CNL-545 Topic 4: Safety Planning and Harm Reduction Planning
ittakesavillage
Intimate_Partner_Violence_Risk.pdf
Violence and Victims, Volume 23, Number 2, 2008
202 © 2008 Springer Publishing Company DOI: 10.1891/0886-6708.23.2.202
Intimate Partner Violence Risk Assessment and Management
P. Randall Kropp, PhD British Columbia Forensic Psychiatric Services Commission
While risk assessment is important in the management of intimate partner violence perpetrators, the science and practice of risk assessment in this field are still in early development. This article reviews the literature on intimate partner violence risk assess- ment. The original intent was to direct discussion to assist the Military Family Advocacy Program (FAP), U.S. Department of Defense, to develop guidelines for the treatment of domestic violence offenders. The article is divided into sections as follows: (a) Defining Risk; (b) The Risk Factors; (c) Models of Risk Assessment; (d) Existing Risk Instruments; (e) The Role of the Victim in Risk Assessment; (f ) Qualifications to Conduct Assessments; (g) Communicating Risk; and (h) Managing Risk. Relevant issues and controversies are raised throughout the article.
Keywords: domestic violence; risk assessment; risk management; issues
T hose working with perpetrators and victims of intimate partner violence regularly must deal with dangerous situations. Violence risk assessment is a method for manag- ing those situations. Although risk assessments have been performed in the domestic
violence field either formally or informally for decades, the practice of risk assessment has received relatively little attention in scientific and professional literature. However, this field is evolving rapidly, and there appears to be growing interest in the development of risk assess- ment technology (Bennett Cattaneo, 2007; Dutton & Kropp, 2000; Hilton & Harris, 2005). With this growth in technology emerge a number of important scientific, professional, and ethical issues that should be considered. This article reviews the literature on intimate partner violence risk assessment in an effort to highlight some of these issues. The original intent of the article was to direct discussion to help the Military Family Advocacy Program (FAP; U.S. Department of Defense) develop guidelines for the treatment of domestic violence offenders, but the discussion is relevant to the domestic violence field in general. The article is divided into the following sections: (a) Defining Risk; (b) The Risk Factors; (c) Models of Risk Assessment; (d) Existing Risk Instruments; (e) The Role of the Victim in Risk Assessment; (f ) Qualifications to Conduct Assessments; (g) Communicating Risk; and (h) Managing Risk.
DEFINING RISK
Although this article is about risk, there is little consensus in the field about what is meant by the term. Most studies on intimate partner violence risk and recidivism appear to define risk in terms of the likelihood that some form of violence will take place sometime in the
Risk Assessment 203
future (Dutton, Bodnarchuk, Kropp, Hart, & Ogloff, 1997; Hanson & Wallace-Capretta, 2000; Rosenfeld, 1992). In practice, however, decisions about risk likely involve con- sideration of the imminence, nature (e.g., emotional, physical, sexual), frequency, and seriousness of the violence, in addition to the likelihood that it will occur (Hart, 2001; Mulvey & Lidz, 1995). Thus, risk is a complex phenomenon; judgments must consider the who, what, where, when, and how of violence. For example, an offender could be at risk for imminent, relatively minor, physical violence against his spouse, such as pushing or shoving, but not at risk for long-term, frequent, sexual violence. These are two rather different scenarios, and they present different implications for victim safety planning, criminal justice intervention, and treatment of the offender. Front-line professionals work- ing with offenders and victims often must consider these nuances. However, research on risk assessment and recidivism rarely has made these distinctions.
These issues have been touched on in the spousal violence literature. Campbell’s (1995) Danger Assessment asks potential victims to document in a calendar the severity and fre- quency of their partners’ abusive behaviors during the past year. Investigators have also discussed the importance of distinguishing among risk for spousal homicide, “severe” vio- lence, and less serious forms of violence (Aldridge & Brown, 2003; Campbell et al., 2003; Dobash, Dobash, & Cavanagh 2007; Weisz, Tolman, & Saunders, 2000; Wilson & Daly, 1993), focusing on factors such as stalking, sexual proprietariness, past use of guns, and estrangement as potential lethality factors. Others have commented on the dynamic nature of risk, noting that risk for violence is influenced by context and by the changing nature of time-varying risk factors (Jones & Gondolf, 2001; Kropp, Hart, Webster, & Eaves, 1999). However, researchers and clinicians alike have yet to agree on an operational definition of risk. The lack of consensus creates several problems. First, it makes it difficult to com- pare risk assessment studies. Second, it makes it difficult to investigate whether different aspects of risk have correspondingly different constellations of risk indicators. Finally, even if the risk factors are similar for the various forms of spousal violence, it could be that the relative importance or weightings of risk factors might vary, but this remains unclear. Researchers can better inform stakeholders and bridge the gap between science and prac- tice by dissecting the definition of risk and exploring the interactions between risk factors and “types” of risk (e.g., Sjöstedt & Grann, 2003).
THE RISK FACTORS
In recent years a number of comprehensive literature reviews have been published on risk factors for domestic violence (Bennett Cattaneo & Goodman, 2005; Dutton & Kropp, 2000; Hilton & Harris, 2005; Riggs, Caulfield, & Street, 2000; Schumacher, Feldbau-Kohn, Slep, & Heyman, 2001; Vest, Catlin, Chen, & Brownson, 2002) and intimate partner homicide (Aldridge & Browne, 2003; Campbell, Sharps, & Glass, 2001). These reviews reflect a burgeoning literature in the past 15 years that has seen hundreds of studies touching on risk issues, and there now appears to be considerable consensus regarding the relevant risk fac- tors. It is impossible to provide a comprehensive review of this literature in this short article, and this has been done effectively elsewhere. To summarize the key risk factors, however, Table 1 describes 10 risk categories as they appear in the Brief Spousal Assault Form for the Assessment of Risk (B-SAFER; Kropp, Hart, & Belfrage, 2005 ), a risk management tool recently developed for law enforcement and other criminal justice agencies (see below). The table includes a short list of selected references for each category.
