/or any
admissions of guilt during the interrogation.
Suspects simultaneously answered parallel questions. First, they
indicated whether they believed the interrogator would perceive
them to be guilty or innocent and rated their confidence in that
judgment on a 10-point scale. Once again, these two items were
combined to form a composite score that could range from ⫺10
(highly confident that they would be judged guilty)to⫹10 (highly
confident that they would be judged innocent). Suspects also rated
how credible their denials were, how much knowledge they had
about the crime, how cooperative they were, how hard the inter-
rogator tried to get them to confess, how friendly the interrogator
was, how confrontational the interrogator was, how anxious they
were, and how stressful the interrogation was for them. These
continuous ratings were made on a scale ranging from 1 (not at all)
to 10 (very). Suspects also gave dichotomous yes/no judgments as
to whether they had made any suspicious remarks and/or any
admissions of guilt. Two suspects (6.25%) neglected to answer one
or more of these items, and thus their data are missing from the
relevant analyses.
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.
233
POLICE REPORTS OF MOCK INTERROGATIONS

Interrogation tactics. Audio recordings of all 32 interroga-
tions were transcribed by a professional transcription service. The
transcripts were then coded by two independent coders who were
blind to our suspicion manipulation and had prior experience
coding interrogation transcripts. On the basis of a previously
published factor analysis of self-reported tactic use in a survey of
police (Kassin et al., 2007), which formed the basis of a coding
scheme introduced by Kassin et al. (2014), we coded transcripts
for the presence of 16 interrogation tactics organized into five
categories (see Table 1).
For three of the five categories (confrontation, maximization,
and leniency), coders counted the number of times each individual
tactic was used. For some analyses, these counts were later coded
as absent (i.e., a count of zero) or present (i.e., any nonzero count).
We also computed the sum of tactic uses within each category,
enabling us to create a dichotomous measure of whether or not
police used at least one tactic from a given category. For the other
two categories (false evidence and miscellaneous), coders merely
noted whether each tactic was present or absent.
Confrontation tactics involved calling the suspect a liar (e.g.,
“You just lied to me,” “You’re pretty much lying to an authority
figure you know”), pointing out inconsistencies in the suspect’s
story (e.g., “You’re flipping your story,” “How do you know there
was a card in the wallet? You said you just looked at it”), outright
accusing the suspect of committing the crime (e.g., “You did it,”
“Okay, well I think that you took the money”), expressing disbelief
in the suspect’s story (e.g., “That’s hard to believe,” “I can tell you
right now just from the evidence that I have that your story is not
believable”), and interrupting the suspect’s attempts to maintain
his innocence (as denoted by midsentence hyphens in the suspect
portions of the transcripts).
Maximization tactics include threatening the suspect with neg-
ative consequences for not confessing (e.g., “If it comes back to
you, you’re done,” “If you don’t tell me up front and then I have
to determine that by evidentiary matter then I slam you in court,
okay?”), and exaggerating the seriousness of the offense (e.g.,
“The amount of cash in the wallet can bring the charges from now
a theft to a much higher degree,” “There’s going to be an addi-
tional charge for hindering prosecution”).
Leniency tactics include developing minimization themes that
excuse, justify, or otherwise downplay the crime, statements that
may lead people to infer leniency in punishment (e.g., “This is
something, it’s so minor,” “It doesn’t make you a bad person, bro”)
and explicit offers of leniency or immunity in exchange for a
confession (e.g., “I got a little bit of juice over at the courthouse
and if you did do it, if you’re honest and up front with me, I can
work with you,” “I’ll get you probation on this,” “I am offering
you a deal here then, you know”).
False evidence tactics include the bluff (the assertion that there
is evidence to be harvested without the added claim that it impli-
cates the suspect; e.g., “Understand I am going to check the
cameras,” “That little padlock and key that you saw, that has been
submitted for DNA analysis”) and specific false claims about the
existence of actual evidence (e.g., “I already told you there is one
person that ID’d you right?” “We have video of you going in that
wallet”).
In addition to these basic Reid-technique approaches, a range of
miscellaneous tactics were coded as well—including irrelevant
small talk aimed at establishing a rapport (e.g., “What kind of food
you like cooking?” “What kind of work do you do?”), asking the
suspect to implicate someone else (e.g., “Did you see any of the
other guys like [John], his brother, whoever, anybody with money
take money out of their pockets, anything?” “What about the, um,
the older gentleman, was he referencing anything about money?”),
encouraging the suspect to admit to other illegal behaviors (e.g.,
“Are you on probation or anything?” “Okay have you ever been
Table 1
Presence and Frequency of 16 Categorized Interrogation Tactics Across 32 Interrogations
Interrogation tactics
% Sessions
used
M(SD)of
times used
Confrontation 56.25 1.63 (1.98)
Calling the suspect a “liar” or accusing him of “lying” 25.00 .63 (1.43)
Pointing out inconsistencies in the suspect’s story 25.00 .53 (1.16)
Directly accusing the suspect of the theft 15.63 .25 (.76)
Making expressions of disbelief toward the suspect 12.50 .19 (.47)
Interrupting the suspect’s denials 3.13 .03 (.18)
Maximization 59.38 2.09 (2.61)
Threatening the suspect with consequences 56.25 1.94 (2.48)
Exaggerating the seriousness of the offense 15.63 .16 (.37)
Leniency 62.50 2.66 (3.82)
Stating minimization themes that imply leniency 53.13 1.44 (2.47)
Making an explicit offer of leniency for confession 46.88 1.22 (1.72)
False evidence 84.38 1.56 (1.02)
ⴱ
Bluffing about future evidence 84.38 1.47 (.98)
Lying about existing evidence 25.00 .38 (.75)
Miscellaneous tactics
Establishing rapport via small talk 75.00 N.A.
Pressing the suspect to implicate someone else 75.00 N.A.
Encouraging the suspect to admit other illegal acts 68.75 N.A.
Appealing to the suspect’s religion/conscience 21.88 N.A.
Praising or flattering the suspect 15.63 N.A.
ⴱ
These values represent the number of different types of evidence about which police participants bluffed or
lied.
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.
234 KASSIN, KUKUCKA, LAWSON, AND DECARLO

arrested?”), appealing to the suspect’s religion or conscience (e.g.,
“Some of the items had some sentimental value that’s what it is. I
just want to get those returned to the owner,” “Are you religious at
all? There is definitely right and wrong in the universe”), and the
use of flattery (e.g., “You seem like a pretty good dude,” “You’ve
obviously got a great relationship with your son, great relationship
with your mom”).
Next we coded for the presence or absence of five discernible
police behaviors (gathering personal information from suspects,
showing crime scene photos to the suspect, asking the suspect to
empty his pockets, issuing a Miranda warning, and obtaining a
Miranda waiver; see Table 2). We also coded for the frequency of
four discernible suspect behaviors (denials of guilt, denials of
crime-relevant knowledge, disclosures of crime-relevant knowl-
edge, and self-incriminating admissions; see Table 2).
