Clinical Psychology Review 34 (2014) 1428

Contents lists available at ScienceDirect

Clinical Psychology Review

Simulation of multiple personalities: A review of research comparing

diagnosed and simulated dissociative identity disorder
Guy A. Boysen a,, Alexandra VanBergen b
a
b

Department of Psychology, McKendree University, USA

State University of New York at Fredonia, USA

H I G H L I G H T S

Twenty studies compared people diagnosed with DID to individuals simulating DID.
Aspects of symptoms, identities, and cognitive processes differed between groups.
Interidentity transfer of information occurred at similar rates in both groups.
Several methodological improvements are needed in simulation research.

a r t i c l e

i n f o

Article history:
Received 18 June 2013
Received in revised form 28 October 2013
Accepted 30 October 2013
Available online 5 November 2013
Keywords:
Dissociative Identity Disorder
Dissociation
Simulation
Malingering
Etiology

a b s t r a c t
Dissociative Identity Disorder (DID) has long been surrounded by controversy due to disagreement about its etiology and the validity of its associated phenomena. Researchers have conducted studies comparing people diagnosed with DID and people simulating DID in order to better understand the disorder. The current research
presents a systematic review of this DID simulation research. The literature consists of 20 studies and contains
several replicated ndings. Replicated differences between the groups include symptom presentation, identity
presentation, and cognitive processing decits. Replicated similarities between the groups include interidentity
transfer of information as shown by measures of recall, recognition, and priming. Despite some consistent ndings, this research literature is hindered by methodological aws that reduce experimental validity.
2013 Elsevier Ltd. All rights reserved.

Contents
1.

2.
3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.1.
Disagreements about DID and dissociation . . . . . . . . . . . . .
1.2.
Evidence provided by simulation studies . . . . . . . . . . . . . .
1.3.
The current research . . . . . . . . . . . . . . . . . . . . . . .
Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.
Methodological controls . . . . . . . . . . . . . . . . . . . . .
3.1.1.
Control group matching . . . . . . . . . . . . . . . . .
3.1.2.
Control group training . . . . . . . . . . . . . . . . . .
3.1.3.
Manipulation checks . . . . . . . . . . . . . . . . . .
3.1.4.
Blinding . . . . . . . . . . . . . . . . . . . . . . . .
3.2.
Differences between diagnosed and simulated DID . . . . . . . . .
3.2.1.
Cognitive processing decits . . . . . . . . . . . . . . .
3.2.2.
Symptom reporting . . . . . . . . . . . . . . . . . . .
3.2.3.
Trauma-focused vs. non trauma-focused identity performance
3.2.4.
Eyesight variability . . . . . . . . . . . . . . . . . . .
3.2.5.
Effect sizes . . . . . . . . . . . . . . . . . . . . . . .
3.3.
Similarities between diagnosed and simulated DID . . . . . . . . .
3.3.1.
Interidentity transfer of information . . . . . . . . . . .
3.3.2.
Specic cognitive effects . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

Corresponding author at: Department of Psychology, McKendree University, 701 College Rd., Lebanon, IL 62254, USA. Tel.: +1 618 537 6882.
E-mail address: gaboysen@mckendree.edu (G.A. Boysen).
0272-7358/$ see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.cpr.2013.10.008

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

15
15
16
16
17
17
17
17
17
17
21
21
21
22
22
22
22
22
22
22

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

4.

Discussion . . . . . . . . . . . . . . .
4.1.
DID and simulator group differences
4.2.
DID and simulator group similarities
4.3.
Research methodology . . . . . . .
4.4.
Limitations . . . . . . . . . . . .
5.
Conclusion . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

1. Introduction
Dissociative Identity Disorder (DID) maintains a unique place in
the eld of psychopathology; it elicits an unprecedented mixture of
acceptance and rejection in the scientic community. Beginning with
acceptance, DID is an ofcially recognized diagnosis in the Diagnostic
and Statistical Manual of Mental Disorders, Fifth Edition (DSM; American
Psychiatric Association, 2013) and the International Classication of
Diseases (ICD-10; World Health Organization, 1992). In addition, the
scientic study of DID produces a modest but steady stream of publications documenting cases of the disorder in countries throughout the
world (Boysen & VanBergen, 2013). On the other hand, some have
dismissed the idea of multiple personalities as an incredible folly
(Piper & Merskey, 2004), and others have argued that interest in dissociative disorders was a scientic fad that peaked in the 1990s and then
quickly faded (Paris, 2012; Pope, Barry, Bodkin, & Hudson, 2006).
One of the longest-standing controversies about DID, however, is if it
represents a socially-enacted role or a special case of posttraumatic dissociation (Spanos, 1994; Spanos, Weekes, & Bertrand, 1985). Can DID be
exhibited after normal social learning processes or is it somehow
unique? One method of exploring this issue would be to compare people
with diagnoses of DID to individuals who are intentionally faking
the symptoms of DID. Differences between these groups would provide evidence for DID's unique nature, and similarities would suggest
a less-than-exceptional nature. The results could also have implications for the understanding of DID's etiology, diagnosis, and basic
features. The purpose of this review was to examine all existing research comparing individuals diagnosed with DID to individuals simulating DID in order to indentify reliable similarities and differences
between the groups.

1.1. Disagreements about DID and dissociation

Etiology is at the center of disagreement about DID. The posttraumatic explanation posits that DID is a reaction to intense trauma,
typically occurring in childhood (Gleaves, 1996; Putnam, 1989). The
model proposes that dissociation functions as a defense mechanism in
the face of inescapable psychological distress, and DID occurs when
this tendency to dissociate becomes too pervasive. Evidence for the
posttraumatic model comes from several different sources. A primary
source of evidence is the high proportion of people with dissociative
disorders who report childhood abuse or trauma (Foote, Smolin,
Kaplan, Legatt, & Lipschitz, 2006; Lewis, Yeager, Swica, Pincus, &
Lewis, 1997; Sar et al., 2007). In fact, studies have documented
the traumadissociation relation across several cultures (Tamar-Gurol,
Sar, Karadag, Evren, & Karagoz, 2008; Xiao et al., 2006). Furthermore,
there is evidence that sexual abuse is uniquely tied to dissociation
among children (Kisiel & Lyons, 2001). Such evidence is cogent, but it
is also correlational and frequently based on retrospective self-reports,
which leaves the posttraumatic model open to criticism.
Skepticism about DID is multifaceted, but the central concern among
critics is that social factors, rather than trauma, may be the predominant
cause of the disorder. Several trends point to DID's origins as a social
phenomena (Lilienfeld et al., 1999; Piper & Merskey, 2004; Spanos,
1994). Due to media depictions in the latter half of the 20th century,

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

15

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

.
.
.
.
.
.
.

23
23
24
25
26
26
26

the concept of having multiple personalities became part of the cultural

landscape in Western countries; at the same time, the prevalence of the
disorder dramatically increased. As cases became more frequent, so did
some of the more incredible aspects of the DID phenomena. For example, the reported number of alternative identities increased and so did
the scope of abuse allegations. Criticism has often focused on the treatment of DID as a potential iatrogenic factor. Some clinicians appear
to diagnose DID disproportionately more than others, and the typical
presentation of DID is unusual in that the core symptoms tend to
emerge only after treatment has started. The DSM states that only a
small minority [of patients] present to clinical attention with observable
alteration of identities (American Psychiatric Association, 2013,
p. 292). In fact, documentation of DID cases that have emerged outside
of treatment appears to be particularly challenging (Boysen, 2011;
Boysen & VanBergen, 2013). Direct support for the power of social inuence comes from experimental laboratory studies indicating that social
cues can lead individuals to enact some of the basic phenomena of DID
such as amnesia and the adoption of an alternative identity (Spanos,
1994; Spanos, Weekes, & Bertrand, 1985). Taken together, these pieces
of evidence have led to the proposal of a sociocognitive model stating
that DID is a social role enacted due to the inuence of culturallydetermined rules for expressing multiple selves (Spanos, 1994). Thus,
DID symptoms are learned in much the same way as any other social
behavior.
Considering the divergent etiological models, a primary subject of
disagreement is if DID's signs and symptoms are somehow special and
unique. Skeptics argue that the exhibition of multiple personalities
consists of acting out a known social role and is brought forth through
nonpathological forms of social inuence (Lilienfeld et al., 1999;
Spanos, 1994); this is not to say that the psychopathology of DID is
not real. Rather, skeptics argue that DID does not need to be explained
as a specialized defense mechanism used in reaction to trauma.
Although some trauma-focused theorists agree that social factors can
be integrated into their models (Sar & Ozturk, 2007), a central assertion
of the trauma model is that enactment of a social role in the absence of
trauma can only resemble DID on a supercial level (Gleaves, 1996). For
example, research showing that undergraduate students can be inuenced into exhibiting symptoms of DID in the laboratory (e.g., Spanos,
Weekes, & Bertrand, 1985) does not directly show that their behavior
is analogous to people diagnosed with DID (Gleaves, 1996). Unlike
many of the conicts between supporters of the sociocognitive and
trauma models, there is empirical evidence that can offer clarication
in this case. Studies comparing individuals diagnosed with DID and individuals attempting to simulate DID have existed in the literature
for almost as long as it has been an ofcial diagnosis in the DSM
(e.g., Coons, Milstein, & Marley, 1982). Comparing people diagnosed with DID to people simulating DID can directly demonstrate
the ways in which having a diagnosis and enacting a role are similar
or dissimilar.
The DID simulation literature can also inform the somewhat analogous debate about the general concept of dissociation. Proponents of
the trauma model of dissociation believe that it is a biologically-based
reaction to threat that includes experiences such as loss of conscious
control over behavior and amnesia, which are typical not only of DID
but also of many other posttraumatic stress reactions (Dalenberg
et al., 2012). Skeptics argue that special, trauma-based explanations