T A
B L
E 1
. S
p ou
sa l
V io
le n
ce R
is k
C at
eg or
ie s
C at
eg or
y D
ef in
it io
n S
el ec
te d
R ef
er en
ce s
V io
le nt
A ct
s A
ct ua
l or
a tt
em pt
ed p
hy si
ca l
ha rm
, in
cl ud
in g
se xu
al v
io le
nc e
an d
us e
of
w ea
po ns
.
C am
pb el
l et
a l.
, 20
01 ;
G on
do lf
, 19
88 ;
H il
to n
et a
l. ,
20 04
; R
ig gs
e t
al .,
20 00
; S
ch um
ac he
r et
a l.
, 20
01 ;
W il
li am
s &
H
ou gh
to n,
2 00
4. V
io le
nt T
hr ea
ts o
r T
ho ug
ht s
T hr
ea te
ni ng
s ta
te m
en ts
o r
in ti
m id
at in
g be
ha vi
or ,
in cl
ud in
g st
al ki
ng .
B ur
ge ss
e t
al .,
19 97
; H
ar t,
1 99
2; M
cF ar
la ne
e t
al .,
20 02
; P
al ar
ea ,
Z on
a, L
an e,
& L
an gh
in ri
ch se
n- R
oh li
ng ,
19 99
; S
on ki
n, 1
98 7.
E sc
al at
io n
In cr
ea se
o ve
r ti
m e
in t
he s
ev er
it y,
f re
qu en
cy ,
or d
iv er
si ty
o f
vi ol
en ce
. C
am pb
el l
et a
l. ,
20 03
; G
re en
la nd
, 19
85 ;
S on
ki n,
1 98
7;
S tu
ar t
& C
am pb
el l,
1 98
9; W
ei sz
e t
al .,
20 00
. V
io la
ti on
o f
C ou
rt O
rd er
s B
re ac
h of
c on
di ti
on s
of b
ai l,
p ro
ba ti
on ,
pa ro
le ,
re st
ra in
in g
or de
rs ,
an d
so f
or th
. A
nd re
w s
& B
on ta
, 19
96 ,
20 03
; C
ar ls
on ,
H ar
ri s,
& H
ol de
n,
19 99
; H
ar re
ll &
S m
it h,
1 99
6; H
ar t,
K ro
pp ,
& H
ar e,
19
88 ;
S he
rm an
, S
m it
h, S
ch m
id t,
& R
og an
, 19
92 .
V io
le nt
A tt
it ud
es B
el ie
fs a
nd v
al ue
s th
at e
nc ou
ra ge
o r
ex cu
se
ab us
iv e
an d
vi ol
en t
be ha
vi or
. D
ut to
n, 1
99 5;
G on
do lf
& W
hi te
, 20
01 ;
R ig
gs e
t al
., 20
00 ;
S au
nd er
s, 1
99 2;
S ch
um ac
he r
et a
l. ,
20 01
. G
en er
al C
ri m
in al
it y
A nt
is oc
ia l
be ha
vi or
t ha
t is
p er
si st
en t,
fr
eq ue
nt ,
or d
iv er
se .
C ad
sk y
& C
ra w
fo rd
, 19
88 ;
H an
so n
& W
al la
ce -C
ap re
tt a,
20
00 ;
H il
to n,
H ar
ri s,
& R
ic e,
2 00
1; T
w ee
d &
D ut
to n,
19
98 .
In ti
m at
e R
el at
io ns
hi p
P ro
bl em
s P
ro bl
em s
su ch
a s
se pa
ra ti
on f
ro m
p ar
tn er
or
e xt
re m
e co
nf li
ct r
eg ar
di ng
r el
at io
ns hi
p st
at us
.
C am
pb el
l et
a l.
, 20
01 ;
D al
y &
W il
so n,
1 99
8; K
yr ia
co u
et a
l. ,
19 99
; R
ig gs
e t
al .,
20 00
; S
ch um
ac he
r et
a l.
, 20
01 .
E m
pl oy
m en
t P
ro bl
em s
P ro
bl em
s su
ch a
s ch
ro ni
c un
em pl
oy m
en t
or
fi na
nc ia
l di
ff ic
ul ti
es .
C am
pb el
l et
a l.
, 20
03 ;
C ar
ls on
e t
al .,
19 99
; H
ot al
in g
&
S ug
ar m
an ,
19 86
; K
yr ia
co u
et a
l. ,
19 99
; S
he rm
an ,
S m
it h,
S
ch m
id t,
& R
og an
, 19
92 .
S ub
st an
ce U
se P
ro bl
em s
S ub
st an
ce u
se t
ha t
ca us
es i
m pa
ir m
en t
of
he al
th o
r so
ci al
f un
ct io
ni ng
. D
ut to
n &
H ar
t, 1
99 2;
G on
do lf
, 20
01 ;
H il
to n
& H
ar ri
s,
20 05
; Jo
ne s
& G
on do
lf ,
20 01
; R
ig gs
e t
al .,
20 00
; T
ol m
an &
B en
ne tt
, 19
90 .
M en
ta l
H ea
lt h
P ro
bl em
s P
ro bl
em s
su ch
a s
de lu
si on
s, a
ng er
, im
pu ls
iv it
y, o
r su
ic id
al it
y. C
am pb
el l
et a
l. ,
20 01
; D
ut to
n, 1
99 5;
G on
do lf
, 19
98 ;
K es
sl er
, M
ol na
r, F
eu re
r, &
A pp
el ba
um ,
20 01
; M
ag do
l et
a l.
, 19
97 ;
S ch
um ac
he r
et a
l. ,
20 00
.
N o te
. T
hi s
m at
er ia
l is
r ep
ro du
ce d
fr om
t he
B ri
ef S
po us
al A
ss au
lt F
or m
f or
t he
E va
lu at
io n
of R
is k
(B -S
A F
E R
) w
it h
th e
pe rm
is si
on o
f P
ro ac
ti ve
R es
ol ut
io ns
I nc
., V
an co
uv er
, B
C ,
C an
ad a
(w w
w .p
ro ac
ti ve
-r es
ol ut
io ns
.c om
).
