Essay Grade

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Training and Education in Professional Psychology
2008, Vol. 2, No. 1, 26 –34
Copyright 2008 by the American Psychological Association
1931-3918/08/$12.00 DOI: 10.1037/1931-3918.2.1.26
An Exploratory Study of Vicarious Trauma Among Therapist Trainees
Shelah A. Adams
Shelley A. Riggs
Texas Woman’s University
University of North Texas
The current study explored vicarious trauma among therapist trainees in relation to history of trauma,
experience level, trauma-specific training, and defense style. Students in graduate clinical and counseling
psychology training programs (N ⫽ 129) completed the Trauma Symptom Inventory, Defense Style
Questionnaire, and an experience questionnaire. Results indicated trauma symptoms were significantly
associated with defense style, which appeared to moderate personal trauma history and experience level.
Trauma-specific training was also independently related to trauma symptoms. Notably, over half the
sample reported a self-sacrificing defense style, which was a risk factor for vicarious trauma. Training
implications of the findings are discussed.
Keywords: vicarious trauma, defense style, trauma history, therapist, trainee
current study specifically examined the association of vicarious
trauma symptoms to personal history of trauma, applied experience
level, trauma-specific training, and further explored the potential
interaction of these variables with defense style.
In the last decade, the mental health field has devoted increasing
attention to the potentially harmful impact of working closely with
traumatized individuals. A growing body of literature explores the
theoretical and clinical implications of trauma work and empirical
evidence is beginning to emerge documenting its negative psychological effects among help-givers, including disaster relief workers, police and medical personnel, and mental health professionals
(Follette, Polusney, & Milbeck, 1994; Sloan, Rozensky, Kaplan, &
Saunders, 1994; Weiss, Marmar, Metzler, & Ronfeldt, 1995). Recent
studies examining possible risk factors for vicarious trauma among
clinicians have reported associations of secondary trauma symptoms
with personal history of trauma and experience level (Chrestman,
1999; Pearlman & Mac Ian, 1995). Pearlman and Saakvitne (1995)
also suggested that the lack of formal trauma coursework and maladaptive defense styles among therapists may create a vulnerability to
trauma-related symptomatology. Although many practitioners’ first
applied experience with trauma clients occurs during practicum training in graduate school and evidence suggests that novice therapists are
more likely to experience difficulties (Pearlman & Mac Ian, 1995),
empirical investigations of vicarious trauma among graduate students
in clinical and counseling psychology programs are lacking. The
Vicarious Traumatization
Various terms have been used to describe the stress resulting
from helping a traumatized person, including “secondary traumatic
stress,” “compassion fatigue,” (Figley, 1995, 1999) and “vicarious
traumatization/trauma” (Pearlman & Saakvitne, 1995; Schauben &
Frazier, 1995). Wilson and Lindy (1994) described this experience
as a form of posttraumatic stress disorder (PTSD) among therapists, who without direct exposure to a traumatic event will display
symptoms almost identical to those of PTSD. Herman (1997)
referred to this response as traumatic countertransference, from
which the therapist experiences the same terror, rage, and anguish
as the patient, albeit to a lesser degree.
Following Pearlman and Saakvitne (McCann & Pearlman, 1990;
Pearlman & Saakvitne, 1995; Saakvitne, 1996; Saakvitne & Pearlman, 1996), the current study conceptualizes vicarious trauma as a
process involving a transformation in the inner experience of the
therapist resulting from empathic engagement with clients’ traumatic material. Whereas countertransference is present in all therapeutic relationships with dynamics unique to each therapist-client
dyad, vicarious traumatization is a cumulative consequence not
specific to any one client, which can be lasting and linked to
multiple aspects of the therapist’s personal and professional life
(Saakvitne, 1996). Taxing psychological effects can interfere with
the therapist’s adaptive assumptions of personal security and a
meaningful world (Chrestman, 1999; Kassam-Adams, 1999), increasing the likelihood of a protective numbing reaction to feelings
of pain and loss (Saakvitne, 1996).