16

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

are not needed and that dissociation is actually associated with increased, but unexceptional, cognitive processing failures such as forgetting and inattention (Giesbrecht, Lynn, Lilienfeld, & Merckelbach, 2008).
Because DID is the hallmark dissociative disorder, the results of DID
simulation research should provide evidence related to the assertion that dissociation phenomena can only be explained as unique,
biologically-based reactions to trauma.
1.2. Evidence provided by simulation studies
A frequent criticism of DID simulation research is that it is incapable
of providing conclusive evidence for one etiological theory over another
(Merckelbach, Devilly, & Rassin, 2002). After all, the same behavior
might emerge for completely separate, unrelated causes. Thus, it is
important to consider how the results of simulation studies should
logically be interpreted. Research on hypnosis provides some useful
guidance for interpretation. Hypnosis simulation research has a long
history (Orne, 1959), and it has had a substantial impact on the understanding of hypnotic phenomena (Kirsch & Lynn, 1998). Theorists assumed for a long time that behaviors associated with hypnosis were
unique and impossible to intentionally simulate (Orne, 1971). However,
researchers began performing studies in which the experimenter was
blind to whether participants were actually hypnotized or acting as if
they were hypnotized (e.g., Green, Page, Handley, & Rasekhy, 2005;
Kirsch & Lynn, 1998; Naish, 2005; Orne, 1971). What they discovered,
to the surprise of many, was that simulated responses and hypnotic
responses were largely indistinguishable.
The results of hypnosis simulation studies provided important
theoretical information. Behaviors that emerged only in the hypnosis
group could be assumed to represent unique consequences to being
hypnotized. In contrast, failure to nd differences between hypnotized
and simulating individuals demonstrated that hypnosis is not required
to produce the behaviors in question, and normal social processes
cannot be eliminated as an explanation for hypnotized behavior
(Orne, 1971). For example, if people who are hypnotized and simulating
hypnosis both experience a reduction in pain sensitivity, social demands
cannot be eliminated as the cause of analgesic effects associated with
hypnosis. If pain reduction occurs in hypnosis but not hypnosis simulation, social demands are unlikely to fully explain the differences
between groups, which may be attributable to unique characteristics
of hypnosis.
Building off of the lessons of hypnosis-simulation research, what
can and cannot be learned from the simulation of DID? Beginning
with similarities between individuals diagnosed with DID and individuals simulating DID, they show that the phenomena of DID cannot be
considered unique. Similarities show that characteristics considered
central to DID, such as exhibiting multiple personalities and amnesia,
can be produced through normal learning and social inuence. In
laboratory experiments, social inuences may include such factors as
instructions, behaviors of the experimenters, and the experimental
setting itself. It is important to note, however, that similarities do not
provide direct evidence for the etiology of DID because the same behavior may result from different causes. Even though direct proof for one
model cannot be offered from simulation research, evidence that DID
phenomena can be created through social inuence is theoretically
important due the centrality of that proposition in the sociocognitive
model (Spanos, 1994).
The interpretation of differences between simulators and people
diagnosed with DID is also important to consider. Differences between
simulators and people diagnosed with DID provide evidence that DID
phenomena is not fully explained by the social inuence exerted
on simulators. However, simulators are likely to differ from people
diagnosed with DID in many ways; thus, confounds due to sample selection prevents simulation research from conclusively showing that
differences are a result of diagnosis rather than other differences
between the groups. Nonetheless, evidence against the ability of

simulators to meaningful enactment DID is theoretically important because it directly contradicts the sociocognitive model and indirectly
supports the trauma model's contention that DID is a unique
phenomena.
There are several other pieces of important information that might
be obtained from simulation research. It is possible that information
about other etiological factors may emerge. The most obvious example
would be if brain functioning in people diagnosed with DID differed
systematically from simulators; this might provide corroborative
evidence for areas of the brain previously identied as related to DID
(e.g., Reinders et al., 2006; Sar, Unal, Kiziltan, Kundakci, & Ozturk,
2001; Vermetten, Schmahl, Lindner, Loewenstein, & Bremner, 2006).
Some researchers have identied neurological correlates of switching
personalities in DID (Reinders et al., 2003); however, without a simulating control group it is impossible to determine if neurological effects are
unique to DID or if they occur in any group enacting the role of having
multiple selves. Simulation studies might also offer perspective on the
validity of the DID diagnostic criteria. According to the DSM, the core
features of DID are amnesia and the existence of different identity states
(American Psychiatric Association, 2013). However, other symptoms
are quite common but are, nonetheless, absent from the DSM criteria
(Dell, 2006). For example, somatoform, derealization, and hallucinatory
symptoms appear to be as common as the classic memory and identity
symptoms. Evidence that these symptoms differentiate simulated from
diagnosed cases of DID would provide an argument for revision of
the diagnostic criteria. Similarly, the research may provide guidance
for differential diagnosis. Some individuals mimic DID symptoms
because they are malingering or because they are exhibiting a factitious
disorder. In fact, factitious DID may be as prevalent as nonfactitious DID
(Friedl & Draijer, 2000). Given the evidence for frequent imitation of
DID (Draijer & Boon, 1999; Reinders, 2008), identifying key points in
differential diagnosis is particularly important. Overall, there is much
to learn about DID from a systematic review of the simulation literature.
Reviewing the DID simulation literature also provides an opportunity to evaluate the methodological rigor used by researchers in the eld.
Condence in the validity of research results must be moderated by the
quality of the methods used, and there are several key methodological
controls that are important to simulation research (Orne, 1971).
Perhaps the most important control is to engage in blinding procedures
so that the experimenter is unaware of participants' diagnostic group.
In terms of simulator preparation, researchers should provide strong
motivation for them to simulate effectively and emphasize the importance of convincing the experimenter that they are actually diagnosed
with DID. During and after the simulation, researchers should engage
in manipulation checks to ensure that simulators are following instructions and adequately performing their tasks. Other necessary controls
are related to more general methodological quality. Considering the
fact that participants in DID simulation research are likely to come
from very different populations, matching of groups on demographic
characteristics would help reduce the potential inuence of selection
confounds. DID is a complex phenomena; as such, people simulating
DID must be trained on how to perform its associated behaviors
and must be given adequate opportunity to practice. Finally, adequate
sample sizes are needed to ensure the validity of statistical conclusions.
1.3. The current research
The current research consisted of a review of all published studies
comparing individuals diagnosed with DID to individuals intentionally
simulating DID. Of primary interest were the similarities and differences
that emerged between the two groups. Similarities provide evidence
that social inuence cannot be ruled out as a factor in DID phenomena,
and differences provide evidence that DID is a unique phenomena not
easily explained by social inuence. However, there are a number of
potential confounds that would diminish the meaningfulness of the
study's results. As such, the methodological quality of DID simulation

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

research must be evaluated before considering the results of the studies.

Considering these issues, there were three primary research questions
for the review. One, what are the methodological controls used in DID
simulation research? Two, are there variables that reliably differentiate
between diagnosed and simulated DID? Three, are there variables
that reliably fail to differentiate between diagnosed and simulated
DID? Answers to these questions may inform diagnostic practices as
well as theories about the etiology of DID.
2. Method
Researchers searched the MEDLINE and PsycINFO databases for
this review. Search limiters included peer-reviewed journal articles
published in English. Search terms included dissociative identity disorder, multiple personality, and dissociative disorder combined with act,
simulation, fake, role play, enactment, mimic, imitation, factitious,
malinger, and feign. Researchers also examined the references of identied articles in order to locate studies that did not emerge using the
database searches. Inclusion criteria for the review were that the
study contain at least one participant diagnosed with DID and at least
one participant simulating the phenomena of DID in some way for the
purposes of the study (e.g., alternate identities, amnesia). Other dissociative disorders, high scores on dissociation measures, or dissociation
in general did not meet the DID inclusion criterion. Also, there are a
few studies comparing people diagnosed with DID and people who
have been identied as intentionally faking DID in clinical settings
(Coons & Milstein, 1994; Kluft, 1987). However, we chose to only include laboratory studies in this review due to their increased experimental control and conceptual similarity. In addition, individuals who
have been caught faking DID have, by denition, failed in their attempt
to simulation the disorder, which makes them a poor comparison group.
These methods led to the identication of 20 DID simulation studies
(see Table 1).
3. Results
The review consists of two major sections: study methods and study
results. The rst section reviews methodological controls utilized in DID
research. Examination of the research methods will help to provide a
frame of reference from which to evaluate research results in the second
part of the review. Subsequent sections include reviews of the reliable
differences and similarities that emerge between simulators and people
diagnosed with DID.
3.1. Methodological controls
3.1.1. Control group matching
A high quality comparison group is an essential methodological control in DID simulation studies. The quality of a control group can be evaluated by size and demographic similarity to the DID group (see Table 1).
In terms of sample size, the average number of people diagnosed with
DID per study was 14 (SD = 8), and the average number of simulators
was 20 (SD = 14). The average sample size in simulation studies is
somewhat lower than the pool of all recent research utilizing people diagnosed with DID, which has mean sample size of 17 (Boysen &
VanBergen, 2013). Overall, the sample sizes in DID simulation research
are small, but the DID and control groups are similar in size. In terms of
the comparability of these samples, there were systematic differences in
the recruitment of participants. Researchers recruited DID samples from
treatment settings. Of the studies specifying details about control
groups recruitment, the samples included volunteers recruited from
the general community (n = 5), health professionals recruited from
the hospitals at which studies were taking place (n = 5), volunteers recruited among the student population (n = 4), actors recruited for
their extensive experience (n = 2), and mixed sources (n = 1). Thus,
the origin of groups was not directly comparable in the majority of

17

studies and simulators always consisted of healthy controls. Most studies (n = 12) included some sort of matching of the samples. However,
matching was not extensive; the majority of studies only matched for
age (n = 9), sex (n = 10), and education (n = 8). Nine studies
matched on at least two of these demographic factors.
3.1.2. Control group training
DID is a complex phenomenon, and effective simulation of the disorder requires training. All simulators followed general instructions to
mimic DID; two exceptions were one group asked to simulate amnesia
by concealing knowledge (Brand, McNary, Loewenstein, Kolos, & Barr,
2006) and one group of actors instructed to role play specic patient
identities (Hopper et al., 2002). Specic wording of instructions or training scripts was not available in the articles (presumably due to space
constraints). The level of training offered to assist in effectively mimicking DID varied considerably. Of the 17 studies outlining a training method, 15 included exposure to educational information about DID in
written or video format (see Table 1). Eight of these studies included
both readings and videos in the training. Educational materials mostly
consisted of general readings or documentaries about DID, but three
studies used the book or movie version of Sybil (Brand et al., 2006;
Miller, 1989; Miller et al., 1991).
Training in 12 studies included opportunity to practice DID simulation (see Table 1). A common training procedure was to encourage control group members to fabricate details about their alternate identities
that corresponded to 17 specic attributes (Huntjens, Peters, et al.,
2005; Huntjens, Peters, Woertman, Van der Hart, & Postma, 2007;
Huntjens et al., 2002; Huntjens, Postma, Peters, Woertman, & Van der
Hart, 2003; Huntjens, Postma, Woertman, Van der Hart, & Peters,
2005; Putnam, Zahn, & Post, 1990; Reinders, Willemsen, Vos, Den
Boer, & Nijenhuis, 2012). Among the studies specifying the duration of
time allowed for practice, the modal amount was one week (Huntjens,
Peters, et al., 2005; Huntjens et al., 2007, 2003; Huntjens, Postma,
et al., 2005; Huntjens, Verschuere, & McNally, 2012), but times
ranged from one day (Silberman, Putnam, Weingartner, Braun, & Post,
1985) to two weeks (at minimum; Reinders et al., 2012). Although
the modal duration was one week, researchers did not provide details
on the actual hours spent on practice or the number of individual practice sessions, which are, perhaps, more valid measures of training
quality.
3.1.3. Manipulation checks
Although the term manipulation check typically refers to the procedures for determining if an independent variable had its intended effect
in experimental research, the concept can be applied in the current context to refer to control group members' ability to effectively simulate
DID. Some basic standard of performance must be set for the quality
of simulation. Otherwise, simulation studies are inherently biased toward nding differences between people diagnosed with DID and simulators. Ten studies included methods that could be broadly construed
as manipulation checks. Reinders' et al. (2012) study included the
most extensive system of checks; it included checks on simulators' compliance with practice instructions, assessment of the quality of their
development of identities, and self-evaluations of their actual performance during the study. Miller et al. (1991) conducted a simple but
elegant manipulation check by having examiners, who were blind to
participants' group membership, rate how much they believed the person they examined had DID. The most common check was testing simulators' knowledge about their identities (Huntjens, Peters, et al., 2005;
Huntjens et al., 2006, 2007, 2002, 2003; Huntjens, Postma, et al., 2005)
or about DID in general (Brand et al., 2006). Finally, Huntjens et al.
(2012) and Kong, Allen, and Glisky (2008) conducted post-study interviews of simulators to assess their effectiveness. Overall, manipulation
checks were minimal, and only one study included an evaluation of
the simulation as it was actually occurring (Miller et al., 1991).