Risk Assessment 205
There is some debate about whether it is acceptable to consider risk factors that have solid theoretical or intuitive appeal but limited empirical support (Douglas, Cox, & Webster, 1999; Hanson & Morton-Bourgon, 2004). Some have argued that only risk factors that have demonstrated reliable empirical association with recidivistic violence should be considered (Hilton, Harris, Rice, Lang, & Cormier, 2004; Quinsey, Harris, Rice, & Cormier, 1998). However, scientific methods might be inadequate to reliably capture rare or difficult to measure risk factors, such as homicidal/suicidal ideation or patriarchal attitudes. An orthodox empirical approach precludes practitioners from considering these risk factors despite the common sense that they might apply in certain cases. Imagine, for example, the professional and ethical issues involved in ignoring suicidal ideation as a risk factor in an offender who has just separated from his intimate partner. Thus a deli- cate balance must be achieved. Scientific rigor is important, as some professionals might have erroneous assumptions about what is and what is not related to violence (Douglas & Kropp, 2002). However, there should also be some license to consider risk factors justi- fied by sound theory and professional consensus, especially given the limitations of social science methodology.
MODELS OF RISK ASSESSMENT
Three models, or methods, of violence risk assessment have been discussed in the lit- erature: unstructured clinical decision making, actuarial decision making, and structured professional judgment.
Unstructured Clinical Decision Making
Unstructured clinical decision making is probably still the most widely used approach to spousal violence risk assessment (Campbell et al., 2001; Dutton & Kropp, 2000). This is a method that involves no constraints or guidelines for the evaluator. Decisions are based on the exercise of professional discretion and usually are justified according to the qualifications and experience of the professional who makes them. Thus, professionals must trust their intuition or “ gut ” when determining who is or who is not dangerous. The approach has been widely criticized in the violence literature for lacking reliability, validity, and accountability (Litwack & Schlesinger, 1999; Quinsey et al., 1998) and has been labeled “ informal, subjective, [and] impressionistic ” (Grove & Meehl, 1996, p. 293). One traditional advantage of unstructured clinical decision making is that it allows for an idiographic analysis of the offender ’ s behavior and a person- and context-specific tailoring of risk management and violence prevention strategies. However, because the approach relies so heavily on professional discretion, it is vulnerable to missing impor- tant factors that require intervention. Recommendations for management strategies—if they are made at all—might be based more on the training, preferences, and biases of the evaluator rather than on (a) well-reasoned consideration of dynamic and criminogenic (i.e., crime-relevant) risk factors, and (b) intervention strategies that are either empiri- cally valid or well-accepted in the field. Given the widespread criticism of this approach, those working with spousal assaulters and their victims are moving away from this prac- tice (Campbell, 1995; Dutton & Kropp, 2000; Hilton & Harris, 2005). At the very least, practitioners should only consider risk factors that have some support in the empirical or clinical literature.
206 Kropp
Actuarial Decision Making
The actuarial method of risk assessment is strongly associated with the prediction para- digm popular in the violence literature (see Heilbrun, 1997). Such methods are designed to predict specific behaviors within a specific time frame. The stated goal of the actuarial method is to predict violence in (a) a relative sense, by comparing an individual to a norm- based reference group; and, (b) an absolute sense, by providing a precise, probabilistic esti- mate of the likelihood of future violence. Grove and Meehl (1996, p. 293) have described this approach as “ mechanical and algorithmic. ” The key strength to this approach is that it improves upon the poor reliability and validity of unstructured clinical assessments (Grove & Meehl, 1996; Litwack, 2001; Quinsey et al., 1998). The actuarial approach can help the evaluator estimate, in a relative sense, the risk posed by an individual over a fixed time period compared to a reference group. In this sense, it is a worthwhile endeavor to develop and test actuarial instruments for spousal violence risk assessments. Indeed, sev- eral attempts have shown correlations between the actuarial approach—that is, the totaling of risk factors to produce a risk “ score ” —and various measures of violent behavior and construct validity (Campbell, 1995; Grann & Wedin, 2002; Hanson & Wallace-Capretta, 2000; Hilton et al., 2004; Kropp & Hart, 2000; McFarlane, Campbell, & Watson, 2002). The Ontario Domestic Assault Risk Assessment, or ODARA (Hilton et al., 2004), is an example of the actuarial approach.
Actuarial approaches have been criticized for their lack of practical utility (Douglas & Kropp, 2002; Hart, 1998; Litwack, 2001). Thus, there is an unresolved schism between science and practice. Practitioners resist using methods that eliminate professional discre- tion. This might be because they see their role as preventing violence rather than predict- ing it (Douglas & Kropp, 2002; Heilbrun, 1997). From a violence prevention perspective, actuarial methods can inform us about the overall level of risk management that might be required (i.e., the greater the risk, the greater the necessary resources). However, they do little to inform us about specific violence prevention strategies. Heilbrun (1997) contrasted “ prediction versus management ” models of risk assessment, noting that the prediction model likely has “ minimal ” implications for management due, in part, to its lack of sensitivity to change. To apply the actuarial approach properly, the evaluator is forced to consider a fixed set of factors and cannot consider unique, unusual, or context-specific variables that might require intervention (Hart, 1998). Moreover, actuarial instruments may lack a “ goodness of fit ” with offender treatment programs: there is incongruence between violence prevention program targets such as “ attitudes toward violence ” or “ denial and minimization ” and risk assessment instruments that fail to consider such things. Finally, although actuarial approaches give the appearance of objectivity and precision, they often yield very modest correlations with violence (Douglas et al., 1999) and are subject to limitations such as statistical shrinkage (incomplete replication on cross-validation in new populations) and measurement error. Moreover, practitioners may feel uncomfortable considering only one “ test ” of risk while ignoring legal, ethical, and professional require- ments to consider all available information from all perspectives (American Psychological Association, 2002).
Legal and professional practice must change considerably before professionals can abandon discretion in favor of strict actuarial methods. Unless and until such changes occur, professionals must decide how to strike the balance between scientific rigor and respect for the context of individual cases. Presumably in an effort to strike such a balance, some actuarial measures employ a professional discretion mechanism whereby the evaluator can
Risk Assessment 207
override an actuarial score (Andrews & Bonta, 1996). However, this approach seems to violate the fundamental premise of the actuarial method—that precise statistical forecasts should not be contaminated by human discretion (e.g., Quinsey et al., 1998).