Posttraumatic symptoms typically assessed by researchers to determine the presence of vicarious traumatization include suspiciousness, anxiety, depression/sadness, somatic symptoms, intrusive
thoughts and feelings, avoidance, emotional numbing and flooding,
and increased feelings of personal vulnerability (Neumann & Gamble,
1995; Pearlman & Mac Ian, 1995; Steed & Downing, 1998). How-
SHELAH A. ADAMS, MA, is a doctoral student in Counseling Psychology
at Texas Woman’s University. She earned her masters of arts degree in
Counseling Psychology from Texas Woman’s University and is currently
a doctoral candidate in the same program. Research interests include
childhood trauma, resulting psychopathology, and disordered eating behaviors.
SHELLEY A. RIGGS, PhD, is currently an assistant professor in the
department of psychology at the University of North Texas. She earned her
doctorate degree in Counseling Psychology from the University of Texas at
Austin. She is currently an associate professor at the University of North
Texas. Research interests include the role of close relationships (family,
romantic, therapeutic and supervisory alliance) and trauma/loss in the
development of psychopathology.
CORRESPONDENCE CONCERNING THIS ARTICLE should be addressed to
Shelley A. Riggs, PhD, P. O. Box 311280, Department of Psychology,
University of North Texas, Denton, TX 76203-1280. E-mail: riggs@
unt.edu
26
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
VICARIOUS TRAUMA AMONG TRAINEES
ever, the effects of vicarious traumatization vary greatly and may
depend on personal factors and training or experience level (Pearlman
& Saakvitne, 1995; Way VanDeusen, Martin, Applegate, & Jandle,
2004). For example, the theoretical literature suggests that therapists
with a personal history of trauma may be more vulnerable to the
impact of working with trauma survivors than therapists without such
history (Figley, 1999; Pearlman & Saakvitne, 1995; Ryan, 1999).
However, empirical findings regarding this relationship are mixed
(Pearlman & Mac Ian, 1995; Schauben & Frazier, 1995; Way et al.,
2004). Consequently, more research is needed to clarify the contribution of personal trauma history and also identify other factors that
potentially interact with personal trauma to produce vicarious trauma
symptoms.
In addition, the extent of applied work with trauma clients and
trauma-specific didactic education may be related to vicarious
traumatization. Several studies using samples of practicing therapists have found that a shorter length of time providing trauma
treatment is associated with more difficulty related to trauma work,
including increases in avoidance, dissociation, anxiety, intrusions,
and other trauma symptoms (Chrestman, 1999; Pearlman & Mac
Ian, 1995; Way et al., 2004). When new therapists encounter
symptoms of vicarious traumatization, they may experience anxiety, shame, and a sense of incompetence, and consequently not
seek adequate supervision and support (Neumann & Gamble,
1995; Pearlman & Mac Ian, 1995). If vicarious traumatization is
left unattended and unresolved, there is the risk that the therapist
may become emotionally distant and unable to maintain a warm,
empathic, and responsive stance to clients (McCann & Pearlman,
1990), which may eventually result in burnout and subsequent
departure from the field (Pearlman & Saakvitne, 1995). Given
earlier research indicating significantly greater stress responses
among practicum and internship trainees compared to more experienced professional staff (Rodolfa, Kraft, & Reilley, 1988), it is
especially important to extend the investigation of vicarious
trauma to graduate students in applied psychology programs.
Pearlman and Saakvitne (1995) argued that without formal traumaspecific training, the trauma therapist is vulnerable to confusion and
potentially harmed by the work. Although there is growing consensus
supporting the need for some degree of training and supervision in
trauma work as part of the graduate curricula (Campbell, Raja, &
Grining, 1999; Figley, 1995), many professional therapists indicate
that their academic training did not provide them with the necessary
skills for working with trauma survivors (Alpert & Paulson, 1990;
Pope & Feldman-Summers, 1992). Whereas students often receive
minimal information regarding trauma therapy in the context of various academic courses or training seminars, fewer students receive
extensive training regarding trauma in a semester long course or
intensive multiday workshop. With no published studies on this issue
to date, the current study explores the question of whether the amount
of formal trauma-specific didactic training is associated with vicarious
trauma symptoms and interacts with therapist characteristics.