For tactics and behaviors that were coded merely as present or
absent (i.e., false evidence tactics, miscellaneous tactics, and po-
lice behaviors), coders exhibited an overall agreement rate of
91.67%, ␬⫽0.83 [95% CI: 0.77, 0.89], p⬍.001 (within each
category, all ␬s⬎.69, ps⬍.001). For those that were coded as
frequency counts (i.e., confrontation, maximization, and leniency
tactics, and suspect behaviors), the overall intraclass correlation
(ICC) for our two raters was near perfect, ICC ⫽.97 [95% CI: .96,
.98], p⬍.001 (within each category, all ICCs ⬎.94, ps⬍.001).
Whatever disagreements or discrepancies that remained between
coders were resolved via discussion.
Police reports. Two additional independent coders read and
coded police reports for disclosures of these same tactics and
behaviors, allowing us to compare the contents of the interrogation
transcripts against the corresponding police accounts of those same
interrogations. For reports, all tactics and behaviors were coded as
either present (i.e., the police participant described the use of a
given tactic or the presence of a given behavior) or absent (i.e., the
given tactic or behavior was not noted in the police report). These
coders exhibited an overall agreement rate of 97.10%, ␬⫽0.89
[95% CI: 0.85, 0.93], p⬍.001 (within each category, all ␬s⬎.65,
ps⬍.001). Once again, disagreements were resolved by discus-
sion.
Phase 1 Results
Length of Interrogations
On average, interrogations lasted for 16.41 min (SD ⫽5.50;
Range: 3.32 – 26.17) and contained 2,438.31 words (SD ⫽984.11;
Range: 866 – 4,481). Our suspicion manipulation had no effect on
the duration, t(15) ⫽0.29, p⫽.780, d⫽0.07 [95% CI: ⫺0.49,
0.63], or word count, t(15) ⫽0.51, p⫽.619, d⫽0.13 [95%
CI: ⫺0.43, 0.68], of interrogations (note that all CIs reported here
and thereafter pertain to effect sizes). On average, interrogations of
the second suspect were longer in word count, t(15) ⫽2.22, p⫽
.042, d⫽0.55 [95% CI: 0.00, 1.11], but not in duration, t(15) ⫽
1.64, p⫽.121, d⫽0.41 [95% CI: ⫺0.15, 0.97]. On average,
56.60% of all words were spoken by the interrogator (SD ⫽
11.63%; Range: 28.05–74.66%), and neither Suspicion, t(15) ⫽
0.56, p⫽.586, d⫽0.14 [95% CI: ⫺0.42, 0.70], nor Order,
t(15) ⫽0.87, p⫽.401, d⫽0.22 [95% CI: ⫺0.34, 0.77] affected
this percentage.
No significant order effects were found on any other dependent
measures, including responses to the postinterrogation question-
naires and the presence and frequency of all coded tactics and
behaviors, with one exception: Suspects who were interrogated
second offered more denials of crime-relevant knowledge (M⫽
3.81, SD ⫽4.04) than those who were interrogated first (M⫽
1.50, SD ⫽1.86), t(15) ⫽2.21, p⫽.043, d⫽0.55 [95%
CI: ⫺0.01, 1.11]. Consequently, all data were collapsed across
order for all subsequent analyses.
Postinterrogation Questionnaire
A 2 (Suspicion: Present vs. Absent) ⫻2 (Source: Police vs.
Suspect) repeated-measures MANOVA was performed on the nine
continuous items on the postinterrogation questionnaire (including
guilt-confidence composite scores) to test for discrepancies be-
tween suspect and police perceptions of the interrogation. A mul-
tivariate effect of Source emerged, Wilks’ ⌳⫽.10, F(9, 5) ⫽5.16,
p⫽.043, with significant differences on four items. Neither the
multivariate effect of Suspicion, Wilks’ ⌳⫽.18, F(9, 5) ⫽2.46,
p⫽.167, nor the Source ⫻Suspicion interaction, Wilks’ ⌳⫽.23,
F(9, 5) ⫽1.86, p⫽.257, was significant.
Follow-up univariate ANOVAs were performed on the four
items that differed as a function of Source (see Figure 1). These
analyses indicated that suspects were more confident that they
would be judged as innocent (M⫽4.07, SD ⫽7.04) than police
were confident in their innocence (M⫽0.54, SD ⫽6.95), F(1,
13) ⫽4.91, p⫽.045, d⫽0.35 [95% CI: ⫺0.07, 0.77]; suspects
rated their denials as more credible (M⫽8.25, SD ⫽1.80) than
police rated their denials (M⫽5.61, SD ⫽2.17), F(1, 13) ⫽
26.92, p⬍.001, d⫽0.88 [95% CI: 0.46, 1.30]; and suspects
rated themselves as more cooperative (M⫽9.11, SD ⫽1.37)
than police rated them (M⫽6.89, SD ⫽2.48), F(1, 13) ⫽
24.21, p⬍.001, d⫽0.84 [95% CI: 0.42, 1.26]. In addition,
police rated the interrogations as more stressful for suspects
(M⫽5.07, SD ⫽2.28) than suspects rated it for themselves
(M⫽3.46, SD ⫽2.15), F(1, 13) ⫽7.57, p⫽.016, d⫽0.63
[95% CI: 0.21, 1.05]. The remaining four items did not differ by
Source. Police and suspects, respectively, rated police as
friendly (Ms⫽6.93 & 6.86, SDs⫽1.88 and 2.19) and mod-
Table 2
Presence and Frequency of Five Police and Four Suspect
Behaviors Across 32 Interrogations
Behaviors coded
% Sessions in
which behavior
occurred
M(SD)of
occurrences
Police behaviors
Obtaining suspect’s personal information 90.63 N.A.
Showing crime scene photos to suspect 37.50 N.A.
Asking suspect to empty his pockets 18.75 N.A.
Issuing a Miranda warning 6.25 N.A.
Obtaining a Miranda waiver 6.25 N.A.
Suspect behaviors
Disclosing crime-relevant knowledge 96.88 16.00 (9.60)
Denying guilt 90.63 7.09 (5.28)
Denying crime-relevant knowledge 75.00 2.66 (3.31)
Making an incriminating admission 12.50 .31 (1.12)
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.
235
POLICE REPORTS OF MOCK INTERROGATIONS

erately confrontational (Ms⫽4.86 and 4.32, SDs⫽2.26 and
2.60), and suspects as moderately anxious (Ms⫽5.18 and 4.32,
SDs⫽2.28 and 2.36) and knowledgeable about the crime
(Ms⫽4.82 and 5.61, SDs⫽2.79 and 2.86).
We separately analyzed three dichotomous items on the postin-
terrogation questionnaire. First, police judged 13 of the 32 suspects
as guilty (40.62%). As with other dependent measures, suspicion
did not affect these judgments, McNemar ␹
2
(1) ⫽0.82, p⫽.549,
nor did it affect suspects’ predictions of whether they would be
judged guilty, McNemar ␹
2
(1) ⫽0.11, p⫽1.00. Police judgments
of the suspect’s guilt were unrelated to suspects’ predictions of
whether they would be judged as guilty, McNemar ␹
2
(1) ⫽0.60,
p⫽.607. Second, police indicated that 13 of the 32 suspects
(40.62%) made suspicious remarks during the interrogation. Once
again, suspicion did not affect police responses to this item,
McNemar ␹
2
(1) ⫽0.14, p⫽1.00 (note that three “suspicious
remarks” suspects were judged innocent; three others not seen as
having made suspicious remarks were judged guilty.). Police and
suspect responses to this item were also unrelated to each other,
McNemar ␹
2
(1) ⫽0.25, p⫽.804. Third, only one suspect was
judged by police as having made an admission of guilt; no suspects
reported having admitted guilt.