Structured Professional Judgment
Structured professional judgment is an approach that attempts to bridge the gap between unstructured clinical and actuarial approaches to risk assessment (Douglas & Kropp, 2002; Hart, 1998). Here, the evaluator must conduct the assessment according to guidelines that reflect current theoretical, professional, and empirical knowledge about violence. Such guidelines provide the minimum set of risk factors that should be considered in every case. The guidelines will also typically include recommendations for information gather- ing (e.g., the use of multiple sources and multiple methods), communicating opinions, and implementing violence prevention strategies. The method is certainly more prescribed than the unstructured clinical approach but much more flexible than the actuarial method. Structured professional judgment does not impose any restrictions for the inclusion, weighting, or combining of risk factors. The flexibility is in the final step of combining risk factors, which is not done algorithmically. Structured professional judgment does not abrogate the professional responsibility and discretion of the evaluator, but it does attempt to improve the consistency and visibility of risk judgments. This approach has been used in the development of the Spousal Assault Risk Assessment Guide, or SARA (Kropp, Hart, Webster, & Eaves, 1994, 1995, 1999), and the Domestic Violence Screening Inventory (Williams & Houghton, 2004).
The primary goal of the structured professional approach to risk assessment is to prevent violence (Douglas & Kropp, 2002). By systematically identifying risk factors— particularly dynamic, or changeable, risk factors—relevant to a case, management strate- gies can be tailored to prevent violence. The structured professional approach allows for a logical, visible, and systematic link between risk factors and intervention in addition to the ability to identify persons who are at higher or lower risk for violence. It is vulnerable to some of the same criticisms as the unstructured clinical approach because it still allows considerable professional discretion. There is some evidence, however, of the reliability and validity of structured professional judgment guidelines such as the SARA (Douglas & Kropp, 2002). For example, a number of studies conducted in North America and Europe indicate that interrater reliability is good to excellent for professional judgments concerning the presence of individual risk factors and overall levels of risk (e.g., Belfrage, 1997; Kropp & Hart, 2000). Furthermore, professional judgments of risk have good criterion-related validity: They correlate substantially with scores on actuarial measures (e.g., Douglas & Webster, 1999; Kropp & Hart, 2000), they discriminate well between known groups of recidivists and nonrecidivists in retrospective research (e.g., Grann & Wedin, 2002; Hanson & Morton-Bourgon, 2004; Kropp & Hart, 2000), and they predict recidivism in prospective research (e.g., Belfrage, Fransson, & Strand, 2000; Williams & Houghton, 2004).
EXISTING RISK INSTRUMENTS
There has been a proliferation of domestic violence risk instruments or tools in recent years (Dutton & Kropp, 2000; Hilton & Harris, 2005; Roehl & Guertin, 1998). Many of these
208 Kropp
were developed by local correctional, police, or victim agencies, and the developers have not reported normative, reliability, and validity data. Others, like the Kingston Screening Instrument for Domestic Violence (K-SID; Gelles, 1998), showed initial promise but have not appeared recently in the literature. It thus appears that there are four risk assessment instruments that currently hold the most promise: Danger Assessment, Domestic Violence Screening Inventory, Ontario Domestic Assault Risk Assessment, and the Spousal Assault Risk Assessment Guide. Each is described below in some detail.
Danger Assessment (DA)
The Danger Assessment (DA) was developed by Jacquelyn Campbell (1995) and col- leagues in consultation with victims of domestic violence, law enforcement officials, those working in shelters, and other experts. It is designed to assess the likelihood for spousal homicide, and the original items were chosen from retrospective studies on homicide or near-fatal injury cases. It is available online: http://www.dangerassessment.com.
The DA consists of two sections. The first is a calendar that asks potential victims to record the severity and frequency of violence in the past year (1 = slap, pushing, no inju- ries, and/or lasting pain through 5 = use of weapon, wounds from weapon). This part of the measure is intended to raise awareness of the woman and reduce minimization of the abuse. In one initial study, 38% of women who initially reported no increase in severity and frequency changed their response to “ yes ” after filling out the calendar (Campbell et al., 2001). The second section consists of a 15-item yes/no list of risk factors associated with intimate partner homicide. The woman can complete the instrument independently or with the assistance of professionals working in the health care, victim advocate, or crimi- nal justice systems. The number of risk factors is then totaled, although the developer does not recommend using cutoff scores for decision making.
Campbell et al. (2001) summarized the results of 10 research studies conducted on the DA. In those studies, interrater reliability coefficients were in the moderate to good range ( r = . 60 to .86). According to Campbell et al., the DA has also been demonstrated to have strong test-retest reliability in two studies ( r = . 89 to .94). Construct validity has also been reported, with the DA discriminating between battered women in an emergency depart- ment and nonabused controls (Campbell, 1995) and with DA correlating strongly with other measures of abusive behavior such as the Index of Spouse Abuse and the Conflict Tactics Scale (Campbell, 1995). The DA is also associated with the severity and frequency of domestic violence (McFarlane et al., 1998).
Campbell and colleagues (2003) completed a multisite case control study to investigate the relative importance of various risk factors for femicide in abusive relationships. The study included many of the items from the original DA. The investigators interviewed 220 proxies of femicide victims along with 343 abused control women. The results indicated that risk factors discriminating between the two groups included perpetrator’s access to a gun and previous threat with a weapon; perpetrator’s stepchild in the home; victim estrangement, especially from a controlling partner; victim leaving abuser for another partner; and the perpetrator’s use of a gun in the homicide. Stalking, forced sex in the relationship, and abuse during pregnancy also bore some significance. All but one of the original 15 DA items were significantly associated with femicide, and the measure was subsequently revised to include additional risk factors not in the original version. Both the original and revised versions of the DA significantly discriminated between the femicide and abused control groups. This was a retrospective study, and Campbell et al. (2001) have
Risk Assessment 209
recommended that prospective studies are still needed to evaluate the predictive validity of the DA.