Defense Style
Since Freud (1940/1964) proposed that the ego unconsciously
adopts habitual psychological defenses to shield itself against
anxiety, the construct of defense mechanism has been widely
accepted in the psychological community, as well as the larger
mainstream society. The Diagnostic and Statistical Manual, 4th
27
edition, Test-Revision (DSM–IV–TR; American Psychiatric Association [APA], 2000) defines defense mechanisms (or coping
styles) as “automatic psychological processes that protect the individual against anxiety and from the awareness of internal or external
dangers or stressors. Individuals are often unaware of these processes
as they operate” (p. 807). The theoretical literature (Cramer, 1998;
Vaillant, 1971) distinguishes between immature defense mechanisms
that emerge early in development and are more unconscious (e.g.,
splitting, acting out, isolation) and mature coping strategies that
emerge later in development and operate more consciously (e.g.,
sublimation, rationalization, humor). According to Bond (1995), the
developmental progression of defense styles is characterized by a
hierarchical shift from basic anxiety regarding controlling raw impulses, to an intermediate level with an overriding focus on the
importance of others, to a more adaptive style showing less preoccupation with others and more creative expression of the inner self. The
number of levels proposed in the maturational hierarchy of defenses
varies from three (Andrews, Singh, & Bond, 1993), to four (Bond,
Garner, Christian, & Sigal, 1983; Vaillant, 1975, 1976), to seven
developmental levels (APA, 2000).
Regardless of how many defense levels are measured, there seems
to be consensus that maladaptive defensive reactions to stressors are
characterized by the tendency to either distort or restrict conscious
perception of a painful reality (Punamaki, Kanninen, Qouta, & ElSarraj, 2002; Silverstein, 1996). The weight of empirical evidence
suggests that immature/maladaptive defenses are associated with poor
psychological functioning, including personality disorders, depression, anxiety disorders, and posttraumatic symptoms (Bond, 2004;
Punamaki et al., 2002; Silverstein, 1996). Because defense mechanisms serve a protective function in helping to maintain psychological
integrity in the face of threat, Punamaki et al. (2002) argued that
defenses are instrumental in determining ongoing adjustment to
trauma and should be treated as moderating variables when
examining the associations between traumatic stress and psychological distress.
Pearlman and Saakvitne (1995) hypothesized a relationship between vicarious traumatization and psychological defenses. Although research examining specific defense styles in relation to
vicarious traumatization is lacking, recent studies have reported
that therapists with healthy coping styles characterized by active,
problem-focused strategies reported fewer PTSD symptoms, less
vicarious traumatization, less negative affect, fewer disruptions in
self-trust schemas, and less burnout than those with avoidant or
emotion-focused coping styles (Schauben & Frazier, 1995; Weaks,
2000). It is also important to examine more unconscious defense
mechanisms that may influence these broad coping strategies
because therapists who are uncomfortable in the presence of powerful emotions, or whose affect tolerance is exceeded, will draw
upon familiar, protective defenses. Herman (1997) suggested that
restrictive defenses (e.g., denial, dissociation, numbing) and overinvolved, intrusive caregiving are commonly employed defense
mechanisms used by therapists in response to vicarious traumatization. Restrictive defenses may result in minimization or avoidance of traumatic material and distancing from the client. Conversely, overinvolved therapists may act in an impulsive, intrusive
manner, characterized by rescue attempts, boundary violations, or
controlling behavior. Consequently, supervisors need to be aware
of these defensive tendencies among trainees in order to assist
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ADAMS AND RIGGS
28
them in processing reactions to clients and foster the development
of more adaptive coping strategies.
Exploratory Study
The current study is a preliminary exploration of the correlates
of vicarious trauma among graduate student trainees, which may
differ from factors previously identified in samples of working
mental health practitioners. The study had two primary objectives:
(a) to examine the relationship between vicarious traumatization
among trainees and variables recognized as potentially influential
in this process among practicing therapists (i.e., history of trauma,
clinical experience, trauma-specific training) and (b) to explore the
relationship between defense style and vicarious traumatization
symptoms, as well as its possible interaction with the previous three
factors in relation to reported symptoms. Based on the existing theoretical and empirical literature, we expected that higher levels of
trauma symptoms would be significantly related to a personal history
of trauma, less applied experience with trauma survivors, and minimal
or no trauma-specific training. Also, compared to other defense styles,
we predicted that the adaptive defense style would be related to
significantly lower levels of trauma symptoms and that defense style
would interact with personal trauma, experience level, and traumaspecific training.