18

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

Table 1
Study method and results for studies comparing people with DID to people simulating DID.
Study

DID/Control N

Control source
and matching

Task and training

Blinding

Manipulation
checks

Comparisons between DID

and simulator groups

Allen & Movius, 2000

4/60

College students
No matching

Simulate amnesia
None

No

No

Brand et al., 2006

20/43

College students
No matching

No

Yes, scores above

70% on test of DID
knowledge

Coons et al., 1982

2/1

Therapist of DID
patient
No matching

No

No

EEG amplitude (no direct

comparisons made, signicant EEG
differences existed within subjects)

Eich et al., 1997

7/9

No

No

Hermans et al., 2006

7/10

Mental health
professionals with
knowledge of DID
No matching
Students
Sex

Simulate DID
Watch or read Sybil,
complete Abnormal
Psychology course,
readings
Simulate DID
Therapist attempted
to mimic patient
personalities
Simulate DID
None

Recognition: forced choice test

(DID less recognition)a
Response latency for responses
(DID slower)a
Electroencephalogram (EEG) P3
amplitude (no statistical
comparisons)
Structured interview of Reported
Symptoms total score and 13
subscales (DID scores lower)a

No

No

Hopper et al., 2002

5/5

Professional actors
Sex, age

Simulate DID
such that one
personality had
memories
of trauma and
the other did not
Readings
Role play identities
Given summaries of
each identity

Perceptual priming (no direct

comparison, but DID showed
signicant priming and control
did not)a
Emotional Stroop task (DID more
efcient in trauma identity)a

No

No

Huntjens et al., 2002

26/25

Not specied
Sex, age, education

Simulate DID
Watched
documentary,
readings, created
identity
details, practiced
for 1 week

No

Yes, researcher
examined
17-item survey
about alter identity

Huntjens et al., 2003

21/25

University staff and

community
members
Sex, age, education

Simulate DID
Watched
documentary,
readings, created
identity
details, practiced
for 1 week

No

Yes, researcher
examined 17-item
survey about
alter identity

Huntjens, Postma,
et al., 2005

27/25

Not reported
Age, education

Simulate DID
Watched
documentary,
readings, created
identity details,
practiced for 1 week

No

Yes, researcher
examined 17-item
survey about
alter identity

Huntjens, Peters,
et al., 2005

19/25

Community
volunteers
Sex, age, education

Simulate DID
Watched
documentary,
readings, created
identity details,
practiced for
1 week

No

Yes, researcher
examined 17-item
survey about alter
identity

EEG coherence (Direct

comparisons between DID patients
and controls
were not reported, but DID patients
showed differences between
identities and simulators did not)a
Perceptual priming: proportion
correct (DID lower)a
Perceptual priming: reaction time
Conceptual priming: proportion
correct
Conceptual priming: reaction time
(DID slower)a
Word stem completion: proportion
correct (DID lower)a
Word stem completion: reaction
time (DID slower)a
Recall: overall memory (DID lower)
a

Recall: number of intrusions from

alternate personality
Recognition: hit rate
Recognition: sensitivity
Recognition: response bias (DID
had more bias)a
Memory list discrimination: hit rate
Memory list discrimination:
response bias
Remember responses to list items
Know responses to list items
Procedural memory: reaction time
to a series-learning task
(DID slower)a
Procedural memory: percent of
correct responses to a serieslearning task
Awareness of series pattern
(no test but percentages
were the same)
Cued recall
Affective ratings for words
Evaluative conditioning
Affective priming: errors
(DID had more errors)a
Affective priming: reaction times
(DID slower)a

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

19

Table 1 (continued)
Study

DID/Control N

Control source
and matching

Task and training

Blinding

Manipulation
checks

Comparisons between DID

and simulator groups

Huntjens et al., 2006

19/25

Community
members
Age, education

Simulate DID
Watched
documentary,
readings, created
identity details,
practiced for 1 week

No

Yes, researcher
examined 17-item
survey about alter
identity

Huntjens et al., 2007

19/25

Community
volunteers
Sex, age, education

Simulate DID
Watched
documentary,
readings, created
identity details,
practiced for 1 week

No

Yes, researcher
examined 17-item
survey about
alter identity

Huntjens et al., 2012

9/23

Actors with an
average of 14 years
of experience
Sex, age, education,
biographical facts

No

Kong et al., 2008

7/34

Community
members
Sex, age, education

Yes, researcher
excluded one
simulator said
during debrieng
that she had not
understood the
instructions no
further details
provided on the
procedure
Yes, interviewed
simulators about
compliance with
instructions and
simulation ability

Miller, 1989

9/9

Nursing and
secretarial staff at
mental hospital
No matching

Simulate DID such

that one personality
had memories of
childhood abuse
and the other did not
Watched documentary, readings,
created identity
details, practiced
for 1 week
Simulated DID
Given information
about DID, spent time
creating identities,
and provided
information about
the identities
Simulate DID
Watch Sybil, watch
video of DID patient

Wechsler Memory Scale-R Logical

Memory: immediate test
(DID lower)a
Wechsler Memory Scale-R Logical
Memory: delayed test
Wechsler Memory Scale-R
Visual Reproduction: immediate test
Wechsler Memory Scale-R Visual
Reproduction: delayed test
Recognition memory:
proportion correct
Plausibility of answers
(DID more plausible)a
Recall of words (DID lower)a
Memory intrusions from
separate identity
Recognition: word list
differentiation sensitivity
(DID less sensitive)a
Recognition: response bias
to old and new words
Sensitivity increases in assigning
words to the correct list
(DID increased more)a
Response bias in assigning
words to lists (DID with less bias)a
Remember responses to list items
Know responses to list items
Reaction times to words related
to current or other identity

Miller et al., 1991

20/20

Nursing,
psychological, and
secretarial staff at
mental hospital
Sex

Simulate DID
Watch Sybil, watch
video of DID patient

No

Yes, examining
ophthalmologist

No

Yes, examining
ophthalmologist

Yes,
ophthalmologist
rated group
membership

Implicit word identication task

Logical memory forced-choice
recognition test (no direct
statistical comparisons but 28% of
DID patients and 35% of controls
performed signicantly worse than
chance, a sign of malingering)
Variability between identities in
visual acuity with correction, visual
acuity without correction, visual
eld, manifest refraction cylinder,
and Maddox Wing Muscle Balance
Test, (DID more for all tests)a
Variability between identities in
manifest refraction sphere and
manifest refraction axis
Number of clinically signicant
changes between personalities
(DID more)a
Variability between identities in
visual acuity with correction and
visual elds (DID more for all tests)a
Variability between identities in
visual acuity without correction,
refraction, keratometry, and
Jaeger visual acuity
Number of clinically signicant
changes between personalities
(DID more)a
Clinically signicant differences
between identities in visual acuity
with correction, visual acuity
without correction, sphere of
manifest refraction, axis of manifest
refraction, keratometry, visual elds,
and eye muscle balance
(DID more for all tests)a
(continued on next page)

20

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

Table 1 (continued)
Study

DID/Control N

Control source
and matching

Task and training

Blinding

Manipulation
checks

Putnam et al., 1990

9/5

Not reported
No matching

Simulate DID
Specied 17
attributes about their
identity, encouraged
to rehearse alter
indentify

No

No

Reinders et al., 2012

11/18

Community
members
Sex

Simulate DID
Readings, given
instructions on
simulating neutral
and trauma-focused
identities, recalled
real emotional
incidents from past,
practiced for at least
2 weeks

No

Yes, checked to see

if instructions were
followed during
practice period,
checked to see if
emotional incident
met requirements,
evaluated ability
to simulate,
self-evaluation of
performance

Silberman et al., 1985

9/10

College students
No matching

Simulate DID
Create identity
details 1 day prior

No

No

Comparisons between DID

and simulator groups
Clinically signicant differences
between identities in Jaeger visual
acuity, cylinder of the manifest
refraction, slit lamp, fundscopy,
pupils, and pupil size
Reaction time variability
(DID more variable)a
Skin conductance
Heart rate
Respiration
Number of orientation responses
Spontaneous uctuations in
response
Habituation
[The authors report analyzing
196 variables but do not
report all results.]
Subjective sensory ratinga
Subjective emotional ratinga
Heart rate frequencya
Systolic blood pressurea
Diastolic blood pressurea
Heart rate variabilitya
[Comparisons occurred between
two diagnostic groups, two levels of
fantasy proneness among the
control groups, and two memory
scripts. Signicant interactions
occurred for all measures, but the
authors do not report specic post
hoc test results. Visual inspection of
Figures indicates that the general
pattern of results across the
variables was for trauma-focused
DID identities to be more reactive
to a trauma-focused memory script
than all other individuals and
conditions.]
Positron emission tomography
(PET)a
[The authors report results for
185 tests of various brain areas
but leave out commonalities
discovered in their whole brain
analysis. Among the areas the
authors claim to be specic to DID
in their analyses are the angular
gyrus, caudate nucleus, cerebellum,
cerebral tonsil, cingulate gyrus,
cingulate sulcus, cuneus, frontal
gyrus, fusiform gyrus, Insula,
Intra-parietal sulcus,
lateral globus pallidus, lingaul
gyrus, medial frontal gyrus,
nucleus dentatus, occipital gyrus,
occipotemporal sulcus,
orbitofrontal cortex,
parahippocapal gyrus,
parietal lobule,
perietal operculum,
postcentral gyrus,
precentral gyrus, precuneus,
putamen, rectal gyrus,
substania nigra,
and temporal gyrus.]
Recall (DID recalled fewer words
between identities, but the authors
caution that the difference was
confounded by extra word
exposure)a
Recognition
Recall compartmentalization
between personalities
Recognition compartmentalization
between personalities
Change in recall between

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

21

Table 1 (continued)
Study

Welburn et al., 2003

DID/Control N

12/10

Control source
and matching

Mental health
nurses
No matching

Task and training

Simulate DID
Read one page
summary of DID
symptoms

Blinding

Yes, interviewers

Manipulation
checks

No

Comparisons between DID

and simulator groups
dissociative and nondissociative
conditions
Change in recognition between
dissociative and nondissociative
conditions (DID showed increases
in compartmentalization of
information)a
Correlation between
compartmentalization in
dissociative and nondissociative
conditions (no direct comparison
but DID showed signicant
correlations and control did not)a
Structured Clinical Interview
for Dissociative Disorders
(DID higher)a
Somatoform Dissociation
Questionnaire-5
Dissociative Experiences Scale
Minnesota Multiphasic
Personality Inventory-2
Hypnotic eye roll sign (DID higher)a
Schneiderian First Rank
Symptoms scale
Millon Multiaxial Clinicial
Inventory-III

Note. Reported N represents the nal number of participants used in analyses. Comments in parentheses indicate the nature of differences between the DID. Comments in brackets outline
additional contextual information about the studies.
a
Signicant difference between people diagnosed with DID and people simulating DID.