Domestic Violence Screening Inventory (DVSI)
The Domestic Violence Screening Inventory (DVSI) was developed by the Colorado Department of Probation Services. The DVSI was designed to be a brief risk assessment instrument that can be completed with a quick criminal history review. It contains 12 social and behavioral factors found to be statistically related to recidivism by domestic violence perpetrators on probation (Williams & Houghton, 2004). The authors also justified includ- ing the risk factors based on a thorough review of the literature, and they consulted judges, law enforcement personnel, lawyers, and victim advocates. The social factors include cur- rent employment and relationship status. The behavioral items essentially summarize the offender’s history of DV and non-DV criminal history. A copy of the DVSI coding sheet is included in an appendix of the Williams and Houghton (2004) validation paper.
The DVSI was validated on a sample of 1,465 male domestic violence offenders on probation, selected consecutively over a 9-month period. Data on reoffending were col- lected in a 6-month follow-up period from a subsample of the victims ( N = 125) of these perpetrators and from official records for all perpetrators during an 18-month follow-up period. The results suggest that the DVSI was administered reliably, although the authors acknowledged that the study’s design required multivariate analyses to conduct “quasi- interrater reliability.” The DVSI also appears to have adequate concurrent validity, cor- relating strongly with ratings of risk to spouses on the Spousal Assault Risk Assessment Guide (SARA). Finally, Williams and Houghton reported statistically significant predictive validity for the DVSI using a prospective (follow-up) design. The authors used a common method for analyzing effect size in risk assessment research labeled Receiver Operating Characteristic (ROC) analysis. The area under the curve (AUC) using this analysis was .60 ( r = .18, p < .001). One way of interpreting the AUC of .60 is as follows: If a recidivist and nonrecidivist were randomly chosen from their respective groups, the probability would be .60 that the recidivist would have a higher DVSI score. There have been no independent validity studies of the DVSI to date.
Ontario Domestic Assault Risk Assessment (ODARA)
The ODARA is a 13-item actuarial instrument recently developed in Ontario, Canada (Hilton et al., 2004). The items were empirically derived from an initial pool of potential risk factors gleaned from police files on 589 domestic violence perpetrators. The study followed the cases for an average of 5 years and coded the risk factors from archival information in several domains that included offender characteristics, domestic violence history, nondomestic criminal history, relationship characteristics, victim characteristics, and index offense. Using setwise and stepwise logistic regression, the developers reduced the item pool to 13. The resulting instrument, the ODARA, correlated well with the DA and the Spousal Assault Risk Assessment Guide (SARA; see below), thus demonstrating adequate convergent validity. The instrument also significantly discriminated between recidivists and nonrecidivists (AUC = .77), and the ODARA total score was also associ- ated with the number, severity, and imminence of new assaults. One shortcoming of the study was that there were no homicides in the construction sample, and the authors have cautioned against using the ODARA for predicting femicide. There is also a need for
210 Kropp
further cross-validation studies to substantiate the precise probabilities associated with each ODARA score.
Spousal Assault Risk Assessment Guide (SARA)
The Spousal Assault Risk Assessment Guide (Kropp, Hart, Webster, & Eaves, 1995, 1999) is a set of guidelines for the content and process of a thorough risk assessment. It consists of 20 items identified by a review of the empirical literature on wife assault and the literature written by clinicians who evaluate male wife abusers. The authors point out that the SARA is not a test. Its purpose is not to provide absolute or relative measures of risk using cutoff scores or norms but rather to structure and enhance professional judg- ments about risk. Since the SARA is not a formal psychological test, professionals other than psychologists can use it. The SARA assessment procedure includes interviews with the accused and victims, standardized measures of physical and emotional abuse, drug and alcohol abuse, and a review of collateral records, such as police reports, victim statements, criminal records, and other psychological procedures.
The authors have evaluated the reliability and validity of judgments concerning risk for violence made using the Spousal Assault Risk Assessment Guide (Kropp & Hart, 2000). SARA ratings were analyzed in six samples of adult male offenders (total N = 2,681). Structural analyses of the risk factors indicated moderate levels of internal consistency and item homogeneity. Interrater reliability was high for judgments concerning the presence of individual risk factors and for overall perceived risk. SARA ratings significantly dis- criminated between offenders with and without a history of spousal violence in one sample ( t = 27.04, p < .0001) and between recidivistic and nonrecidivistic spousal assaulters in another ( r = .36, p < .0001; or AUC = .70). Finally, SARA ratings showed good convergent and discriminant validity with respect to other measures related to risk for general and violent criminality (Kropp & Hart, 2000).
Two recently published studies support the validity of the SARA. Williams and Houghton (2004), in their evaluation of the DVSI, included the SARA in some of the analyses. Thus, the results also supported the concurrent validity of the SARA, and the AUC for the SARA in the 18-month follow-up exceeded that of the DVSI (.65 vs. .60, although the difference was not statistically significant). Similarly, Hilton et al. (2004) reported an AUC for the SARA of .64 in a 5-year follow-back study. The authors con- cluded that in this study the SARA did not postdict violence as well as the ODARA. However, that conclusion was significantly limited by the fact the authors did not use the administration procedure recommended in the SARA manual. They employed the SARA solely as an actuarial instrument, which it is not, and they noted that the “integrity” (p. 271) of the SARA scores could not be guaranteed because they were coded from archival data only.
Finally, due to calls from the field, particularly from law enforcement agencies, to have briefer risk assessment tools to conduct time-limited assessments, the authors of the SARA have developed the Brief Spousal Assault Form for the Evaluation of Risk, or B-SAFER (Kropp, Hart, & Belfrage, 2005 ). It consists of 10 risk factors, which were derived from the 20 SARA risk factors using factor analysis. The B-SAFER has been piloted in Canada and Sweden, and preliminary findings suggest that the B-SAFER ratings are associated with the type and number of management strategies recommended by police. Further, it appears that the use of the B-SAFER contributed to reduced recidivism rates in a sample of Swedish offenders (Kropp, 2004a, 2007).