Method
Participants
Participants were recruited from the APA-accredited clinical
and counseling psychology graduate programs at state universities
in Texas. Out of 355 packets distributed to graduate students, 134
packets were returned yielding a response rate of 37.7%. However,
five packets were excluded because the student currently was not
enrolled in the program or was not working with trauma clients,
thus reducing the response rate to 36.3%. Of the 129 participants
included in the study, 83.7% were female and 85.3% were Caucasian. Participants ranged in age from 22 to 55 years, with a mean
of 31.21 (SD ⫽ 8.69), median of 28, and mode of 26. Doctoral
level counseling (43.4%; n ⫽ 56) and clinical (25.6%; n ⫽ 33)
students comprised the majority of the sample, with masters students making up the remaining 31% (counseling n ⫽ 39; clinical
n ⫽ 1).
Instrumentation
Background information. An experience questionnaire was
developed for this study to gather demographic information. In
addition, participants were asked to indicate whether they had
substantial (e.g., multiple workshops, semester-long course, other
extensive formal training), minimal (e.g., one workshop or seminar), or no trauma-specific training. In order to assess for previous
trauma work experience, the questionnaire inquired about previous
number of semesters spent working with trauma clients in the
capacity of a therapist in a practicum, internship, paid employment, or volunteer experience. Finally, a history of personal
trauma was assessed by asking whether or not the student therapist
had ever personally been (a) involved in a natural disaster; (b) a
witness or participant in combat; (c) a victim of a violent crime; (d)
a victim of physical, sexual, or emotional abuse as a child; (e) an
adult victim of sexual assault or rape; (f) involved in a physically
abusive relationship; or (g) a witness to someone being seriously
injured or killed.
Vicarious traumatization. The Trauma Symptom Inventory
(TSI; Briere, 1995; Briere, Elliot, Harris, & Cotman, 1995) is a
widely used self-report instrument consisting of 100 items describing trauma symptoms, which are rated on a 4-point scale of
frequency of occurrence over the past 6 months. Because TSI
subscales that are broad and not definitive of trauma (i.e., depression, anger/irritability, tension reduction) or primarily relate to
sexual trauma (i.e., sexual concerns, dysfunctional sexual behavior) are less relevant to vicarious trauma, five subscales totaling 42
items were selected to represent vicarious trauma: (a) Anxious
Arousal measures posttraumatic hyperarousal symptoms like
jumpiness and tension; (b) Intrusive Experiences measures reexperiencing symptoms such as flashbacks and nightmares; (c) Defensive Avoidance measures both cognitive and behavioral avoidance strategies, (d) Dissociation measures dissociative
experiences, including depersonalization, derealization, numbing;
and (e) Impaired Self-Reference measures self-concept problems,
such as identity confusion and low self-esteem. Across four validation studies, Briere (1995) reported relatively high internal consistency for these subscales, with alpha reliability coefficients
ranging from .82–.87 for anxious arousal, .87–.90 for intrusive
experience, .87–.90 for defensive avoidance, .82–.88 for dissociation, and .85–.88 for impaired self-reference. The TSI has also
demonstrated good convergent and predictive validity in clinical
and nonclinical samples (Briere et al., 1995; Briere & Elliott,
1997).
Defense style. Derived from a list of defense mechanisms currently being considered for inclusion as a separate axis in the DSM
nosology (Bond, 1995), the Defense Style Questionnaire (DSQ;
Bond et al., 1983; Bond & Wesley, 1996) is the most widely used
self report of defense mechanisms (Bond, 2004). The original
instrument consists of 88 items on a 9-point Likert scale. Based on
factor analysis, the DSQ identifies a hierarchy of four basic defense styles: (a) Maladaptive action style consists of the most
immature defense mechanisms that reflect an inability to manage
impulses by taking constructive action, for example, withdrawal,
acting out, regression, inhibition, passive aggression, and projection; (b) Image-distorting style consists of defenses involving
splitting the image of self and other into good and bad and includes
derivatives of omnipotence, splitting, and primitive idealization;
(c) Self-sacrificing style reflects a need to maintain an image of the
self as kind, helpful, and never angry, which is accomplished
through consists of reaction formation and pseudoaltruism; and (d)
Adaptive style represents positive coping strategies and consists of
the most mature defense mechanisms such as suppression, sublimation, and humor.