3.1.4. Blinding
One of the most basic research controls is blinding researchers to the
experimental condition of participants. Similarly, it is generally important to keep participants blind to a study's hypothesis. Despite the
importance of blinding, only three studies mentioned blinding of participants or researchers (Miller, 1989; Miller et al., 1991; Welburn et al.,
2003). In all three cases an examiner or interviewer was unaware of
participants' group membership.
3.2. Differences between diagnosed and simulated DID
The next section of the review outlines replicated differences between people diagnosed with DID and simulators. Results within the
section are further broken down, as needed, into subcategories based
on if they were replicated by independent researchers or the same researcher. The section also includes a brief discussion of effect sizes for
the differences.
3.2.1. Cognitive processing decits
The most widely replicated difference between people diagnosed
with DID and simulators is the decit in cognitive processing shown
by the former group in terms of memory and reaction times. Two independent research groups have shown that people diagnosed with DID
have lower recognition of previously seen material than simulators
regardless of which identity was exposed to the material (Allen &
Movius, 2000; Huntjens et al., 2002). Measures of recognition included
forced choice and word stem completion tests. The memory decit is
not attributable to interidentity amnesia, which, as will be outlined
below, is similar across people diagnosed with DID and simulators. It
is important to note that some studies have not replicated the memory
decit (Huntjens et al., 2006, 2003; Silberman et al., 1985). There were
no obvious methodological differences to explain the failure to replicate.
Also, although the Huntjens' et al. (2006, 2003) studies did not replicate
a decit in recognition memory, they did nd a decit in recall memory,
which will be discussed below.

Independent replication also shows that people diagnosed with DID,

compared to simulators, have signicantly slower reaction times (Allen
& Movius, 2000; Huntjens et al., 2002; Huntjens, Postma, et al., 2005).
This effect seems generalizable because each of the three studies
measured reaction times to different stimuli as part of a different task.
Reaction time measures include yes or no responses to words
(Allen & Movius, 2000), completion of word stems (Huntjens et al.,
2002), and prediction of the next event in a patterned series
(Huntjens, Postma, et al., 2005). However, the nding does not always
replicate (Huntjens et al., 2002, 2012). One of the replication failures
occurred for reaction times to previously seen pictures in a study that
found reaction time differences for previously seen words (Huntjens
et al., 2002); low statistical power may explain the failure to replicate
because the group difference had a medium effect size and nearly
reached the conventional cutoff for signicant results (p = .065). The
other replication failure involved reaction times to autobiographically
meaningful information rather than the autobiographically neutral
material used in other studies; both simulators and people diagnosed
with DID were slower to respond to words associated with an alternate
identity than to identity-neutral words, which is an example of
interidentity transfer of information outlined in greater detail below
(Huntjens et al., 2012).
A nal cognitive decit involves reduced memory on measures
of recall. One research group has demonstrated that DID patients
have lower recall scores than simulators regardless of which identity was exposed to the material (Huntjens et al., 2006, 2007, 2003).
However, the effect does not occur on all tests of recall (Huntjens,
Peters, et al., 2005; Huntjens et al., 2006; Silberman et al., 1985).
Failure to replicate in one study might be explained by the fact
that its procedure utilized cued recall rather than the free recall
used in other studies (Huntjens, Peters, et al., 2005), and the interpretation of one of the other failed replications is difcult because,
as the authors point out, there was a confound in that the DID
group received extra exposure to the memory stimuli (Silberman
et al., 1985).

22

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

3.2.2. Symptom reporting

Surprisingly few studies have examined simulators' ability to mimic
DID on diagnostic measures; however, there is independent replication
of signicant differences in the reporting of symptoms. One study
showed that the Structured Clinical Interview for DSM-IV Dissociative
Disorders yielded higher symptom severity scores for people diagnosed
with DID than for simulators (Welburn et al., 2003). Another study
demonstrated higher scores for simulators on four out of 14 scales of
the Structured Interview of Reported Symptoms, a measure of faking
psychiatric symptoms (Brand et al., 2006). Although both studies documented differences in symptom reporting, the effects were in opposite
directions, the constructs being measured were different, and both studies included other measures or subscales that showed no differences
between the groups.
3.2.3. Trauma-focused vs. non trauma-focused identity performance
One potential criticism of DID simulation research is that the alternate identities created by simulators do not reect the common differences between the identities of people diagnosed with DID. For
example, one identity might have no trauma-related experiences
whereas another identity is focused on past trauma. Two research
groups have documented differences between such identities that
could not be accurately simulated. In one study, identities with and
without trauma memories completed an emotional Stroop task, which
is a measure of information processing efciency (Hermans, Nijenhuis,
Van Honk, Huntjens, & Van der Hart, 2006). Signicant differences
emerged between people diagnosed with DID and simulators, but only
for identities that had no memory of trauma. Identities without trauma
memories tended to be more efcient in processing emotional material
in people diagnosed with DID relative to simulators, but identities
with trauma memory did not show a statistically signicant difference
between groups. One other study showed DID-specic differences
between trauma-focused and neutral identities using a large number
of physiological and brain activation measures (Reinders et al., 2012).
Despite broad conceptual similarities, these studies contained considerable differences in their procedures and measures, and not all studies
nd signicant differences between identities with and without trauma
memories. Huntjens et al. (2012) found that biographical memory
transferred at similar rates between trauma-focused and non traumafocused identities for both people diagnosed with DID and simulators.
3.2.4. Eyesight variability
One researcher has conducted two studies on eyesight differences
between identities (Miller, 1989; Miller et al., 1991). The studies included eye exams conducted by an optometrist who did not have knowledge of individuals' group membership; each identity received
a separate eye exam, and variability between identities was the main
dependent measure. In both studies, people diagnosed with DID
showed greater variability between identities than simulators in terms
of visual acuity, visual eld, and the total number of clinically signicant
differences between identities. The ophthomological measures varied
in their objectivity and vulnerability to self-report bias, but signicant
differences emerged for measures with both higher and lower
objectivities (Miller et al., 1991). However, it is essential to note
that these effects have not been independently replicated, and the
lead author of the studies abandoned the research after reaching the
conclusion that the effects were not reliable (Scott D. Miller personal
communication, March 16, 2013).
3.2.5. Effect sizes
Although the purpose of the current review is not to offer a metaanalysis, effect sizes provide some perspective on the differences that
emerged in the research. Some studies did not include enough information to calculate effect sizes, but the available data were remarkably
consistent. Calculation of Cohen's d for replicated differences (n = 26)
showed that effect sizes were almost all large. In fact, the average effect

size was 1.02 (SD = 0.54). The only replicated differences with effect
sizes below 0.50 were for recognition memory (Allen & Movius,
2000), recall (Huntjens et al., 2006), and reaction time (Huntjens
et al., 2002). Effect sizes for nonreplicated differences (n = 7) were
large as well (M = 0.81, SD = 0.44). Given the small sample sizes
used in the studies, the large effect sizes are not surprising. With 20 or
fewer participants per group, the studies may have lacked statistical
power to detect smaller differences.
3.3. Similarities between diagnosed and simulated DID
The nal section of the review outlines replicated similarities
between people diagnosed with DID and simulators. There are inherent
difculties in interpreting null statistical ndings (Keppel, 1991).
Failures to nd signicant group differences can be a result of poor
statistical or methodological practices. However, in order to increase
the potential that the similarities are meaningful, the focus of this
review is on replicated ndings.
3.3.1. Interidentity transfer of information
The most prominent nding in the DID simulation literature is that
the self-reported amnesia between identities is incomplete. In contrast
to subjective reports of memory compartmentalization, there is transfer
of information between identities. Five independent research groups
have documented interidentity transfer of information, and they have
done so using a wide variety of explicit and implicit measures. Starting
with explicit measures ones that include conscious, intentional cognition such as free recall or cued recall both people diagnosed with DID
and simulators recall information presented to one identity when they
are responding as a separate, purportedly amnesic, identity (Huntjens
et al., 2007, 2003; Kong et al., 2008; Silberman et al., 1985). The same
effect holds true for recognition memory (Allen & Movius, 2000;
Huntjens et al., 2006, 2007, 2003; Silberman et al., 1985). Typically,
the procedure for these memory studies includes presenting a large
amount of separate but similar information (e.g., word lists) to the
identities and then determining how much material from one identity
is accidentally recalled or recognized by the other identity.
Implicit measures also show interidentity transfer of information
for both people diagnosed with DID and simulators. The dening feature
of implicit measures is that they do not require conscious, intentional
processing of information. For example, priming is the process whereby
exposure to a stimulus facilitates responses to a later stimulus in
some way. In a typical priming study, exposure to a negative word
(e.g., abuse) allows faster recognition of subsequent negative words
than if a positive word (e.g., love) had been seen. Four studies conducted by two independent research groups have shown that priming
between purportedly amnesic identities occurs in both people diagnosed with DID and simulators. Types of priming investigated have
included recognition of line drawings seen by an alternate identity
(Eich, Macaulay, Loewenstein, & Dihle, 1997; Huntjens et al., 2002),
responses to words seen by a alternate identity (Huntjens et al.,
2002), and responses to words that had been conditioned to have an
emotional connotation for an alternate identity (Huntjens, Peters,
et al., 2005). In term of implicit measures not related to priming, two
independent research groups have shown that reaction times are
slower for words related to an alternate identity than for neutral
words, an indication of transferred knowledge about the alternate
identity (Allen & Movius, 2000; Huntjens et al., 2012). The implicit
transfer of information occurred at similar rates for people diagnosed
with DID and simulators. Once again, it is worth reemphasizing that
the explicit and implicit transfer of information occurs at similar rates
for people diagnosed with DID and simulators.
3.3.2. Specic cognitive effects
Within the broader concept of interidentity transfer of information,
there are three specic cognitive effects that researchers have