Risk Assessment 211
Summary
There are now a number of risk assessment instruments circulating in the field. The con- tent of these tools is similar, and all of them have demonstrated some reliability and valid- ity. Moreover, it is apparent from the literature that the field is devoting more attention to issues of risk assessment and management. However, these instruments still only yield modest associations with recidivism (i.e., correlations roughly between .30 and .40), and this is unlikely to improve as we simply do not possess the technology to accurately predict violent behavior. Future research should therefore focus on the ability of these instruments to inform management practices and prevent violence.
THE ROLE OF THE VICTIM IN RISK ASSESSMENT
A risk assessment is only as reliable as the information upon which it is based. It is usually inadvisable to place much weight on self-reported information from the (alleged) offender. Such assessments will likely result in an underestimate of risk level. Those accused or con- victed of spousal violence typically are reluctant to disclose information that may affect their sentencing dispositions or release opportunities and conditions. Moreover, offenders often are in a state of denial or significantly minimize their responsibility for violence. These points underscore the important role of collateral informants in risk assessment. In this respect, the evaluator of spousal assault risk has a unique opportunity to involve potential victims or survivors, which is a distinct advantage over those performing risk assessments where the potential victims are usually unknown (e.g., sex offending, general violence).
Most agree that risk assessments should be victim-informed. Campbell’s Danger Assessment, for example, is designed entirely for use with victims. The authors of the Spousal Assault Risk Assessment Guide (SARA; Kropp et al., 1999) caution strongly against performing risk assessments without consulting known victims. They emphasize that a victim can provide crucial information regarding an offender’s violent past, person- ality, attitudes, and mental health. The importance of victim information has also been empirically demonstrated. Wiesz, Tolman, and Saunders (2000) reported that survivors’ predictions of reassault were significantly associated with the reoccurrence of severe violence. Similarly, Gondolf (2001) found that in a 30-month follow-up of court-mandated batterers, the most significant predictors of reoffense were offender drunkenness and wom- en’s perceptions of safety. Whittemore and Kropp (2001) reported a study in which SARA ratings of risk were made using offender and file information only and then compared to rat- ings made with additional victim-reported information. The results revealed that risk ratings made with the added victim information were higher than those made without. It seems, therefore, that victims can provide some critical information related both to professionals’ perceptions of risk and recidivism. Bennett Cattaneo’s (2007) thoughtful review explores these issues in more detail. It is important to remember, however, that victims’ perceptions of risk are not always accurate (Weisz et al., 2000). Victims can also grossly minimize or underestimate the risk posed by their partners. Campbell et al. (2001) reported the results of an investigation of actual and attempted femicides. Proxy informants gathered informa- tion regarding the actual homicides. Campbell et al. noted that victims underestimated their spouse’s risk in 47% and 53% of the actual and attempted femicides, respectively. Thus, for a variety of reasons (e.g., fear, denial, absence of objective risk factors), there
212 Kropp
are likely many situations in which victims are not able to recognize the risks posed by their partners.
Although few would argue against including victims in the risk assessment process, real-life situations present some complications. First, a victim might simply be unavailable due to relocation or a general reluctance to cope with the justice system. Confidentiality of information is another central problem. Victims may be reluctant to provide risk-relevant information and/or violence predictions if they perceive that it will place them at further risk. It is conceivable, for example, that the risk posed to the victim will increase in the short term following her participation in a court-related assessment of the abuser. Asking the victim to “predict” her abuser’s future behavior places her in a difficult position, espe- cially if that prediction is going to be used to make decisions about the offender’s life.
Thus, victims will be reluctant to participate in a risk assessment for all the same rea- sons they are reluctant to involve the criminal and civil justice systems in the first place (Barnett, 2001). They might fear for their own safety, be protecting the offender, or be concerned for their children’s safety and welfare. For example, a woman may fear cor- rectly that her child will be apprehended should she predict that her husband will assault her again in the presence of that child. A victim might also be required to endure cross- examination should her information be presented to court. Therefore, informed consent prior to participation is of paramount importance, as victims need to know exactly how the information they provide will be used. This is easier said than done, however, as it is difficult to know how a court will use sensitive information. Jurisdictions that advocate for victim participation in the risk assessment process must therefore confront the ethical, legal, and practical barriers involved. It must be emphasized that if victim information is not available, risk assessments should be appropriately qualified. Of course, these consid- erations apply to the use of all collateral informants whether they are spouses, other family members, friends, and so forth.
QUALIFICATIONS TO CONDUCT ASSESSMENTS
Despite the widespread application of spousal violence risk assessment there exist no professional standards for (a) the minimal qualifications of those conducting the assess- ments; (b) “best practices” for applying the assessments; (c) training of the assessors; and (d) evaluation and monitoring of the assessments (Borum, 1996). Thus, professionals are told what to do but not how to do it. Discussion of such issues appears to be absent from the spousal violence literature despite growing attention to risk technology (Campbell et al., 2001; Dutton & Kropp, 2000; Hilton et al., 2004). Of course, standards must vary according to the context of the risk assessment. Some risk assessments will be necessar- ily brief, such as those conducted during a police investigation or in a hospital emergency room. Other assessments will be of greater duration and complexity, such as those com- pleted for a presentencing hearing in criminal court. Professionals—for example, police, nurses, victim service workers, and psychologists—also will have varying degrees of edu- cation, training, and experience with respect to spousal violence. For example, it is unclear how nonmental health professionals such as police and probation officers should assess risk markers involving mental health problems such as personality disorders and suicidal ideation (Dutton & Kropp, 2000; Kropp , 2004b).
It is time for administrators, licensing bodies, and government agencies to set and enforce standards for risk assessment practice. There is a precedent for this in the area
Risk Assessment 213
of sexual violence risk assessment, where the Association for the Treatment of Sexual Abusers (ATSA) has made a position statement on risk assessment (Hanson, 2000). At the very least, those conducting risk assessments should have some expertise and experi- ence in interviewing and assessing offenders and victims. Moreover, they should have considerable knowledge of the dynamics of spousal violence. Finally, assessments should be completed with the assistance of risk assessment guidelines or tools that have some acceptance in the scientific and professional communities. Training and monitoring should be implemented to fill any gaps in qualifications that might exist.