Following the manual (Bond & Wesley, 1996), the DSQ scale
scores were used to group participants in the current study into
these four prototypical defense styles. The DSQ reliably identifies
defense styles, which correspond to hypothesized patterns of unconscious psychological mechanisms (Andrews et al., 1993). The
developmental hierarchy of styles from least mature to most mature is supported by theory (Semrad, Grinspoon, & Feinberg, 1973;
Vaillant, Bond, & Vaillant, 1986) and evidence of expected correlations with ego strength and ego development scores (Bond,
1995). Internal consistency of the DSQ was demonstrated through
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
VICARIOUS TRAUMA AMONG TRAINEES
significant item-total correlations ( p ⬍ .001) and factor analyses
showing four theoretically meaningful clusters. The DSQ has
demonstrated test–retest reliability, construct and criterion validity, and the ability to distinguish patients from nonpatients (Andrews et al., 1993; Bond et al., 1983, 1989; Perry & Cooper, 1989;
Sammallahti, Aalberg, & Pentinsaari, 1994). Andrews et al. reported comparable findings regarding internal reliability and test–
retest reliability for long and short versions of the DSQ when a
three factor solution was used to identify immature, neurotic (⬃
self-sacrificing), and mature defense styles. Internal consistency
on the long and short versions, respectively, were .89 and .80 on
the immature factor, .72 and .58 on the neurotic factor, and .59 and
.68 on the mature factor.
Procedures
To explore graduate training in Texas, the public state universities with doctoral psychology programs in clinical or counseling
psychology were identified; if the university also offered masters
programs, these were included in recruitment. Program directors of
11 graduate programs at five universities were contacted. The
directors of nine programs (PhD ⫽ 7; MA ⫽ 2) at four universities, who agreed to distribute questionnaire packets to students
enrolled in practicum or internship in fall 2002, were mailed
packets containing a recruitment letter with a brief description of
the study, consent forms, an experience questionnaire, the Trauma
Symptom Inventory, the Defense Style Questionnaire, and a
stamped envelope to return the completed forms. The number of
packets sent to each program was based on the number of enrolled
students in each program. The manner in which the packets were
distributed was negotiated with each program director. A
follow-up call was placed to each program director 2 weeks after
the packets were sent to ensure their distribution.
Results
Descriptive Data and Analyses
Over a third of the sample (38.7%; n ⫽ 50) reported a history of
personal trauma. The majority of the sample (74.3%) reported
some formal didactic training in trauma work, with 38.7% (n ⫽
50) participants reporting minimal training and 35.6% (n ⫽ 46)
participants reporting substantial training. However, a full quarter
of the sample (n ⫽ 33) reported working with trauma clients with
no prior training related to trauma. Number of previous semesters
experience with trauma work ranged from 0 to 38 (M ⫽ 4.86,
SD ⫽ 6.09); however, the median number of semesters was 3, and
the mode was 2, with 40% (n ⫽ 52) of the sample reporting two
or fewer semesters working with trauma clients.
TSI scale means were not clinically significant, with only
8 –15% of the sample exceeding the clinical cut-off score on each
scale. However, taken together, 31% of the sample exceeded the
clinical cut-off score on one or more TSI scale. Low cell counts
necessitated collapsing maladaptive (n ⫽ 3) and image-distorting
(n ⫽ 6) DSQ categories into one style, which yielded 7.0%
maladaptive/image-distorting (n ⫽ 9), 51.2% self-sacrificing (n ⫽
66), and 41.8% adaptive (n ⫽ 54). This decision is supported by
the theory, which distinguishes immature from mature defenses by
the tendency to either distort or restrict conscious perception of
29
painful reality (Cramer, 1998; Punamaki et al., 2002; Silverstein,
1996; Vaillant, 1971). Moreover, these three defense styles correspond to the 3-factor solution identified by Andrews et al. (1993)
for the DSQ-40, a brief version of the DSQ that identifies mature
(e.g., humor, sublimation), intermediate/neurotic (e.g., pseudoaltruism, reaction formation), and immature defenses. The latter
category combines maladaptive and image-distorting items (e.g.,
acting out, passive-aggressiveness, splitting, projection). Due to
the low number of maladaptive/image-distorting participants, analyses proceeded on a strictly exploratory basis for this category.