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

independently replicated and found to be similar among people diagnosed with DID and simulators. Thus, these results overlap with those
in the previous section, but it is important to point out which specic
forms of transfer have been replicated. Three separate research groups
have documented similar transfer of information between identities
for measures of recall (Huntjens et al., 2007, 2003; Kong et al., 2008;
Silberman et al., 1985). Similarly, three research groups have documented transfer of information between identities for measures of
recognition (Allen & Movius, 2000; Huntjens et al., 2006, 2007, 2003;
Silberman et al., 1985). Two research groups each have demonstrated
slowed reaction times to information concerning an alternate identity
(Allen & Movius, 2000; Huntjens et al., 2012) and perceptual priming
between identities (Eich et al., 1997; Huntjens et al., 2002). Perceptual
priming included exposing one identity to line drawings and then testing the alternate identity's ability to recognize incomplete versions of
the same drawings. Despite the fact that both studies documented the
existence of priming, it is important to note that the pattern of results
differed between studies; Eich et al. (1997) found that people diagnosed
with DID had stronger priming, and Huntjens et al. (2002) found that
they had weaker priming.
4. Discussion
The purpose of the current review was to answer three questions
concerning research on the simulation of DID. (1) What are the methodological controls used in DID simulation research? Most researchers engage in some form of control group matching and training. In contrast,
blinding of researchers and direct checks on simulation quality are rarely utilized. (2) Are there variables that reliably differentiate between diagnosed and simulated DID? If replication is set as the standard for
reliability, there are several reliable differences between people diagnosed with DID and simulators. Consistent with the idea that DID is different from socially inuenced mimicry, simulators and people
diagnosed with DID show signicant differences on some clinical measures. Another reliable difference is that people with DID show cognitive
decits in memory and reaction time that are generalized and
not specic to interidentity amnesia. There is also evidence that people
diagnosed with DID show differences between trauma-focused and
trauma-neutral identities that are not shown by simulators. Overall,
the effect sizes of these differences were large. (3) Are there variables
that reliably fail to differentiate between diagnosed and simulated
DID? Despite the fact that amnesia is the most commonly reported
symptom of DID (Spiegel et al., 2011), people with DID and simulators
demonstrated similar levels of interidentity recall, recognition, and
priming. The pattern of results among both groups clearly indicates
that information is transferred between identities despite reports of
amnesia.
The results of this study can inform the ongoing debate about the
validity of dissociative disorders and their associated etiological
theories. Taken together, the results provide more support for
sociocognitive/skeptical models than the trauma model. Only one
replicated nding provided direct support for the trauma model and
that was the difference in cognitive processes among trauma-focused
and non trauma-focused identities. In contrast, interidentity transfer
of information, despite self-reports of amnesia, was replicated numerous times by independent researchers using a variety of methods; this
nding supports the adoption of skeptical stance toward the accuracy
of self-reported memory experiences among people diagnosed with
DID. The fact that, in terms of key symptoms of the disorder, people
taught to simulate DID are largely indistinguishable from people actually diagnosed with DID shows that social inuence cannot be eliminated
as a possible etiological factor. Furthermore, the failure to nd objective
evidence of amnesia among people diagnosed with DID is consistent
with the sociocognitive assertion that people with DID are playing a
social role rather than exhibiting a special dissociative state that results
in a split of consciousness.

23

The results also inform the broader debate about the nature of
dissociation. In an article proposing a skeptical alternative to the trauma
model of dissociation, Giesbrecht et al. (2008) concluded that dissociation, despite being associated to some subtle cognitive decits, is a
product of normal failures of cognition such as forgetting and inattention. They characterized the mental processes of people with dissociative
disorders as unremarkable, and the dearth of differences emerging in
the DID simulation literature must be considered supporting evidence
for their contention (Giesbrecht et al., 2008, p. 632). Just as Giesbrecht
and colleagues would predict, people diagnosed with DID appear to
process information less efciently than people simulating DID, but
the differences are not indicative of any specialized, dissociative memory
capacity for isolating information within one identity. However, because
people diagnosed with DID represent only a fraction of people who
dissociate (Dalenberg et al., 2012), the current review's results may not
generalize to all disorders that include dissociation.
4.1. DID and simulator group differences
Arguably, the most important potential difference between people
diagnosed with DID and simulators is in their basic symptom presentation. The average clinician has only limited resources with which to
accurately diagnose patients, and the ability to discern genuine from
intentionally faked DID using only diagnostic measures would be highly
valuable. Some evidence emerged for reliable differences between people diagnosed with DID and simulators. Control group simulators
received signicantly lower scores on the standard diagnostic tool for
DID (Welburn et al., 2003) and signicantly higher scores on a measure
of faking psychiatric symptoms (Brand et al., 2006). However, it is important to note that signicant differences emerged on a minority of
the latter measure's subscales and that the study's authors characterized
simulators as successful at feigning without being detected (p. 78).
Although these measures did differentiate between average scores
of people diagnosed with DID and simulators with relatively little
experience feigning DID, the fact that the samples were different
on basic measures of symptomology leads to questions about the
quality of the simulation, which is a point discussed in greater detail
below.
One assumption about alter identities in DID is that they have different functions than the primary identity. Differences in function between
trauma-aware and trauma-unaware identities is a subtle feature of DID
and may be difcult to fake. Indeed, the current review indicates that
test performance of trauma-aware and trauma-unaware personalities
function somewhat differently among simulators and people with DID
(Hermans et al., 2006; Reinders et al., 2012). However, the methods
used in the pertinent studies did not overlap, and not all studies nd
such identity differences (Huntjens et al., 2012). Thus, this difference
must be considered tentative until further research is conducted, but
it represents a potentially fruitful new area of DID simulation research.
The hypnosis simulation literature provides an interesting perspective on cases where simulators appear more distressed or dysfunctional
than people diagnosed with DID (Brand et al., 2006; Hermans et al.,
2006). Hypnosis simulators sometimes fake too well such that their
behavior exemplies the ideal hypnotic subject more so than individuals who are actually hypnotized (e.g., Spanos, James, & de Groot,
1990). For example, simulators may report more complete amnesia
than people who are actually under hypnosis. Although the similarity
in results is intriguing, interpretations must be tentative considering
that people diagnosed with DID exhibited many signs of cognitive
decits not shown by simulators.
Evidence for decits in recognition, recall, and reaction times
emerged across several studies, and the nding is consistent with previous reviews of the cognitive processes associated with dissociation
(Dorahy, 2001; Giesbrecht et al., 2008). There are many possible
etiological explanations for cognitive differences between people with
DID and healthy controls. Given the similar effects with PTSD

24

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

(Vasterling et al., 2002), it could be that decits are an aftereffect of the

trauma that is so frequently reported among people with DID. An intriguing possibility, also from the study of PTSD, is that the cognitive
decits may not be caused by the trauma; rather, they could represent
a preexisting vulnerability to experiencing trauma-based disorders
(Gilbertson et al., 2006). As such, the cognitive decits may have existed
before the DID. Recent research has tied dissociation with disturbances
in sleep (Koffel & Watson, 2009; van der Kloet, Merckelbach,
Merckelbach, Giesbrecht, & Lynn, 2012). Disruptions in normal sleep
patterns could explain both the dissociation and the types of memory
disturbances that characterize DID, or all three factors could share a
common cause that is yet to be discovered.
The cognitive decits associated with DID could also be unrelated to
etiology. Cognitive decits are seen among people with anxiety disorders, mood disorders, and schizophrenia (e.g., Austin, Mitchell, &
Goodwin, 2001; Castaneda, Tuulio-Henriksson, Marttunen, Suvisaari,
& Lnnqvist, 2008; Heinrichs & Zakzanis, 1998); this suggests that the
decits could be an effect of psychopathology in general rather than
any one specic disorder. DID is also known to be a disorder with high
comorbidity (Rodewald, Wilhelm-Gling, Emrich, Reddemann, &
Gast, 2011), and this means that cognitive decits may be explained
by the disorders that co-occur with DID rather than DID itself. Although
the current review cannot shed light on the origins of cognitive decits
related to DID, it is a topic worthy of further attention, especially due to
its potential etiological signicance.
An often-reported feature of DID is the existence of distinct differences between identities' biological characteristics. Several studies in
the current review included physiological measures, but the results
often failed to include clear statistical comparisons (Allen & Movius,
2000; Coons et al., 1982; Hopper et al., 2002; Putnam et al., 1990)
or were not replicated (Reinders et al., 2012). However, two studies
conducted by Miller (Miller, 1989; Miller et al., 1991) showed that
people diagnosed with DID had greater variation in eyesight between
identities than did simulators. Although some of the differences
involved subjective, self-reported measures of eyesight, others were
more objective and physical in nature. However, these effects were
not independently replicated, and vision research appears to have
ended after these publications. Miller stated that the main reason for
the absence of follow-up research was that the fad of multiple personality disorder passed (personal communication, March 16, 2013). In
addition, Miller stated that variations were not stable within personalities and much of what we were seeing was a co-construction between
helper and client (personal communication, March 16, 2013). The
experiences of one researcher should be generalized only with extreme
caution, but this candid statement from a former DID researcher illustrates much about the potential for social inuence on dissociative
phenomena.
Overall, the relative dearth of evidence for objective differences
between DID and simulated DID is surprising. Given claims for the
uniqueness of multiple personalities and the divergent interpretations
of DID symptoms that are induced in the laboratory using normal, social
processes (Gleaves, 1996; Spanos, 1994), the topic would seem to be of
high importance to the understanding of DID. However, a valid criticism
of DID simulation research is its inability to provide convincing evidence
for one etiological theory over another (Merckelbach et al., 2002). An
often-used analogy is that the ability to fake a broken leg says nothing
about the reality of broken legs or their causes. However, showing
that both people with a broken leg and those faking a broken leg can,
despite their claims to the contrary, walk without assistance is meaningful, and, as outlined below, evidence of just this type exists in the
DID simulation literature.
4.2. DID and simulator group similarities
Documentation of similarities between people diagnosed with DID
and simulators has been one of the primary outcomes of DID research,