COMMUNICATING RISK
Regardless of how well a violence risk assessment is conducted, the mode of communica- tion will greatly affect how the information is received and utilized (Heilbrun et al., 2000 ; Litwack, 1997). Effective risk communication can and should prevent violence. Domestic violence fatality reviews (Denver Metro Domestic Violence Fatality Review Committee, 2000; Websdale, 1999; Websdale, Sheeran, & Johnson, 1998) tell us that in many cases of spousal homicide, many risk indicators were present and known but not necessarily documented or communicated to those who needed to know such as the victims, offender treatment providers, police, correctional agencies, and so forth.
There are several principles of sound risk communication. First, professionals offering risk opinions must support their opinion clearly, concisely, and with appropriate evidence. To do so they must have a language for communicating risk. Those conducting assess- ments in the field are often left with relying on their instincts without a way to articulate their concerns. Existing risk assessment guidelines or checklists, such as the DA, ODARA, SARA, and the DVSI, can be helpful in this respect. Such tools can help structure and support opinions about risk. Moreover, a risk language can help articulate concerns and can be more compelling for the audience (see Hilton, Harris, Rawson, & Beach, 2005). For example, a risk assessment vaguely presented and worded is less powerful than a risk opinion supported by a concise list of risk factors (e.g., “Mr. B is at risk because of risk factors X, Y, and Z”).
Second, whenever possible, risk assessments should be communicated to the potential victim(s). This is especially important in high-risk cases or situations but can be useful information for the victim regardless of risk level. Risk assessments, properly commu- nicated, can inform a victim regarding overall level of risk, so she can take precautions but also be educated about specific risk factors. Many risk factors—such as mental health problems, employment instability, and substance abuse—can be perceived by potential victims and others as sympathy factors rather than causal factors of violence (Kropp, Hart, Lyon, & LePard, 2002). Thus, the very process of risk communication can be enlightening for those victims who are naïve about the existing danger or who mini- mize their partners’ violence and can serve as an important component of a safety plan (Campbell, 1995).
Third, risk opinions must be appropriately qualified and must be accompanied by an explanation of the limitations of the assessment. For example, if an interview with the victim is not possible, it should be explained to those concerned that this could seriously affect the validity of the assessment. In this sense, there is nothing more dangerous than a risk assessment based on inadequate information that does not include appropriate qualifi- cations. This can lead to an underestimate of risk and mislead victims, protective agencies,
214 Kropp
and other interested parties. Risk assessments should also, whenever possible, discuss the nature, frequency, severity, likelihood, and imminence of the violence. Such an assessment can be more informative for risk management (discussed below) and safety planning than simply a global prediction of likelihood. One way of accomplishing all of these objectives is the anchored narrative approach, a method by which the evaluator constructs possible risk scenarios, listing the relevant risk factors, management strategies, and possible out- comes of each (Hart, 2002). Note that if actuarial predictions are offered, the limitations of these estimates must also be specified. The audience for the risk assessment must be given enough information to interpret the actuarial scores. Thus, the evaluator should discuss base rates of violence, reliability of the instrument used, measurement error, cross- validation shrinkage, and so forth. In this case, the evaluator must also be qualified to provide interpretation (American Educational Research Association, 1999).
MANAGING RISK
Risk assessment should help evaluators identify risk management strategies. Developing risk management plans is a difficult business. Optimally, it requires familiarity with and cooperation among a number of different professionals working in different agencies, each with a different skill set and mandate. The development and implementation of comprehensive, integrated, multidisciplinary risk management plans is best accomplished with the assistance of a guiding policy and procedure manual (Kropp, Hart, Lyon, et al., 2002). Evaluators should consider the initiation or implementation of four basic kinds of risk management activities: monitoring, treatment, supervision, and victim safety planning (Kropp, Hart, Lyon, et al., 2002).
Monitoring, or repeated assessment, is always a part of good risk management. The goal of monitoring is to evaluate changes in risk over time so that risk management strat- egies can be revised as appropriate. Monitoring services may be delivered by a diverse range of mental health, social service, law enforcement, corrections, and private security professionals. Monitoring, unlike supervision, focuses on surveillance rather than on con- trol or restriction of liberties; it is therefore minimally intrusive. Monitoring strategies may include contacts with the client as well as with potential victims and other relevant people (e.g., therapists, correctional officers, family members, coworkers) in the form of face-to- face or telephonic meetings. Where appropriate, they may also include field visits (e.g., at home or work), electronic surveillance, polygraphic interviews, drug testing (urine, blood, or hair analysis), and inspection of mail or telecommunications (telephone records, fax logs, e-mail, etc.). Frequent contacts by the client with health care and social service professionals are an excellent form of monitoring; missed appointments with treatment providers are a warning sign that the client’s compliance with treatment and supervision may be deteriorating.
Treatment involves the provision of (re-) habilitative services. The goal of treatment is to improve deficits in the individual’s psychosocial adjustment. One important form of treatment is directed at mental disorder that is causally related to the individual’s his- tory of violence. Although there is as yet no direct evidence that various treatments for mental disorder decrease violence, it is possible—and even likely—that they will have a beneficial impact. Treatments may include individual or group psychotherapy; psycho- educational programs designed to change attitudes toward violence; training programs designed to improve interpersonal, anger management, and vocational skills; psychoactive
Risk Assessment 215
medications, such as antipsychotics or mood stabilizers; and chemical dependency pro- grams. Another important form of treatment is the reduction of acute life stresses, such as physical illness, interpersonal conflict, unemployment, legal problems, and so forth.
Supervision involves the restriction of the individual’s rights or freedoms. The goal of supervision is to make it (more) difficult for the individual to engage in further violence. An extreme form of supervision is incapacitation, that is, involuntary institutionalization of the individual in a correctional or health care facility. Incapacitation clearly is an effective means of reducing the individual’s access to potential victims. Community supervision is much more common than institutionalization. Typically, it involves allowing the individual to reside in the community with restrictions on activity, movement, association, and com- munication. In general, supervision should be implemented at a level of intensity commen- surate with the risks posed by the individual. This helps protect the individual’s civil rights and also helps reduce the liability of people involved in providing supervision services.