Preliminary analyses were conducted to determine whether study
variables were systematically related to demographic variables. With
few exceptions, age, gender, ethnicity, and type of program (counseling PhD, clinical PhD, counseling/clinical MA) were not associated
with TSI scores. Although an overall multivariate analysis of variance
(MANOVA) was nonsignificant for TSI scales, univariate results
indicated that ethnicity was significantly associated with TSI intrusive
experience, F(1, 127) ⫽ 5.43, p ⫽ .02, with ethnic minority participants reporting lower levels of intrusive experience. Because of the
prevalence of female participants, we also examined whether gender
was associated with each of the independent variables but found no
gender differences in defense style, history of personal trauma, applied experience, or formal trauma-specific training. Based on these
findings, MANOVAs were conducted without controlling for demographic variables, but follow-up univariate tests for TSI intrusive
experience used ethnicity as a covariate.
Hypothesis Testing
Three multifactorial MANOVAs, which control for the possibility of inflated overall Type I error rate and also incorporate
correlations among variables into the test statistic (Stevens,
1996), were conducted to test the association of five trauma
symptoms with the three defense styles and either personal
trauma, applied experience, or trauma-specific training. Wilks’s
lambda F and an alpha value of .05 were used to determine
significance of MANOVAs. To better account for Type I error
in follow-up univariate tests, the Bonferonni correction indicated that a .01 level was needed to determine significance.
Table 1 shows the results of the three MANOVAs and follow-up
analysis of variances (ANOVAs). The first MANOVA with defense style and personal trauma (yes/no) yielded a significant main
effect for defense style (F ⫽ 3.40, df ⫽ 10, 236, p ⬍ .000), a
significant interaction term (F ⫽ 2.86, df ⫽ 10, 236, p ⫽ .002), but
a nonsignificant effect for personal trauma (F ⫽ 2.21, df ⫽ 5, 118,
p ⫽ .058). Follow-up ANOVAs for defense style were significant
for all trauma scales: anxious arousal, F ⫽ 5.19, df ⫽ 2, p ⫽ .007,
intrusive experiences, F ⫽ 5.92, df ⫽ 2, p ⫽ .004, defensive
avoidance, F ⫽ 7.82, df ⫽ 2, p ⫽ .001, dissociation, F ⫽ 8.88,
df ⫽ 2, p ⫽ .001, and impaired self-reference, F ⫽ 8.40, df ⫽ 2,
p ⬍ .000. Post hoc comparisons indicated that self-sacrificing style
was associated with significantly higher scores than adaptive style
on all five trauma scales. Maladaptive/image-distorting style was
also associated with significantly higher scores than adaptive style
on impaired self-reference, and higher scores than both adaptive
and self-sacrificing styles on dissociation. The interaction of defense style and personal trauma was significant for intrusive experience, F ⫽ 7.01, df ⫽ 2, p ⫽ .001 and defensive avoidance, F ⫽
6.65, df ⫽ 2, p ⫽ .002, and also approached significance for
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ADAMS AND RIGGS
30
Table 1
Factorial MANOVA for Defense Style and Personal Trauma, Applied Experience, and Formal Training (N ⫽ 128)
TSI scales
MANOVA
ANOVAs
Anxious
arousal
Intrusive
experience
Defensive
avoidance
Dissociation
Impaired self
reference
Defense style
F
Personal
trauma
F
Defense ⫻
trauma
interaction
F
3.40*** (.126)
2.21 (.085)
2.86** (.108)
5.19
**
*
Defense style
F
Applied
experience
F
Defense ⫻
experience
interaction
F
Defense style
F
Training
F
Defense ⫻
training
interaction
F
2.06* (.080)
3.37** (.124)
2.07* (.080)
2.10* (.083)
3.58*** (.134)
1.17 (.059)
*
*
(.078)
.006 (.000)
4.04 (.062)
3.93 (.060)
.45 (.004)
1.42 (.023)
3.32 (.057)
1.93 (.031)
.96 (.028)
5.92** (.089)
4.75* (.038)
7.01*** (.104)
2.48 (.039)
.23 (.002)
.62 (.010)
5.25** (.081)
2.01 (.033)
.21 (.007)
7.82
***
(.114)
8.88*** (.127)
8.40*** (.121)
**
9.13
**
(.070)
6.65 (.098)
.02 (.000)
.88 (.007)
1.32 (.021)
1.32 (.021)
**
4.39
**
(.067)
.001 (.000)
1.16 (.019)
6.07 (.092)
.36 (.006)
.52 (.017)
5.91** (.088)
4.15** (.063)
5.80* (.045)
8.61** (.065)
4.81** (.073)
1.76 (.028)
4.35** (.068)
4.75** (.073)
3.44* (.054)
4.01* (.063)
1.67 (.053)
1.75 (.055)
Note. Ethnicity was used as a covariate in ANCOVAs for Intrusive Experience.