Interrogation Tactics and Behaviors
Frequencies and descriptive statistics for all 16 coded interro-
gation tactics and five tactic categories are shown in Table 1. Out
of 32 interrogations, 56.25% featured the use of one or more
confrontation tactics, 59.38% featured one or more maximization
tactics, 62.50% featured one or more leniency tactics, and 84.38%
featured one or more false evidence tactics. The average interro-
gation featured 1.63 uses of confrontation (SD ⫽1.98), 2.09 uses
of maximization (SD ⫽2.61), and 2.66 uses of leniency (SD ⫽
3.82).
With respect to presentations of false evidence, 84.38% of interro-
gations featured bluffs about evidence, and the average interrogation
included bluffs about 1.47 different types of evidence (SD ⫽0.98).
Police most often bluffed about having surveillance footage (75% of
interrogations), followed by fingerprints (46.88%), DNA (18.75%),
and eyewitnesses (3.13%). One quarter of interrogations featured
outright lies about existing evidence, and the average interrogation
included lies about 0.38 different types of evidence (SD ⫽0.75).
Police most often lied about incriminating surveillance footage
(21.88% of interrogations), followed by DNA (9.38%), fingerprints
(3.13%), and eyewitnesses (3.13%).
Descriptive statistics for the five coded police behaviors and
four suspect behaviors are shown in Table 2. Most police solicited
personal information from the suspect (90.63%); fewer showed
crime scene photos (37.50%), asked the suspect to empty his
pockets (18.75%), or read Miranda warnings (6.25%). Virtually all
suspects (96.88%) disclosed some crime-relevant knowledge dur-
ing the interrogation. On at least one occasion, most suspects
denied guilt (90.63%) and denied crime-relevant knowledge
(75.00%); relatively few suspects made incriminating statements
(12.50%).
Suspicion Manipulation
Suspicion did not affect any of the coded tactics, namely,
whether the interrogator exhibited one or more uses of confronta-
tion, maximization, leniency, or false evidence, all McNemar
␹
2
sⱕ1.00, ps⬎.60; the number of uses of confrontation,
maximization, or leniency tactics, all ts⬍1, ps⬎.75, ds⬍0.10;
or the number of different bluffs, t(15) ⫽0.59, p⫽.566, d⫽0.15
[95% CI: ⫺0.41, 0.70], or lies, t(15) ⫽0.19, p⫽.849, d⫽0.05
[95% CI: ⫺0.51, 0.61]. Suspicion likewise did not influence the
likelihood of any of the five coded police behaviors, all McNemar
␹
2
ⱕ2.00, ps⬎.50, the likelihood of any of the four coded
suspect behaviors, all McNemar ␹
2
ⱕ0.50, ps⬎.70, or the
number of times that any of the four suspect behaviors occurred,
all tsⱕ1.05, ps ⬎.30, dsⱕ0.15.
Comparison of Transcripts and Reports
In comparison to the interrogation transcripts, which con-
tained an average of 2,438.31 words (SD ⫽984.11; Range:
866 – 4,481), police reports contained an average of 1,224.44
words (SD ⫽365.52; Range: 570 –1,875). Descriptions of the
crime scene contained fewer words (M⫽258.06, SD ⫽76.31)
than descriptions of either the first (M⫽452.44, SD ⫽174.94)
or the second (M⫽478.31, SD ⫽183.06) interrogation, F(2,
30) ⫽23.01, p⬍.001, ␩p
2⫽.61, which did not differ in word
count.
We sought to measure both the accuracy and completeness of
police reports by noting errors of commission and omission in
relation to the coded transcripts. For this purpose, we compared the
Figure 1. Effects of source (suspect vs. police) on four (of nine) continuous self-report items (Phase 1). Note.
Higher scores on the composite measure indicate greater confidence in the suspect’s projected innocence.
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.
236 KASSIN, KUKUCKA, LAWSON, AND DECARLO

presence of all coded tactics and behaviors in the interrogation
transcripts against the presence of those same tactics and behaviors
in the corresponding police reports. Errors of commission were
extremely rare. Across 25 coded tactics and behaviors in 32
interrogations (a total of 800 coded details), only four errors of
commission were found (0.50%). One interrogator mistakenly
reported having lied about evidence to one suspect and having shown
crime scene photos to the other; a second interrogator mistakenly
reported using minimization tactics; and a third mistakenly reported
having offered leniency in exchange for a confession.
In contrast, errors of omission were prevalent. Figure 2 shows
the frequencies of usage and reporting for all coded tactics and
behaviors. As noted earlier, a majority of interrogations included
the use of one or more tactics involving confrontation (56.25%),
maximization (59.38%), leniency (62.50%), and false evidence
(84.38%). When used, however, the corresponding reports de-
scribed one or more of these tactics only 22.22%, 15.79%, 40%,
and 66.67% of the time, respectively (e.g., in 77.78% of reports of
interrogations in which confrontation was used one or more times,
no mention of any confrontation tactic was made). In short, a
number of different tactics were frequently used during interroga-
tions but did not appear in corresponding police reports.
With respect to suspect behaviors, virtually all suspects
(96.88%) disclosed some crime-relevant knowledge regarding the
mock theft, and these disclosures were almost always reported by
police (96.77%). Most suspects also issued denials of guilt
(90.63%), which were also typically reported (89.66%). However,
while 75% of suspects denied crime-relevant knowledge at least
once, only 33.33% of the corresponding reports noted such denials
(in contrast, three out of four suspects who made incriminating
admissions were reported to have done so).
To sum up: Two sets of findings emerged from Phase 1 of our
study. First, police and suspects diverged in their perceptions of
the interrogations, with suspects seeing themselves as more cred-
ible in their denials, more cooperative, and more confident in their
presentation of innocence. Second, a comparison of interrogation
transcripts and police reports indicated that errors of omission
were prevalent, with police underreporting the use of various
tactics. In light of these results, Phase 2 assessed whether observ-
ers’ impressions of the interrogations were differently influenced
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Making an incriminang admission
Denying crime-relevant knowledge
Denying guilt
Revealing crime-relevant knowledge
SUSPECT BEHAVIORS:
Obtaining a Miranda waiver
Issuing a Miranda warning
Asking suspect to empty his pockets
Showing crime scene photos t o suspect
Obtaining suspect's personal informaon
INTERROGATOR BEHAVIORS:
Praising or flaering the suspect
Appealing to the suspect's religion / conscience
Encouraging suspect to admit to other illegal ac ts
Asking the suspect to implicate someone else
Establishing rapport via small talk
MISCELLANEOUS TACTICS:
Lying about exisng evidence
Bluffing about future evidence
FALSE EVIDENCE:
Making an explicit offer of leniency for confession
Offering minimizaon themes that imply leniency
LENIENCY:
Exaggerang the seriousness of the offense
Threatening the suspect with consequences
MAXIMIZATION:
Interrupng the suspect's denials
Making expressions of disbel ief toward the suspect
Directly accusing the suspect of the the
Poinng out inconsistencies in the suspect's story
Calling the suspect a "liar" or accusing him of "lying"
CONFRONTATION:
Used and reported Used, but not reported Not used
Figure 2. Occurrence and reporting of interrogation tactics, police behaviors, and suspect behaviors across 32
interrogations and corresponding police reports.