and this is problematic from an interpretive standpoint. Basic statistical

theory states that the null hypothesis (i.e., no difference between
groups) can only be disconrmed, not conrmed (Keppel, 1991). Null
results are especially difcult to interpret in the current research
given the small sample sizes. Larger samples may have resulted in
the detection of more small- or medium-sized differences. Nonetheless,
researchers have provided strong evidence for similarity by administering validated cognitive tests, demonstrating that the tests show the
expected effects, and replicating the results. Despite statistical pitfalls,
similarities between groups are theoretically important. Simulators'
behavior is only explained by social inuence; thus, similarities
between their behavior and the behavior of people diagnosed with
DID illustrate that social inuence cannot be eliminated as a potential
causal explanation (Orne, 1971). In laboratory settings, social inuence
might include explicit instructions, interactions with the experimenters,
and the research setting. In addition, similarities provide perspective on
the objective accuracy of self-reported symptoms such as amnesia.
Among all of the similarities that emerged in this review, the
most important was the failure of interidentity amnesia among people
diagnosed with DID and simulators. Interidentity transfer of information was, in fact, the most reliable nding in the entire review. It
was supported by multiple independent replications using implicit
and explicit measures and multiple tests within those categories; thus,
the nding cannot be dismissed as an artifact of any one researcher or
methodology. An important subnding within the interidentity transfer
results is the evidence for systematic avoidance of information
presented to one identity by a purportedly amnesic second identity.
Some simulators and people diagnosed with DID showed evidence
of scores below chance on tests of recognition of material seen by
an alternate identity (Huntjens et al., 2006; Kong et al., 2008); this
is a classic sign of malingering (Merten & Merckelbach, 2013). Accurate diagnoses is necessary to determine appropriate treatment, and
combining data from cases of legitimate and malingering DID harms
the validity of research results. Thus, both clinicians and researchers
should take care to identify cases of intentionally feigned DID so
that the individuals can be effectively treated or eliminated from
research samples.
There is little agreement about how the memories of people diagnosed with DID can or should function (Merckelbach et al., 2002), but
Spiegel et al. (2011) offered this broad summary: Memory is available,
but it is not accessible (p. 835). In other words, memories are still
stored neurologically even though the person with DID cannot call
them to mind. However, the large amount of objective evidence for
interidentity transfer of information shows that people with DID can
actually access memories despite subjectively feeling like they cannot.
Thus, the current review suggests an amendment to Spiegel and
colleagues' summary: Memory is available, but it is not subjectively
accessible.
What is to be made of this failure to nd objective evidence of amnesia among people with DID? To take an extreme position, failure to
substantiate reports of amnesia constitutes evidence that one of the
dening features of DID is not strictly accurate and could be construed
as evidence against the validity of the diagnosis. By way of analogy,
imagine if people with sexual arousal disorder were actually found to
exhibit genital arousal to sexual stimuli. Just such a difference between
self-reported arousal and objective measurement of arousal lead to the
deletion of Female Sexual Arousal Disorder as a stand-alone diagnosis in
the DSM-5 (American Psychiatric Association, 2013; Graham, 2010) On
the other hand, misinterpretation is a hallmark of mental disorders. In
Panic Disorder, symptoms are perceived as signs of imminent danger;
in hypochondriasis, minor physical abnormalities are believed to be
evidence of illness; in Schizophrenia, internal thoughts are perceived
as external; in Body Dysmorphic Disorder, normal aspects of the body
are viewed as seriously awed; in DID, belief that one has amnesia is
perceived as amnesia (Giesbrecht et al., 2008). To take a less extreme
position, the absence of full amnesia among people diagnosed with

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

DID is completely consistent with what is known about cognitive

psychology, the differences between organic and functional amnesia,
and the modern view of alters as metaphorical rather than literal aspects of the person's mind (Merckelbach et al., 2002; Spiegel et al.,
2011; Staniloiu & Markowitsch, 2012). Complete amnesia would only
exist if the brains of people diagnosed with DID were damaged or if
they truly had separate minds, and neither of these characterizations
are accurate. Even supporters of the trauma model recognize that
personalities in DID are not truly separate and distinct (Dalenberg
et al., 2012).

4.3. Research methodology

The differences and similarities that emerged in this review must be
couched in the context of the methodological rigor used in producing
them, and there are several ways research methods in the DID simulation literature might be improved. One improvement would be to
increase the standards for equivalence between DID and simulation
groups. Researchers recruited people with DID and simulators from
unrelated populations and matched them on basic demographic characteristics, if at all. Such practices are entirely appropriate if the
researchers' hypotheses concern the ability of healthy subjects to
mimic DID, but the use of healthy controls as a sample of convenience
could be improved upon. One possible methodological improvement
would be to recruit simulators from treatment populations similar to
those used to recruit people diagnosed with DID. Matching could
also occur based on general distress or impairment. Not even the most
ardent skeptics argue that DID is created among individuals who are
otherwise completely healthy. In addition, people diagnosed with DID
have high comorbidity rates (Rodewald et al., 2011), and a control
group sampled from a healthy community population is likely to be
different from people diagnosed with DID in many fundamental ways
not accounted for by researchers' attempts to match basic demographics. The decits in cognitive processing identied in the current
review support this contention. Cognitive decits are common among
people with anxiety and mood disorders (e.g., Austin et al., 2001;
Castaneda et al., 2008; Vasterling et al., 2002), and recruiting simulators
from treatment populations with these diagnoses could better ensure
that differences found in simulation studies are not simply a product
of population differences unrelated to the specic diagnosis of DID.
Although there are many other possible characteristics on which to
match subjects, one more that deserves special mention is fantasy
proneness. Matching on this trait could be particularly important
because of the evidence that it is a key cognitive process in dissociation
(Giesbrecht et al., 2008).
Another methodological improvement would be to engage in rigorous evaluations of simulators' motivation and ability to mimic the
signs and symptoms of DID. Hypnosis simulation research provides
an exemplar for these procedures. Standard practice in the hypnosis
literature dictates that simulators should be well motivated and that
researchers should assess simulators' abilities to simulate (Orne,
1971). To illustrate the methodological rigor of hypnosis research
with just one example, Spanos, de Groot, Tiller, Weekes, and
Bertrand (1985) instructed simulators to behave throughout the
whole experiment as if they were an excellent hypnotic subject.
The researchers told participants that their task was to fool the hypnotist who would not know if they had been told to simulate hypnosis or actually become hypnotized. To ensure motivation the
researchers also told simulators that the hypnotist would stop the
study if their faking was detected. At the end of the study, simulators
provided detailed feedback on their experiences during the experiment using a questionnaire and in-depth interview (Experimental
Analysis Technique; Sheehan, McConkey, & Cross, 1978). This example clearly illustrates the great lengths hypnosis researchers go to
ensure valid simulations.

25

In contrast to hypnosis simulation research, few studies in the

DID simulation literature systematically assessed simulators' actual
effort and ability to feign DID. The level of motivation among DID
simulators is largely unknown. One study offered the best simulator
a cash prize to increase motivation (Brand et al., 2006), but the motivation of simulators in other studies to engage in quality simulation
is unknown. In terms of assessment of actual performance, many
studies measured factual knowledge about created identities, but
creating a biography for an identity is not the same as successfully
enacting that identity. The interviewing of participants to assess
their experience, effort, and consistency with instructions during
the study is rare in this literature (see Reinders et al., 2012 for an exception). Given the lack of data on simulators' motivation and performance, an alternative explanation for the differences that
emerged in this review is that simulators simply lacked appropriate
motivation or skills.
A higher standard for simulation is needed. Consider research on
the effective discrimination of real from simulated DID using diagnostic tools (Brand et al., 2006). Success is dened in this research
area when diagnostic tools discriminate between real and simulated
DID, but the problem posed by this approach is that people who are
detected as feigners have inherently been ineffective in their simulation. What is the minimum quality of faking that is acceptable? Using
hypnosis research as an example again, the standard for performance in simulators is to actually fool the hypnotist into believing
that they have been hypnotized (Orne, 1971). Perhaps an effective,
albeit extreme, manipulation check would be the ability of simulators to receive a clinical diagnosis of DID. Such a suggestion comes
with many practical difculties increased sample sizes, more simulator training, and extensive clinical interviewing but the resultant methodological rigor would substantially increase the validity
of simulation results, which currently suffer from an almost total absence of manipulation checks.
A potential difculty with a manipulation check consisting of
clinical diagnosis is that researchers would have to remain blind to
participants' group membership; however, blinding is one of the
methodological improvements most needed in this research area.
There is a long and rich history demonstrating the power of expectations among both experimenters and participants to inuence research results (Orne, 1962; Schulz & Grimes, 2002). DID simulation
research is no less vulnerable to these validity concerns. Nonetheless, only three studies in this literature included blinding of researchers to diagnostic group (Miller, 1989; Miller et al., 1991;
Welburn et al., 2003). A methodological improvement would be to
keep all researchers who interact with participants during data collection blind to diagnostic group. Some might react to this suggestion by insisting that such a procedure would never work because
of simulators' inability to fool the experimenters. However, this concern goes against what is known from the current review and hypnosis research. Hypnosis experimenters believed that they could not be
fooled by people simulating hypnosis at least they did until they
were consistently fooled (Orne, 1971). Keeping researchers blind to
experimental group would also allow for an interesting new analysis: comparison of the people researchers believe are diagnosed
DID to the people they believe are simulating DID. In addition to
keeping experimenters blind to diagnostic group, participants
should be blind to the specic hypotheses of the research in order
to prevent them from, intentionally or unintentionally, adjusting
their responses based on the expected results.
Some additional methodological observations are worth noting. The
DID samples contained only individuals who could switch personalities
on demand. Although this is a methodological necessity for this
research, exclusive use of participants with that characteristic could
lead to problems in generalizing the results to individuals who are
unaware of alternate identities and cannot switch on demand.
Researchers performed a huge number of statistical comparisons we

26

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

would estimate the number at around 1000 in these 20 studies on a

host of measures to yield a relatively small number of differences. A
potential problem is the tendency to conduct a large number of statistical comparisons without specic a priori predictions. For example,
Putnam et al. (1990) analyzed 196 variables per participant in their
study with no guiding hypotheses. Differences found in exploratory
research can be valid and important, but researchers should exert
appropriate caution when interpreting a small number of signicant
differences that result from a large number of tests. The sample sizes
of some studies may also be leading to statistical error. Mean sample
size per group in this literature was below 20 participants, and this
may be insufcient for some statistical tests. In terms of Type 2
error, such small sample sizes may hide real differences due to low
power. The fact that effect sizes of differences were almost universally large lends suggests that DID simulation studies are systematically
underpowered.
Small samples may also lead to Type 1 error by exacerbating the
effects of random or extreme scores. Twenty participants per group
is the accepted minimum sample size for avoiding false positives in
statistical results (Simmons, Nelson, & Simonsohn, 2011), but 75%
of the studies in this literature analyzed DID samples with fewer
than 20 people. Obtaining large sample sizes is an endemic problem
in DID research (Boysen, 2011; Boysen & VanBergen, 2013), but it
would be poor science to ignore statistical standards simply due to
the difculty of obtaining samples in this research area. The within
subject designs used by many DID simulation researchers increases
statistical power, but the fact remains that some researchers conduct
a large number of tests on a small number of people. For example,
Reinders et al. (2012) utilized samples of 11 people diagnosed with
DID and 18 simulating controls; they further broke down the
control group based high and low fantasy proneness. These
methods resulted in statistical comparisons being conducted on
groups with ns of 11, 5, and 8. Using these small samples,
Reinders et al. (2012) report statistical results for over 200 specic comparisons between simulators and people diagnosed
DID. However, they compared neural activation between groups
across the whole brain and explicitly stated that areas of common
activation were left out of the article. Thus, the reported results
are likely to be a gross underestimate of the total number of statistical
comparisons conducted by the researchers. Although the researchers can
be commended for their use of a priori predictions, the small sample
sizes, large number of statistical tests, and selective reporting of results
are troubling both in this particular study and in the broader research
area.