p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001. Numbers in parentheses ⫽ partial ␩2. Effect sizes for partial ␩2 are .01 ⬃ small, .06 ⬃ medium, .14 ⬃ large.
*
anxious arousal, F ⫽ 4.04, df ⫽ 2, p ⫽ .020. Whereas participants
with an adaptive defense style did not appear to differ based on
personal trauma, participants with a self-sacrificing defense style
and personal trauma reported higher levels of symptoms than
self-sacrificing participants without personal trauma. In contrast,
participants with a maladaptive/image-distorting style and personal trauma reported lower levels of symptoms than maladaptive/
image-distorting participants without personal trauma.
The second MANOVA yielded significant main effects for defense
style (F ⫽ 2.07, df ⫽ 10, 238, p ⫽ .028) and applied experience (F ⫽
3.37, df ⫽ 5, 119, p ⫽ .007), and a significant interaction (F ⫽ 2.07,
df ⫽ 10, 238, p ⫽ .028). Follow-up univariate tests with defense style
were similar to the first MANOVA for defensive avoidance, dissociation, and impaired self-reference, but did not meet the .01 level for
anxious arousal, F ⫽ 3.93, df ⫽ 2, p ⫽ .022, and intrusive experience,
F ⫽ 2.48, df ⫽ 2, p ⫽ .088. Follow-up ANOVAs for applied
experience were significant at the .01 level only for impaired selfreference, F ⫽ 8.61, df ⫽ 1, p ⫽ .004 and approached significance for
dissociation, F ⫽ 5.80, df ⫽ 1, p ⫽ .017. In each case, participants
with two or fewer semesters of applied experience reported higher
levels of trauma symptoms. The interaction of defense style and
applied experience was significant only for dissociation, F ⫽ 4.81,
df ⫽ 2, p ⫽ .010. Although no substantial differences were noted for
self-sacrificing or adaptive defense styles, participants with a
maladaptive/image-distorting style and two or fewer semesters of
applied experience reported significantly higher levels of dissociation
compared to maladaptive/image-distorting participants with three or
more semesters of applied experience.
The final MANOVA yielded significant main effects for defense style (F ⫽ 2.10, df ⫽ 10, 232, p ⬍ .000) and formal
trauma-specific training (F ⫽ 3.58, df ⫽ 10, 232, p ⬍ .000), but a
nonsignificant interaction (F ⫽ 1.47, df ⫽ 20, 386, p ⫽ .088).
Results of follow-up univariate tests for defense style were similar
to the previous MANOVAs but did not reach the .01 level for
anxious arousal, F ⫽ 3.62, df ⫽ 2, p ⫽ .03, and dissociation, F ⫽
4.35, df ⫽ 2, p ⫽ .015. Despite a highly significant multivariate
main effect for formal trauma training, follow-up univariate tests
showed only trends ( p ⬍ .05) for dissociation, F ⫽ 3.44, df ⫽ 2,
p ⫽ .035 and impaired self-reference, F ⫽ 4.01, df ⫽ 2, p ⫽ .021.
Based on medium effect sizes, exploratory post hoc comparisons
were conducted and indicated that participants with substantial
trauma-specific training reported lower levels of dissociation than
those with minimal training, and also lower levels of impaired
self-reference than those with both minimal and no training.
Implications
Although current findings must be considered preliminary
and require replication with a larger, more representative sample, overall results of this exploratory study generally support
theoretical predictions regarding vicarious trauma and have
important implications for the training and supervisio