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.
237
POLICE REPORTS OF MOCK INTERROGATIONS

by the police reports that were generated relative to full and
objective records.
Phase 2 Method
Participants and Design
Participants in Phase 2 were 96 undergraduates, 64 of whom
were female, with a mean age of 19.45 (SD ⫽2.84). A total of
38.54% self-identified as Hispanic, 17.71% as White, 15.63% as
Asian, 12.50% as Black, and 15.63% as multiracial or Other.
To determine whether outside fact finders would form different
or more accurate impressions when they had access to verbatim
transcripts of suspect interrogations, compared to secondhand po-
lice reports, participants read a brief summary of the case and were
randomly assigned to either read a police report or an interrogation
transcript of one of the 32 Phase 1 interrogations. These conditions
enabled us to compare directly the two alternative written accounts
of the process. Next we sought to assess what impressions partic-
ipants would form if they also listened to audiotapes of these same
sessions, thereby having access to vocal and paralinguistic cues
emanating from both police and suspect participants (e.g., prosod-
ics such as pitch, volume, stress, tempo, pitch, rhythm, pauses,
points of stress, and intonation—the kinds of cues that often lead
laypeople to draw erroneous inferences of truth and deception; see
DePaulo et al., 2003;Bond & DePaulo, 2006). Thus, a third group
of participants— equipped with headphones and an MP3 player—
listened to the audio recording while following along on the
transcript. This group enabled us to compare the two verbatim
conditions— one written, the second accompanied by audio cues.
Because no significant effects of manipulated suspicion were
obtained in Phase 1, this second experiment combined rather than
distinguished these two groups of suspects in all analyses.
Procedure
Participants completed the study in sessions that included from
four to eight individuals. All participants in a given session were
in the same experimental condition, but they read and/or listened
via headphones to different Phase 1 interrogations. After giving
their informed consent and providing basic demographic informa-
tion (i.e., age, gender, and race), all participants read the same
one-page background summary of the crime and investigation that
occurred during Phase 1. The summary described the theft of an
unknown amount of cash from a briefcase inside a police station,
described the crime scene and the contents of the briefcase, and
explained that two suspects were identified who were known to
have been in the vicinity of the briefcase around the time of the
theft. Participants were told that they would now be given an
account of the interrogation of one of these two suspects. Partic-
ipants then read the police report (report condition) or the full
transcript (transcript condition), or they listened to an audio re-
cording (audio condition) of one of the 32 Phase 1 interrogations.
When they were finished, they completed a questionnaire that
measured their perceptions of the interrogator and suspect, after
which they were fully debriefed.
Dependent Measures
The questionnaire administered to Phase 2 participants con-
tained 13 items that paralleled those answered by police and
suspects in Phase 1. First, participants indicated whether they
believed the suspect to be guilty or innocent and rated their
confidence in that judgment on a scale from 1 (not at all)to10
(very). As in Phase 1, these items were combined to form a
guilt-confidence composite score that could range from ⫺10
(highly confident guilty judgment)to⫹10 (highly confident inno-
cent judgment). Participants also gave continuous ratings of how
credible the suspects’ denials were, how knowledgeable the sus-
pect was about the crime, how cooperative the suspect was, how
hard the interrogator tried to get the suspect to confess, how
friendly the interrogator was, how confrontational the interrogator
was, how anxious the suspect was, and how stressful they thought
the interrogation was for the suspect. One new continuous item
was added for Phase 2, which asked participants to rate how much
pressure the interrogator placed on the suspect during the interro-
gation. These nine continuous ratings were each given on a scale
from 1 (not at all)to10(very). Participants also gave dichotomous
yes/no judgments as to whether the suspect had made any suspi-
cious remarks and/or any admissions of guilt during the interro-
gation. Four participants (4.17%) neglected to answer one or more
items, so their data are missing from the relevant analyses.
Phase 2 Results
Two a priori sets of comparisons framed our analyses. First and
foremost, we sought to compare the two written accounts of the
interrogations—secondhand police reports versus verbatim tran-
scripts. Next we compared the two verbatim accounts—written
transcripts alone versus accompanied by an audiotape.
Written Police Reports Versus Transcripts
Our most important prediction was that reading transcripts of
interrogations would improve fact-finding accuracy relative to
reading the accounts contained in police incident reports by ren-
dering participants less likely to judge innocent suspects as guilty.
Supporting this prediction, participants in the Transcript condition
misjudged the suspect as guilty less often than did those in the
Report condition (9.38% vs. 31.25%, respectively), ␹
2
(1) ⫽4.73,
p⫽.030, ␸⫽.27, OR ⫽4.39 [95% CI: 1.08, 17.86].
We then compared the Transcript and Report conditions in
terms of their continuous ratings of the interrogator and suspect on
our questionnaire. A one-way MANOVA on these 10 items did not
reach a conventional level of significance, Wilks’ ⌳⫽.74, F(10,
47) ⫽1.65, p⫽.123, but it did reveal a large effect size (␩p
2⫽.26)
and suboptimal power (1-␤⫽.71). In light of this, and to more
precisely focus on our most pertinent system-relevant measures,
we performed univariate ttests designed to directly test our pre-
diction that presentation medium would in particular impact per-
ceptions of guilt and coercion.
Significant univariate differences were found on three items,
each of which showed a medium to large effect size (see Figure 3;
Cohen, 1988). Consistent with the aforementioned binary judg-
ments, participants who read a transcript were more confident in
the suspect’s innocence (M⫽5.86, SD ⫽4.82) than were those
who read the corresponding police report (M⫽2.34, SD ⫽7.12),
t(56) ⫽2.20, p⫽.032, d⫽0.59 [95% CI: ⫺0.95, 2.14]. Impor-
tantly as well, compared to those in the Report condition, partic-
ipants who read a transcript also rated the interrogator as having
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.
238 KASSIN, KUKUCKA, LAWSON, AND DECARLO

exerted more pressure on the suspect (Ms⫽6.93 and 5.59, SDs⫽
2.36 and 2.43), t(56) ⫽2.14, p⫽.037, d⫽0.57 [95% CI: 0.03,
1.18], and as having tried harder to obtain a confession (Ms⫽7.59
and 5.79, SDs⫽2.50 and 2.70), t(56) ⫽2.62, p⫽.011, d⫽0.70
[95% CI: 0.04, 1.36].