4.4. Limitations
Several limitations of this review are worth noting. The most important limitation is the exclusion of research not published in English. DID research occurs in many countries, and there are likely to
be ndings published in non-English journals that we did not review.
Another potential problem is the lack of a comparison group for the
results. Research on the simulation of mental disorders appears to
be unique to DID, and this makes it impossible to know if some of
the trends that emerged are also unique. For example, we cannot determine if the matching and blinding that occurs in DID research is
more or less rigorous than is typical. Similarly, we do not know if
simulation of other disorders would also lead to few group differences and contradiction of self-reported symptoms. Finally, research
in this review was dominated by the many studies of Huntjens and
colleagues from the Netherlands. Although their research meets
the highest standards of cognitive experimentation, the fact that
their studies represent one third of the total DID simulation research
output means that they had an especially large inuence on the results of this review.

5. Conclusion
Some have dismissed research of DID simulation by healthy controls
as a logical dead end in terms of its ability to solve theoretical disputes
(Gleaves, 1996), but this review suggests that simulation has a great
deal to offer. The extant differences between DID and simulators offer
means for improving the quality of differential diagnosis and understanding basic phenomena associated with the disorder. Differences
showed that DID may not be as simple as enacting social role. In contrast, overlap among DID and simulators illustrated how self-reports
may not be a reliable indication of actual abilities and symptoms if
patients and clinicians can be fooled by the subjective experience of
amnesia, what else could be an illusion? Despite some meaningful
results, conclusions from this research are hampered by methodological aws. Future researchers should consider (a) remaining
blind to diagnostic group, (b) matching simulation and DID groups
more comprehensively, (c) increasing the standards for effective
simulation, and (d) increasing statistical power. With methodological improvements such as these, continued research on the simulation of DID is likely to produce further theoretically meaningful
results.

References
Allen, J. J. B., & Movius, H. L. (2000). The objective assessment of amnesia in Dissociative Identity Disorder using event-related potentials. International Journal of Psychophysiology, 38,
2141. http://dx.doi.org/10.1016/S0167-8760(00)00128-8.
American Psychiatric Association (2013). Diagnostic and statistical manual of mental
disorders (5th ed.)Washington, DC: Author.
Austin, M. P., Mitchell, P., & Goodwin, G. M. (2001). Cognitive decits in depression:
Possible implications for functional neuropathology. The British Journal of Psychiatry,
178, 200206. http://dx.doi.org/10.1192/bjp.178.3.200.
Boysen, G. A. (2011). The scientic status of childhood dissociative identity disorder: A
review of published research. Psychotherapy and Psychosomatics, 80, 329334.
http://dx.doi.org/10.1159/000323403.
Boysen, G. A., & VanBergen, A. (2013). A review of published research on adult dissociative identity disorder: 20002010. Journal of Nervous and Mental Disorders, 201, 511.
http://dx.doi.org/10.1097/NMD.0b013e31827aaf81.
Brand, B.L., McNary, S. W., Loewenstein, R. J., Kolos, A.C., & Barr, S. R. (2006). Assessment of genuine and simulated Dissociative Identity Disorder on the structured
interview of reported symptoms. Journal of Trauma & Dissociation, 7, 6385.
http://dx.doi.org/10.1300/J229v07n01_06.
Castaneda, A. E., Tuulio-Henriksson, A., Marttunen, M., Suvisaari, J., & Lnnqvist, J. (2008).
A review on cognitive impairments in depressive and anxiety disorders with a focus
on young adults. Journal of Affective Disorders, 106, 127. http://dx.doi.org/10.
1016/j.jad.2007.06.006.
Coons, P.M., & Milstein, V. (1994). Factitious or malingered multiple personality disorder:
Eleven cases. Dissociation: Progress in the Dissociative Disorders, 7, 8185.
Coons, P.M., Milstein, V., & Marley, C. (1982). EEG studies of two multiple personalities
and a control. Archives of General Psychiatry, 39, 823825. http://dx.doi.org/
10.1001/archpsyc.1982.04290070055010.
Dalenberg, C. J., Brand, B.L., Gleaves, D. H., Dorahy, M. J., Loewenstein, R. J., Cardea, E.,
et al. (2012). Evaluation of the evidence for the trauma and fantasy models of dissociation. Psychological Bulletin, 138, 550588. http://dx.doi.org/10.1037/a0027447.
Dell, P. F. (2006). A new model of dissociative identity disorder. Psychiatric Clinics of North
America, 29, 126. http://dx.doi.org/10.1016/j.psc.2005.10.013.
Dorahy, M. J. (2001). Dissociative Identity Disorder and memory dysfunction: The current
state of experimental research and its future directions. Clinical Psychology Review, 21,
771795. http://dx.doi.org/10.1016/S0272-7358(00)00068-4.
Draijer, N., & Boon, S. (1999). The limitations of dissociative identity disorder: Patients at
risk, therapists at risk. Journal of Psychiatry & Law, 27, 423458.
Eich, E., Macaulay, D., Loewenstein, R. J., & Dihle, P. H. (1997). Implicit memory,
interpersonality amnesia, and dissociative identity disorder: Comparing patients
with simulators. In J.D. Read, & D. S. Lindsay (Eds.), Recollections of trauma: Scientic
evidence and clinical practice (pp. 469474). New York: Plenum Press.
Foote, B., Smolin, Y., Kaplan, M., Legatt, M. E., & Lipschitz, D. (2006). Prevalence of
dissociative disorders in psychiatric outpatients. The American Journal of Psychiatry,
163, 623629. http://dx.doi.org/10.1176/appi.ajp.163.4.623.
Friedl, M. C., & Draijer, N. (2000). Dissociative disorders in Dutch psychiatric inpatients.
The American Journal of Psychiatry, 157, 10121013. http://dx.doi.org/10.1176/
appi.ajp.157.6.1012.
Giesbrecht, T., Lynn, S. J., Lilienfeld, S. O., & Merckelbach, H. (2008). Cognitive processes in
dissociation: An analysis of core theoretical assumptions. Psychological Bulletin, 134,
617647. http://dx.doi.org/10.1037/0033-2909.134.5.617.
Gilbertson, M. W., Paulus, L. A., Williston, S. K., Gurvits, T. V., Lasko, N.B., Pitman, R. K., et al.
(2006). Neurocognitive function in monozygotic twins discordant for combat
exposure: Relationship to posttraumatic stress disorder. Journal of Abnormal
Psychology, 115, 484495. http://dx.doi.org/10.1037/0021-843X.115.3.484.

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

Gleaves, D. H. (1996). The sociocognitive model of dissociative identity disorder: A reexamination of the evidence. Psychological Bulletin, 120, 4259. http://dx.doi.org/
10.1037/0033-2909.120.1.42.
Graham, C. A. (2010). The DSM diagnostic criteria for female sexual arousal disorder.
Archives of Sexual Behavior, 39, 240255. http://dx.doi.org/10.1007/s10508009-9535-1.
Green, J. P., Page, R. A., Handley, G. W., & Rasekhy, R. (2005). The 'hidden observer' and
ideomotor responding: A real-simulator comparison. Contemporary Hypnosis, 22,
123137. http://dx.doi.org/10.1002/ch.8.
Heinrichs, R., & Zakzanis, K. K. (1998). Neurocognitive decit in schizophrenia: A quantitative review of the evidence. Neuropsychology, 12, 426445. http://dx.doi.org/
10.1037/0894-4105.12.3.426.
Hermans, E. J., Nijenhuis, E. R. S., Van Honk, J., Huntjens, R. J. C., & Van der Hart, O. (2006).
Identity state-dependent attentional bias for facial threat in dissociative identity
disorder. Psychiatry Research, 141, 233236. http://dx.doi.org/10.1016/j.psychres.
2005.03.007.
Hopper, A., Ciorciari, J., Johnson, G., Spensley, J., Sergejew, A., & Stough, C. (2002). EEG
coherence and dissociative identity disorder: Comparing EEG coherence in DID
hosts, alters, controls and acted alters. Journal of Trauma & Dissociation, 3, 7588.
http://dx.doi.org/10.1300/J229v03n01_06.
Huntjens, R. J. C., Peters, M. L., Postma, A., Woertman, L., Effting, M., & Van der Hart, O.
(2005). Transfer of newly acquired stimulus valence between identities in Dissociative Identity Disorder (DID). Behaviour Research and Therapy, 43, 243255.
http://dx.doi.org/10.1016/j.brat.2004.01.007.
Huntjens, R. J. C., Peters, M. L., Woertman, L., Bovenschen, L. M., Martin, R. C., & Postma, A.
(2006). Inter-identity amnesia in dissociative identity disorder: A simulated memory
impairment? Psychological Medicine, 36, 857863. http://dx.doi.org/10.1017/
S0033291706007100.
Huntjens, R. J. C., Peters, M. L., Woertman, L., Van der Hart, O., & Postma, A. (2007). Memory transfer for emotionally valenced words between identities in dissociative identity disorder. Behaviour Research and Therapy, 45, 775789. http://dx.doi.org/
10.1016/j.brat.2006.07.001.
Huntjens, R. J. C., Postma, A., Hamaker, E. L., Woertman, L., Van Der Hart, O., & Peters, M.
(2002). Perceptual and conceptual priming in patients with dissociative identity disorder. Memory & Cognition, 30, 10331043. http://dx.doi.org/10.3758/BF03194321.
Huntjens, R. J. C., Postma, A., Peters, M. L., Woertman, L., & Van der Hart, O. (2003).
Interidentity amnesia for neutral, episodic information in dissociative identity
disorder. Journal of Abnormal Psychology, 112, 290297. http://dx.doi.org/10.1037/
0021-843X.112.2.290.
Huntjens, R. J. C., Postma, A., Woertman, L., Van der Hart, O., & Peters, M. L. (2005).
Procedural memory in dissociative identity disorder: When can inter-identity
amnesia be truly established? Consciousness and Cognition: An International
Journal, 14, 377389.
Huntjens, R. J. C., Verschuere, B., & McNally, R. J. (2012). Inter-identity autobiographical amnesia in patients with dissociative identity disorder. PLoS ONE, 7,
18. http://dx.doi.org/10.1371/journal.pone.0040580.
Keppel, G. (1991). Design and analysis: A researcher's handbook. Prentice-Hall, Inc.
Kirsch, I., & Lynn, S. (1998). Dissociation theories of hypnosis. Psychological Bulletin, 123,
100115. http://dx.doi.org/10.1037/0033-2909.123.1.100.
Kisiel, C. L., & Lyons, J. S. (2001). Dissociation as a mediator of psychopathology among
sexually abused children and adolescents. The American Journal of Psychiatry, 158,
10341039. http://dx.doi.org/10.1176/appi.ajp.158.7.1034.
Kluft, R. P. (1987). The simulation and dissimulation of multiple personality disorder.
American Journal of Clinical Hypnosis, 30, 104118. http://dx.doi.org/10.1080/
00029157.1987.10404170.
Koffel, E., & Watson, D. (2009). Unusual sleep experiences, dissociation, and schizotypy:
Evidence for a common domain. Clinical Psychology Review, 29, 548559.
http://dx.doi.org/10.1016/j.cpr.2009.06.004.
Kong, L. L., Allen, J. J. B., & Glisky, E. L. (2008). Interidentity memory transfer in dissociative
identity disorder. Journal of Abnormal Psychology, 117, 686692. http://dx.doi.org/
10.1037/0021-843X.117.3.686.
Lewis, D., Yeager, C. A., Swica, Y., Pincus, J. H., & Lewis, M. (1997). Objective
documentation of child abuse and dissociation in 12 murderers with
dissociative identity disorder. The American Journal of Psychiatry, 154,
17031710.
Lilienfeld, S. O., Kirsch, I., Sarbin, T. R., Lynn, S. J., Chaves, J. F., Ganaway, G. K., et al. (1999).
Dissociative Identity Disorder and the sociocognitive model: Recalling the lessons
of the past. Psychological Bulletin, 125, 507523. http://dx.doi.org/10.1037/
0033-2909.125.5.507.
Merckelbach, H., Devilly, G. J., & Rassin, E. (2002). Alters in Dissociative Identity
Disorder metaphors or genuine entities? Clinical Psychology Review, 22, 481498.
http://dx.doi.org/10.1016/S0272-7358(01)00115-5.
Merten, T., & Merckelbach, H. (2013). Symptom validity testing in somatoform and
dissociative disorders: A critical review. Psychological Injury and Law, 6, 122137.
http://dx.doi.org/10.1007/s12207-013-9155-x.
Miller, S. D. (1989). Optical differences in cases of multiple personality disorder. Journal
of Nervous and Mental Disease, 177, 480486. http://dx.doi.org/10.1097/
00005053-198908000-00005.
Miller, S. D., Blackburn, T., Scholes, G., White, G. L., White, G. L., & Mamalis, N. (1991).
Optical differences in multiple personality disorder: A second look. Journal of
Nervous and Mental Disease, 179, 132135. http://dx.doi.org/10.1097/00005053199103000-00003.
Naish, P. N. (2005). On the inevitability of nding hypnosis-simulator equivalence.
Contemporary Hypnosis, 22, 154157. http://dx.doi.org/10.1002/ch.5.
Orne, M. T. (1959). The nature of hypnosis: Artifact and essence. The Journal of Abnormal
and Social Psychology, 5, 277299. http://dx.doi.org/10.1037/h0046128.