No differences were found on the other continuous ratings, all
ts⬍1.15, ps⬎.26. Overall, participants in the Transcript and
Report conditions, respectively, rated the interrogation as moder-
ately stressful (Ms⫽5.28 and 5.66, SDs⫽2.83 and 3.11) and the
suspect as moderately anxious (Ms⫽5.24 and 5.48, SDs⫽2.69
and 2.89), credible (Ms⫽6.72 and 6.21, SDs⫽1.79 and 2.32),
and knowledgeable (Ms⫽5.48 and 5.38, SDs⫽2.52 and
2.62)—and as highly cooperative (Ms⫽8.52 and 8.14, SDs⫽
1.38 and 2.01). Transcript and Report participants also rated the
interrogator both as somewhat friendly (Ms⫽6.07 and 5.31,
SDs⫽2.88 and 2.27) and somewhat confrontational (Ms⫽6.07
and 5.41, SDs⫽2.30 and 2.15).
Two additional questionnaire items asked whether suspects had
made any suspicious remarks or admissions of guilt. Participants in
the Transcript condition were no more or less likely than those in
the Report condition to believe that the suspect had made suspi-
cious remarks (37.50% vs. 28.13%, respectively), ␹
2
(1) ⫽0.64,
p⫽.424, ␸⫽.10, OR ⫽1.53 [95% CI: 0.54, 4.39], nor were they
more or less likely to believe that he had admitted guilt (13.79%
vs. 6.90%, respectively), ␹
2
(1) ⫽0.74, p⫽.389, ␸⫽.11, OR ⫽
2.16 [95% CI: 0.36, 12.84].
Verbatim Transcripts Versus Audiotapes
Next, we compared the judgments of participants in the Tran-
script and Audio conditions who received the same verbal content
but presented in a different modality (in writing vs. audio record-
ing). Interestingly, we found that participants in the Audio condi-
tion misjudged the suspect as guilty more often than did those in
the Transcript condition (37.50% vs. 9.38%, respectively), ␹
2
(1) ⫽
7.05, p⫽.008, ␸⫽.33, OR ⫽5.80 [95% CI: 1.45, 23.23], a
difference that is not surprising in light of research showing that
people are not intuitively accurate judges of truth and deception.
Audio participants were not, however, more likely to believe that
their suspect had made suspicious remarks (53.13% vs. 37.50%,
respectively), ␹
2
(1) ⫽1.58, p⫽.209, ␸⫽.16, OR ⫽1.89 [95%
CI: 0.70, 5.12], or that he had made an admission of guilt (12.90%
in both conditions), ␹
2
(1) ⫽0.00, p⫽1.00, ␸⫽.00, OR ⫽1.00
[95% CI: 0.23, 4.42].
A one-way MANOVA on continuous ratings of the interrogator
and suspect did not reach significance, Wilks’ ⌳⫽.73, F(10,
51) ⫽1.85, p⫽.075. Once again, however, this analysis revealed
a large effect size (␩p
2⫽.27) and a relative lack of power (1-␤⫽
.78). Follow-up ttests on the most relevant continuous items
revealed two significant univariate differences with medium to
large effect sizes. First, participants in the Audio condition were
less confident in the suspect’s innocence (M⫽1.19, SD ⫽7.55)
than were those in the Transcript condition (M⫽5.97, SD ⫽
4.68), t(60) ⫽2.99, p⫽.004, d⫽0.77 [95% CI: ⫺0.77, 2.31].
Second, participants in the Audio condition rated the interrogator
as having exerted less pressure on the suspect (M⫽5.39, SD ⫽
3.13) than did those in the Transcript condition (M⫽6.89, SD ⫽
2.46), t(60) ⫽2.10, p⫽.040, d⫽0.54 [95% CI: ⫺0.15, 1.23]. No
significant differences were found on the remaining items, all ts⬍
1.87, ps⬎.06.
Discussion
In light of the problems associated with confession evidence,
numerous social scientists, legal scholars, and practitioners have
recommended a policy reform that would require the electronic
recording of entire suspect interviews and interrogations—not just
the resulting confessions (e.g., American Psychology-Law Society
white paper—Kassin et al., 2010). Historically, such a requirement
has proved controversial, drawing opponents from the law enforce-
ment community who have speculated about the possible adverse
effects on police, suspects, and juries (e.g., Inbau et al., 2001; for
an overview of opposition arguments see Sullivan, 2008;Thurlow,
2005).
The present research was designed to test for an important
possible benefit of audio recording full interrogations: that these
recordings provide a more accurate factual account of the interro-
gation behavior of police and suspects than would otherwise be
derived from police reports. Indeed, a common argument for
recording interrogations is that it is the most effective way to
memorialize the process by which a statement was taken and,
hence, increase the fact-finding accuracy of judges and juries. In
current practice, whereby detectives produce reports from contem-
poraneous or retrospective notes, disputes often arise as to a
number of issues concerning the behavior of police and suspects—
most notably, whether the police Mirandized suspects in a timely
manner, whether they used certain coercive tactics, and whether
0
1
2
3
4
5
6
7
8
Guilt-Confidence Composite Interrogator Effort Pressure on Suspect
Self-Reported Rang (1-10)
Report Transcript
Figure 3. Comparison of written transcripts and reports on Phase 2 perceptions of interrogators and suspects.
Note. Higher scores on the composite measure indicate greater confidence in the suspect’s innocence.
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.
239
POLICE REPORTS OF MOCK INTERROGATIONS

the details contained within a confession originated with the sus-
pect or came about through a process of contamination.
Police notes from interviews and interrogations are routinely ac-
cepted in the courts in lieu of an objective recording. Yet the accuracy
and impact of these notes has never been tested in the context of a
suspect interrogation. In light of the importance of the policy in
question as well as basic research on memory for conversations, we
conducted a two-phased experiment. In Phase 1, experienced police
officers investigated a mock crime scene, interviewed two innocent
suspects— one of whom was presented as suspicious on the basis of
his behavior—and then filed an incident report. All sessions were
covertly audio recorded; these recordings were later used to assess the
accuracy of the reports. In Phase 2, lay participants read either the
police report or the interrogation transcript, or they listened to an
audiotape. Our goal was to compare the accuracy of these “fact
finders” as a function of the information they were provided.
In Phase 1, two notable sets of results were obtained. First, in
parallel postinterrogation questionnaires, police and their suspects
diverged in their perceptions of the suspect’s behavior. Specifi-
cally, suspects believed that they would be seen as more innocent
than the police actually believed them to be; they also rated their
denials as more credible, their stress levels as lower, and their
behavior as more cooperative. Along with the finding that police
judged 41% of suspects as guilty, and has having made suspicious
remarks, this result demonstrates a phenomenon often seen in real
trials: That detectives and defendants often testify disparately as to
what transpired during the process of interrogation.
The second key result emerged from the comparison between the
tactics that police actually used during their interrogations, as later
coded from the tapes, and the tactics police said they used in their
incident reports. Consistent with past research using a mock crime
paradigm (e.g., Kassin et al., 2014), in naturalistic field settings
involving real interrogations (e.g., Feld, 2013;Leo, 1996), and in
self-report surveys of police (e.g., Kassin et al., 2007), we found that
police participants commonly used confrontation, maximization, min-
imization and leniency, and presentations of false evidence in the
form of bluffing and outright lies— even in the context of interroga-
tions that were brief and without high-stakes consequences.
In this regard, two important points are worth noting. First,
closely replicating the mock interrogation results reported by Kas-
sin et al. (2014), the police participants in our study used the false
evidence ploy at a very high rate. This result contrasts with more
realistic field-based data (e.g., Leo, 1996) and self-report surveys
(e.g., Kassin et al., 2007) indicating that this tactic is used more
sparingly. Second, when the usage frequencies were compared
with the incident reports, even though police participants were
instructed to state as close to verbatim as possible what tactics they
used, numerous errors of omission were observed. Specifically,
whereas most interrogations involved the use of confrontation,
maximization, leniency, and false evidence at rates of 56%, 59%,
63%, and 84%, respectively, corresponding reports described these
tactics only 22%, 16%, 40%, and 67% of the time, respectively. In
actual cases, such underreporting, whether purposeful or inadver-
tent, could influence judges who rule on the voluntariness of
statements taken and juries who rule on their credibility.
In Phase 2, we sought primarily to determine whether outside
fact finders would form different or more accurate impressions
when they had access to written transcripts of suspect interviews
compared to the written police reports. On the most important
dependent measure, perceptions of the suspect’s guilt or inno-
cence, the results confirmed expectations: Participants were sig-
nificantly less likely to misjudge the innocent suspect as guilty
when they read a written transcript of the interrogation than when
they read the corresponding police report (9% vs. 31%). Moreover,
participants who read a police report, compared to those who read
the full transcript, believed that police officers had applied less
pressure on the suspect and tried less hard to get the suspect to
confess. In short, observers whose information basis was one of
the police reports relative to one of the transcripts, saw the process
as less pressure-filled and the innocent suspect as more guilty.
In addition to comparing the two alternative forms of written
information—police reports and transcripts—we wondered what
impressions observers would have if they listened to an interroga-
tion audiotape. On the one hand, this richer medium of presenta-
tion presents exactly the same verbal content along with the
transcripts—perhaps, therefore, with the same result. On the other
hand, audiotapes add complex vocal and paralinguistic cues from
both police and suspect participants—the kinds of cues that often
lead laypeople to draw unwarranted and erroneous inferences of
truth and deception (e.g., DePaulo et al., 2003;Bond & DePaulo,
2006). Suggesting that the latter cues influenced perceptions, apart
from verbal content, several results showed that participants who
heard an audiotape, relative to those who merely read a transcript,
fell prey to some of these effects. Most notably, for example, 38%
misjudged the suspect to be guilty, a number that was significantly
higher than in the transcript-only condition.
As a matter of policy, one could argue from our results that
perhaps strictly content-focused interrogation transcripts—without
access to audio and visual cues—would provide a sufficient basis
for fact finding. Such a conclusion would not be warranted. While
a transcript fully communicates the verbal text of a police–suspect
interaction, it does not depict potentially important aspects of the
suspect (e.g., his or her physical condition, appearance, attire,
voice, and demeanor; whether he or she is seated in a corner or
handcuffed) or the police officers (e.g., their number, size, and
proximity to the suspect; whether they are uniformed or in plain
clothes; whether weapons are visible; whether they raise their
voices). In this regard, extensive research indicates that fact finders
render more balanced and accurate judgments from “equal focus”
video recordings that show both the suspect and police rather than
one or the other (Lassiter, Diamond, Schmidt, & Elek, 2007;
Lassiter, Geers, Handley, Weiland, & Munhall, 2002).
Summary, Implications, and Limitations
The most important signal to emerge from our study is one that
strongly supports a requirement that all suspect interviews and
interrogations be recorded and transcribed in order to provide a
more accurate account of the process and improve the fact-finding
performance of judges and juries. This signal is embodied in the
following main findings: (a) Police and suspects diverged in their
perceptions of the suspect’s behavior during the interrogation
sessions in which both parties were present; (b) Police committed
frequent errors of omission in their Incident Reports, underreport-
ing their use of confrontation, maximization, leniency, and pre-
sentations of false evidence; and (c) Phase 2 participants who read
a police report, compared to those who read a full verbatim
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.
240 KASSIN, KUKUCKA, LAWSON, AND DECARLO

transcript, perceived the process as less pressure-filled and were
more likely to misjudge innocent suspects as guilty.
Taken as a whole, these findings help to explain the second
problem with false confessions (the first being that they occur):
That they are too often believed by judges and juries equipped only
with secondhand information about the process by which the
statements were elicited (Kassin, 2012). On this issue, results from
Phase 2 suggest two important points. First, participant observers
were aided by having access to the more accurate content provided
by verbatim transcripts, suggesting, perhaps, this previously ne-
glected potential benefit of electronic recording. Second, partici-
pant observers did not benefit from the addition of audio record-
ings of the interrogations, leaving open the empirical question,
untested in our study, of whether additional access to equal-focus
video would improve fact-finding performance (for a discussion,
see Snyder, Lassiter, Lindberg, & Pinegar, 2009).
One might wonder whether the errors of omission we observed, in
the form of underreporting the use of interrogation tactics, were
purposeful or inadvertent. As a strictly empirical matter, our data do
not permit us to weigh in on this issue. However, given the context
(i.e., these errors were observed in a mock-crime-and-investigation
study; police participants wrote Incident Reports that would not later
be shown to a prosecutor, entered into evidence at trial, or become the
subject of sworn testimony), common sense would suggest that these
errors were inadvertent. Such inadvertence is often seen in the con-
tamination of false confessions with accurate crime details, a phe-
nomenon observed in actual cases (to illustrate, see Trainum, 2007;
for reviews, see Garrett, 2010,2015; also see Lamb et al., 2000). In
this regard, it is important to note that we did not test for—and our
estimates do not account for the possibility of— conscious or moti-
vated failures of police to recall or report certain aspects of the
interrogations they conduct. Because we asked for all reports to be
submitted within 48 hours, we also did not test for the possibility that
reporting errors would increase as a function of longer intervals
between interrogations and reports.
The present research is potentially limited in important ways.
Our police participants were trained and seasoned professionals,
ranging in age from 26 to 55 and having an average of 16 years of
law enforcement experience, which included numerous suspect
interviews. In the context of our mock-crime-and-investigation
paradigm, however, these participants were limited to 5 minutes
for crime scene analysis and two relatively brief interrogations—
each lasting an average of only 17 min, and a 48-hr time limit for
submission of reports. Whether our results would generalize to
longer and more consequential investigations remains an important
empirical question for further research. On the one hand, one
might argue that our police participants were not as motivated to
recall the details of their mock interrogations as they are in actual
practice, causing us to overestimate errors of omission and their
effects. On the other hand, one might argue from a cognitive
perspective that because it is so much easier to recall brief con-
versations that last only a few minutes, rather than hours-long
interrogations, our results underestimate the potential for errors of
omission and do not adequately test for possible errors of com-
mission, including source attribution errors.