27

Orne, M. T. (1962). On the social psychology of the psychological experiment: With

particular reference to demand characteristics and their implications. American
Psychologist, 17, 776783.
Orne, M. T. (1971). The simulation of hypnosis: Why, how, and what it means. International
Journal of Clinical and Experimental Hypnosis, 19, 183210. http://dx.doi.org/
10.1080/00207147108407167.
Paris, J. (2012). The rise and fall of dissociative identity disorder. The Journal of
Nervous and Mental Disease, 200, 10761079. http://dx.doi.org/10.1097/NMD.
0b013e318275d285.
Piper, A., & Merskey, H. (2004). The persistence of folly: A critical examination of Dissociative Identity Disorder. Part I. The excesses of an improbable concept. The Canadian
Journal if Psychiatry, 49, 592600.
Putnam, F. W. (1989). Diagnosis and treatment of multiple personality disorder. New York:
Guilford.
Pope, H. G., Barry, S., Bodkin, A., & Hudson, J. I. (2006). Tracking scientic interest in
the dissociative disorders: A study of scientic publication output 19842003.
Psychotherapy and Psychosomatics, 75, 1924. http://dx.doi.org/10.1159/
000089223.
Putnam, F. W., Zahn, T. P., & Post, R. M. (1990). Differential autonomic nervous system
activity in multiple personality disorder. Psychiatry Research, 31, 251260.
http://dx.doi.org/10.1016/0165-1781(90)90094-L.
Reinders, A. A. T. S. (2008). Cross-examining dissociative identity disorder: Neuroimaging
and etiology on trial. Neurocase, 14, 4453. http://dx.doi.org/10.1080/
13554790801992768.
Reinders, A. A. T. S., Nijenhuis, E. R., Paans, A.M., Korf, J., Willemsen, A. T. M., & Den
Boer, J. A. (2003). One brain, two selves. NeuroImage, 20, 21192125.
Reinders, A. A. T. S., Nijenhuis, E. S., Quak, J., Korf, J., Haaksma, J., Paans, A. J., et al. (2006).
Psychobiological characteristics of Dissociative Identity Disorder: A symptom
provocation study. Biological Psychiatry, 60, 730740. http://dx.doi.org/10.1016/
j.biopsych.2005.12.019.
Reinders, A. A. T. S., Willemsen, A. T. M., Vos, H. P. J., Den Boer, J. A., & Nijenhuis, E. R. S.
(2012). Fact or factitious? A psychobiological study of authentic and simulated
dissociative identity states. PLoS ONE, 7, 117. http://dx.doi.org/10.1371/
journal.pone.0039279.
Rodewald, F., Wilhelm-Gling, C., Emrich, H. M., Reddemann, L., & Gast, U. (2011). Axis-I
comorbidity in female patients with Dissociative Identity Disorder and Dissociative
Identity Disorder not otherwise specied. Journal of Nervous and Mental Disease,
199, 122131. http://dx.doi.org/10.1097/NMD.0b013e318208314e.
Sar, V., Koyuncu, A., Ozturk, E., Yargic, L., Kundakci, T., Yazici, A., et al. (2007). Dissociative
disorders in the psychiatric emergency ward. General Hospital Psychiatry, 29, 4550.
http://dx.doi.org/10.1016/j.genhosppsych.2006.10.009.
Sar, V., & Ozturk, E. (2007). Functional dissociation of the self: A sociocognitive approach to trauma and dissociation. Journal of Trauma & Dissociation, 8, 6989.
http://dx.doi.org/10.1300/J229v08n04_05.
Sar, V., Unal, S. N., Kiziltan, E., Kundakci, T., & Ozturk, E. (2001). HMPAO SPECT study of
regional cerebral blood ow in dissociative identity disorder. Journal of Trauma &
Dissociation, 22, 525. http://dx.doi.org/10.1300/J229v02n02_02.
Schulz, K. F., & Grimes, D. A. (2002). Blinding in randomised trials: Hiding who got what.
Lancet, 359, 696700.
Sheehan, P. W., McConkey, K. M., & Cross, D. (1978). Experiential analysis of hypnosis:
Some new observations on hypnotic phenomena. Journal of Abnormal Psychology,
87, 570573. http://dx.doi.org/10.1037/0021-843X.87.5.570.
Silberman, E. K., Putnam, F. W., Weingartner, H., Braun, B. G., & Post, R. M. (1985).
Dissociative states in multiple personality disorder: A quantitative study.
Psychiatry Research, 15, 253260. http://dx.doi.org/10.1016/0165-1781(85)
90062-9.
Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive psychology:
Undisclosed exibility in data collection and analysis allows presenting anything as
signicant. Psychological Science, 22, 13591366. http://dx.doi.org/10.1177/
0956797611417632.
Spanos, N.P. (1994). Multiple identity enactments and multiple personality disorder: A
sociocognitive perspective. Psychological Bulletin, 116, 143165. http://dx.doi.org/
10.1037/0033-2909.116.1.143.
Spanos, N.P., de Groot, H. P., Tiller, D. K., Weekes, J. R., & Bertrand, L. D. (1985).
Trance logic duality and hidden observer responding in hypnotic, imagination
control, and simulating subjects: A social psychological analysis. Journal
of Abnormal Psychology, 94, 611623. http://dx.doi.org/10.1037/0021843X.94.4.611.
Spanos, N.P., James, B., & de Groot, H. P. (1990). Detection of simulated hypnotic amnesia. Journal of Abnormal Psychology, 99, 179182. http://dx.doi.org/10.1037/
0021-843X.99.2.179.
Spanos, N.P., Weekes, J. R., & Bertrand, L. D. (1985). Multiple personality: A
social psychological perspective. Journal of Abnormal Psychology, 94, 362376.
http://dx.doi.org/10.1037/0021-843X.94.3.362.
Spiegel, D., Loewenstein, R. J., Lewis-Fernandez, R., Sar, V., Simeon, D., Vermetten, E.,
et al. (2011). Dissociative disorders in DSM-5. Depression and Anxiety, 28,
824852. http://dx.doi.org/10.1002/da.20874.
Staniloiu, A., & Markowitsch, H. J. (2012). Towards solving the riddle of forgetting in functional amnesia: recent advances and current opinions. Frontiers in Psychology, 3, 123.
http://dx.doi.org/10.3389/fpsyg.2012.00403.
Tamar-Gurol, D., Sar, V., Karadag, F., Evren, C., & Karagoz, M. (2008). Childhood emotional
abuse, dissociation, and suicidality among patients with drug dependency in Turkey.
Psychiatry and Clinical Neurosciences, 62, 540547. http://dx.doi.org/10.1111/
j.1440-1819.2008.01847.x.
van der Kloet, D., Merckelbach, H., Merckelbach, H., Giesbrecht, T., & Lynn, S. J. (2012).
Fragmented sleep, fragmented mind: The role of sleep in dissociative symptoms.

28

G.A. Boysen, A. VanBergen / Clinical Psychology Review 34 (2014) 1428

Perspectives on Psychological Science, 7, 159175. http://dx.doi.org/10.1177/

1745691612437597.
Vasterling, J. J., Duke, L. M., Brailey, K., Constans, J. I., Allain, A. N., & Sutker, P. B. (2002).
Attention, learning, and memory performances and intellectual resources in
Vietnam veterans: PTSD and no disorder comparisons. Neuropsychology, 16, 514.
http://dx.doi.org/10.1037/0894-4105.16.1.5.
Vermetten, E., Schmahl, C., Lindner, S., Loewenstein, R. J., & Bremner, J. (2006). Hippocampal and amygdalar volumes in dissociative identity disorder. The American Journal of
Psychiatry, 163, 630636. http://dx.doi.org/10.1176/appi.ajp.163.4.630.

Welburn, K. R., Fraser, G. A., Jordan, S. A., Cameron, C., Webb, L. M., & Raine, D.
(2003). Discriminating Dissociative Identity Disorder from Schizophrenia and
feigned dissociation on psychological tests and structured interview. Journal
of Trauma & Dissociation, 4, 109. http://dx.doi.org/10.1300/J229v04n02_07.
World Health Organization (1992). International statistical classication of diseases and related health problems (10th ed., revised ) (Geneva).
Xiao, Z., Yan, H., Wang, Z., Zou, Z., Xu, Y., Chen, J., et al. (2006). Trauma and dissociation in
China. The American Journal of Psychiatry, 163, 13881391. http://dx.doi.org/
10.1176/appi.ajp.163.8.138.