Our results are also limited by the fact that our manipulation of
prior suspicion had no significant effects whatsoever—not on the
length of interrogations in minutes or words, participants’ perceptions
of the exchange, police perceptions of the suspect’s guilt, or the coded
interrogation tactics that were used. We had hoped to assess whether
the accuracy of police reports and the impressions they elicited in
observers were moderated by investigators’ preexisting degree of
suspicion. In prior research, suspicion was varied by the presentation
of a base rate (i.e., 80 vs. 20% of suspects in this study are guilty of
the mock crime; see Kassin, Goldstein, & Savitsky, 2003). In this
study, however, because each investigator was set to conduct two
interrogations, we manipulated relative suspiciousness by describing
the demeanor of the two suspects— one as calm, the other as anxious
and evasive. Lacking a manipulation check to ensure that investiga-
tors drew the intended inferences from the experimenter’s description
of the two suspects, we cannot adequately evaluate the possible
effects of suspicion on accuracy and bias in police reports. Testing this
confirmation bias hypothesis thus remains an important avenue for
follow-up research.
We should also comment on the mock-crime-and-investigation
paradigm we used, modeled after that previously reported in Kas-
sin et al. (2014). Conducted in a police station and involving a
collaboration of experienced law enforcement participants, this
experiment contained a high level of ecological validity. As noted
by Kassin et al. (2014), however, such data are difficult to collect
during the workday from on-duty and off-duty officers, detectives,
and sergeants. For that reason, our sample was smaller than we had
hoped it would be, thereby limiting the power of our study and
hence our ability to analyze for individual differences in experi-
ence or training among police participants.
One final limitation concerns the medium through which Phase 2
participants, mimicking fact finders, made their judgments. Specifi-
cally, they read a brief description of the crime followed by the “raw
data” of a police report or verbatim transcript or audio recording of an
interrogation. Police participants who had interrogated mock suspects
did not testify as to their experience—and they were not cross exam-
ined. Although research suggests that the process of cross examina-
tion may have variable effects—for example, helping jurors to be-
come more discerning of scientific experts (Austin & Kovera, 2015),
yet impairing an eyewitness’s memory (Valentine & Maras,
2011)—it is nevertheless the natural process through which fact
finders are informed about interviews and interrogations. More re-
search is needed to determine if cross-examination serves to correct
for the underreporting of tactics used.
With an accumulation of DNA exonerations illuminating the
problem of false confessions, and with research indicating the dual
risk that innocent people might confess to crimes they did not
commit and that judges and juries may well believe these false
confessions, it is easy to understand calls to reform that focus on
the recording of interrogations. Limitations notwithstanding, the
present study adds to a growing empirical literature indicating the
need for such a requirement, if only to ensure the accuracy and
completeness of memorial accounts of key transactions between
police and their suspects—accounts that form the basis of deci-
sions routinely made by judges and juries.
References
American Bar Association. (2004). Resolution 8A—Videotaping custodial
interrogations. Approved February, 9, 2004.
American Psychological Association. (2014). Resolution on interrogations
of criminal suspects. Retrieved from http://www.apa.org/about/policy/
interrogations.aspx
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.
241
POLICE REPORTS OF MOCK INTERROGATIONS

Appleby, S. C., Hasel, L. E., & Kassin, S. M. (2013). Police-induced
confessions: An empirical analysis of their content and impact. Psychol-
ogy, Crime & Law, 19, 111–128. http://dx.doi.org/10.1080/1068316X
.2011.613389
Austin, J. L., & Kovera, M. B. (2015). Cross-examination educates jurors
about missing control groups in scientific evidence. Psychology, Public
Policy, and Law, 21, 252–264. http://dx.doi.org/10.1037/law0000049
Boetig, B. P., Vinson, D. M., & Weidel, B. R. (2006). Revealing incom-
municado. FBI Law Enforcement Bulletin, 75, 1– 8.
Bond, C. F., Jr., & DePaulo, B. M. (2006). Accuracy of deception judg-
ments. Personality and Social Psychology Review, 10, 214 –234. http://
dx.doi.org/10.1207/s15327957pspr1003_2
Bruck, M., Ceci, S. J., & Francoeur, E. (1999). Accuracy of mothers’
memories of conversations with their preschool children. Journal of
Experimental Psychology: Applied, 5, 89 –106. http://dx.doi.org/10
.1037/1076-898X.5.1.89
Buckley, D. M., & Jayne, B. C. (2005). Electronic recording of interro-
gations. Eagle River, WI: Hahn Printing,.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences
(2nd ed.). Hillsdale, NJ: Erlbaum.
Deffenbacher, K. A., Bornstein, B. H., Penrod, S. D., & McGorty, E. K.
(2004). A meta-analytic review of the effects of high stress on eyewit-
ness memory. Law and Human Behavior, 28, 687–706. http://dx.doi.org/
10.1007/s10979-004-0565-x
DePaulo, B. M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton,
K., & Cooper, H. (2003). Cues to deception. Psychological Bulletin,
129, 74 –118. http://dx.doi.org/10.1037/0033-2909.129.1.74
Drizin, S. A., & Reich, M. J. (2004). Heeding the lessons of history: The
need for mandatory recording of police interrogations to accurately
assess the reliability and voluntariness of confessions. Drake Law Re-
view, 52, 619 – 646.
Federal Rules of Evidence. (2015). Arlington, VA: Federal Evidence
Review. Retrieved from http://federalevidence.com/downloads/rules.of
.evidence.pdf
Feld, B. C. (2013). Kids, cops, and confessions: Inside the interrogation
room. New York, NY: New York University Press.
Garrett, B. L. (2010). The substance of false confessions. Stanford Law
Review, 62, 1051–1119.
Garrett, B. L. (2015). Contaminated confessions revisited. Virginia Law
Review, 101, 395– 453.
Geller, W. A. (1993). Videotaping interrogations and confessions: A
preliminary examination of issues. Washington, DC: National Institute
of Justice.
Gudjonsson, G. H. (2003). The science of interrogations and confessions:
A handbook. Chichester, UK: Wiley.
Inbau, F., Reid, J., Buckley, J. P., & Jayne, B. C. (2001). Criminal
interrogation and confessions (4th ed.). Gaithersburg, MD: Aspen.
Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitor-
ing. Psychological Bulletin, 114, 3–28. http://dx.doi.org/10.1037/0033-
2909.114.1.3
Kassin, S. M. (1997). The psychology of confession evidence. American Psychol-
ogist, 52, 221–233. http://dx.doi.org/10.1037/0003-066X.52.3.221
Kassin, S. M. (2005). On the psychology of confessions: Does innocence
put innocents at risk? American Psychologist, 60, 215–228. http://dx.doi
.org/10.1037/0003-066X.60.3.215
Kassin, S. M. (2012). Why confessions trump innocence. American Psy-
chologist, 67, 431– 445. http://dx.doi.org/10.1037/a0028212
Kassin, S. M., Drizin, S. A., Grisso, T., Gudjonsson, G. H., Leo, R. A., &
Redlich, A. D. (2010). Police-induced confessions: Risk factors and
recommendations. Law and Human Behavior, 34, 3–38. http://dx.doi
.org/10.1007/s10979-009-9188-6
Kassin, S. M., Goldstein, C. C., & Savitsky, K. (2003). Behavioral con-
firmation in the interrogation room: On the dangers of presuming guilt.
Law and Human Behavior, 27, 187–203. http://dx.doi.org/10.1023/A:
1022599230598
Kassin, S. M., & Gudjonsson, G. H. (2004). The psychology of confession
evidence: A review of the literature and issues. Psychological Science in
the Public Interest, 5, 33– 67. http://dx.doi.org/10.1111/j.1529-1006
.2004.00016.x
Kassin, S. M., Kukucka, J., Lawson, V. Z., & DeCarlo, J. (2014). Does
video recording alter the behavior of police during interrogation? A
mock crime-and-investigation study. Law and Human Behavior, 38,
73– 83. http://dx.doi.org/10.1037/lhb0000047
Kassin, S. M., Leo, R. A., Meissner, C. A., Richman, K. D., Colwell, L. H., Leach,
A.-M., & La Fon, D. (2007). Police interviewing and interrogation: A self-report
survey of police practices and beliefs. Law and Human Behavior, 31, 381– 400.
http://dx.doi.org/10.1007/s10979-006-9073-5
Kleider, H. M., Pezdek, K., Goldinger, S. D., & Kirk, A. (2008). Schema-
driven source misattribution errors: Remembering the expected from a
witnessed event. Applied Cognitive Psychology, 22, 1–20. http://dx.doi
.org/10.1002/acp.1361
Lamb, M. E., Orbach, Y., Sternberg, K. J., Hershkowitz, I., & Horowitz, D.
(2000). Accuracy of investigators’ verbatim notes of their forensic
interviews with alleged child abuse victims. Law and Human Behavior,
24, 699 –708. http://dx.doi.org/10.1023/A:1005556404636
Lassiter, G. D., Diamond, S. S., Schmidt, H. C., & Elek, J. K. (2007).
Evaluating videotaped confessions: Expertise provides no defense
against the camera-perspective effect. Psychological Science, 18, 224 –
226. http://dx.doi.org/10.1111/j.1467-9280.2007.01879.x
Lassiter, G. D., Geers, A. L., Handley, I. M., Weiland, P. E., & Munhall, P. J.
(2002). Videotaped interrogations and confessions: A simple change in camera
perspective alters verdicts in simulated trials. Journal of Applied Psychology, 87,
867– 874. http://dx.doi.org/10.1037/0021-9010.87.5.867
Lassiter, G. D., & Meissner, C. A. (Eds.). (2010). Police interrogations and
false confessions: Current research, practice, and policy recommenda-
tions. Washington, DC: American Psychological Association. http://dx
.doi.org/10.1037/12085-000
Leo, R. A. (1996). Inside the interrogation room. Journal of Criminal Law
and Criminology, 86, 266 –303.
MacLeod, C. M., Gopie, N., Hourihan, K. L., Neary, K. R., & Ozubko,
J. D. (2010). The production effect: Delineation of a phenomenon.
Journal of Experimental Psychology: Learning, Memory, and Cogni-
tion, 36, 671– 685. http://dx.doi.org/10.1037/a0018785
Morgan, C. A., III, Hazlett, G., Doran, A., Garrett, S., Hoyt, G., Thomas,
P., & Southwick, S. M. (2004). Accuracy of eyewitness memory for
persons encountered during exposure to highly intense stress. Interna-
tional Journal of Law and Psychiatry, 27, 265–279. http://dx.doi.org/10
.1016/j.ijlp.2004.03.004
Neisser, U. (1981). John Dean’s memory: A case study. Cognition, 9, 1–22.
http://dx.doi.org/10.1016/0010-0277(81)90011-1
Schacter, D. L. (2001). The seven sins of memory: How the mind forgets
and remembers. Boston, MA: Houghton Mifflin.
Schmidt, M. S. (2014, May 23). In policy change, justice dept. to require
recording of interrogations. The New York Times, p. A14.
Snyder, C. J., Lassiter, G. D., Lindberg, M. J., & Pinegar, S. K. (2009).
Videotaped interrogations and confessions: Does a dual-camera ap-
proach yield unbiased and accurate evaluations? Behavioral Sciences &
the Law, 27, 451– 466. http://dx.doi.org/10.1002/bsl.875
Sullivan, T. P. (2004). Police experiences with recording custodial inter-
rogations. Chicago, IL: Northwestern University Law School, Center on
Wrongful Convictions.
Sullivan, T. P. (2008). Recording federal custodial interviews. The Amer-
ican Criminal Law Review, 45, 1297–1345.
Sullivan, T. P., Vail, A. W., & Anderson, H. W. (2008). The case for
recording police interrogation. Litigation, 34, 1– 8.
The Justice Project. (2007). Electronic recording of custodial interroga-
tions: A policy review. Washington, DC: Author.
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.
242 KASSIN, KUKUCKA, LAWSON, AND DECARLO

Thurlow, M. D. (2005). Lights, camera, action: Video cameras as tools of
justice. The John Marshall Journal of Computer & Information Law, 23,
771– 813.
Trainum, J. (2007, September 20). I took a false confession–so don’t tell
me it doesn’t happen! The California Majority Report. Retrieved from
http://seeingtheforest.com/i-took-a-false-confession-so-dont-tell-me-it-
doesnt-happen/
Valentine, T., & Maras, K. (2011). The effect of cross-examination on the
accuracy of adult eyewitness testimony. Applied Cognitive Psychology,
25, 554 –561. http://dx.doi.org/10.1002/acp.1768
Appendix
Self-Report Questionnaire
1) In your opinion, is the suspect you just interrogated guilty or
innocent?
(In your opinion, did the officer who interrogated you believe
you were guilty or innocent?)
2) How confident are you in this opinion?
(How confident are you that this was his/her opinion?)
3) In your opinion, how credible were (the suspect’s/your)
denials?
4) How much did this suspect seem to know about the crime?
(In your opinion, how much did you know about the crime
before you were questioned?)
5) In your opinion, how open and cooperative (was this suspect/
were you) in (his/your) attitude and demeanor?
6) Did (this suspect/you) make any suspicious or self-
incriminating remarks?
7) How hard did (you/the officer) try to get (this suspect/you)to
confess?
[Phase 2 only: How much pressure did the police officer place
on the suspect during the interrogation?]
8) How friendly, sympathetic, and understanding was (your
behavior toward this suspect/the officer toward you)?
9) How confrontational (were you toward this suspect/was the
officer toward you)?
10) How stressful do you think the interrogation was for this
suspect?
(How stressful was this interrogation for you?)
11) Overall, how anxious (was this suspect/were you) during the
interrogation?
12) Did (the suspect/you) make any admissions or confessions
of guilt?
Received October 24, 2015
Revision received October 2, 2016
Accepted October 4, 2016 䡲
E-Mail Notification of Your Latest Issue Online!
Would you like to know when the next issue of your favorite APA journal will be available
online? This service is now available to you. Sign up at https://my.apa.org/portal/alerts/ and you will
be notified by e-mail when issues of interest to you become available!
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.
243
POLICE REPORTS OF MOCK INTERROGATIONS