Victim safety planning involves improving the victim’s dynamic and static security resources, a process sometimes referred to as “target hardening.” The goal is to ensure that, if violence recurs—despite all monitoring, treatment, and supervision efforts—any negative impact on the victims’ psychological and physical well-being is minimized. Victim safety planning services may be delivered by a wide range of social service, human resource, law enforcement, and private security professionals. These services can be deliv- ered regardless of whether the individual is in an institution or the community. Victim safety planning is most relevant in situations that involve “targeted violence,” that is, where the identity of the likely victims of any future violence is known.
CONCLUDING COMMENTS
The body of literature on domestic violence risk assessment is growing rapidly, but there is still much to be done. Any agency considering the implementation of risk assessment protocols must therefore recognize that this is an imperfect enterprise. I have attempted to present the “state of the art” in this article, and certain themes have emerged. In sum- mary:
1. Risk should be defined broadly and should not focus only on the narrow criterion of “likelihood” of reoffense. The nature, severity, frequency, and imminence of the violence must also be considered. That said, existing risk instruments are not precise enough to reliably discriminate among different types of risk (e.g., who will commit homicide vs. who will commit less serious violence). Therefore, some professional discretion will likely always be necessary.
2. The risk factors for domestic violence—that is, the variables that are reliably asso- ciated with this form of violence—are well-established in the literature. We know what they are, and it is unlikely that any new risk factors will turn up soon. However, there is some debate about whether to consider risk factors with solid theoretical foundation but limited empirical support.
3. The choice of a method of risk assessment is complicated. The most viable options are either a structured professional judgment approach or an actuarial procedure. The strengths and weaknesses of both approaches have been presented.
4. There are several existing risk instruments, all of which have similar content and some of which have established psychometric reliability and validity.
216 Kropp
5. It is important for those conducting risk assessments to not rely solely on self- reported information from the offender. It is also critical to use victim-based infor- mation somehow. This must be done sensitively with attention paid to ethical and safety issues.
6. Not everyone can or should perform risk assessments. Such evaluations require specialized knowledge and experience. Those conducting risk assessments should understand the dynamics of domestic violence, and they should have experience working with offenders and victims. Proper risk assessment training is extremely important.
7. Risk assessment training should include issues regarding report writing and commu- nication. Risk assessments that are not effectively communicated to decision-makers and victims are essentially useless.
8. Finally, while risk assessment can inform us about who should be a priority to receive spousal violence treatment, it is important to recognize that risk management involves far more than just domestic violence programming. Other specialized treat- ments might be necessary, as well as proper monitoring and supervision. Moreover, victim safety planning is crucial, as offender intervention is far from perfect for preventing future violence.
REFERENCES
American Educational Research Association, American Psychological Association, and the National Council on Measurement in Education. (1999). Standards for educational and psychological testing. Washington, DC: American Educational Research Association.
American Psychological Association. (2002). Ethical principles of psychologists and code of con- duct. Washington, DC: Author.
Andrews, D. A., & Bonta, J. (1996). The level of service inventory-revised: User ’ s manual. Toronto: Multi-Health Systems, Inc.
Andrews, D. A., & Bonta, J. (2003). The psychology of criminal conduct (3rd ed.). Cincinnati, OH: Anderson.
Barnett, O. (2001). Why battered women do not leave, part 2: External inhibiting factors—social support and internal inhibiting factors. Trauma, Violence, & Abuse, 2 (1), 3–35.
Belfrage, H. (1997). The interrater reliability of the HCR-20 scheme for risk assessment: A study of 6 clinicians’ assessments of the same 43 patients. Nordic Forensic Medicine, 3, 79–81.
Belfrage, H., Fransson, G., & Strand, S. (2000). Prediction of violence using the HCR-20: A prospective study in two maximum security correctional institutions. Journal of Forensic Psychiatry, 11, 167–175.
Bennett Cattaneo, L. B. (2007). Contributors to assessments of risk in intimate partner violence: How victims and professionals differ. Journal of Community Psychology, 35 (1), 57–75.
Bennett Cattaneo, L. B., & Goodman, L. A. (2005). Risk factors for reabuse in intimate partner vio- lence: A cross-disciplinary critical review. Trauma, Violence, & Abuse, 6 (2), 141–175.
Borum, R. (1996). Improving the clinical practice of violence risk assessment. American Psychologist, 51 (9), 945–953.
Burgess, A. W., Baker, T., Greening, D., Hartman, C. R., Burgess, A. G., Douglas, J. E., et al. (1997). Stalking behaviors within domestic violence. Journal of Family Violence, 12, 389–403.
Cadsky, O., & Crawford, M. (1988). Establishing batterer typologies in a clinical sample of men who assault their female partners. Special Issue: Wife battering: A Canadian perspective. Canadian Journal of Community Mental Health, 7, 119–127.
Risk Assessment 217
Campbell, J. C. (1995). Prediction of homicide of and by battered women. In J. C. Campbell (Ed.), Assessing dangerousness: Violence by sexual offenders, batterers, and child abusers (pp. 96–113). Thousand Oaks, CA: Sage.
Campbell, J. C., Sharps, P., & Glass, N. (2001). Risk assessment for intimate partner homicide. In G. F. Pinard & L. Pagani (Eds.), Clinical assessment of dangerousness: Empirical contributions (pp. 137–157). New York: Cambridge University Press.
Campbell, J. C., Webster, D., Koziol-McLain, J., Block, C., Campbell, D., Curry, M. A., et al. (2003). Risk factors for femicide in abusive relationships: Results from a multi-site case control study. American Journal of Public Health, 93, 1089–1097.
Carlson, M. J., Harris, S. D., & Holden, G. W. (1999). Protective orders and domestic violence: Risk factors for reabuse. Journal of Family Violence, 14 (2), 205–226.
Daly, M., & Wilson, M. (1998). An evolutionary psychological perspective on homicide. In M. D. Smith & M. Zahn (Eds.), Homicide: A sourcebook of social research (pp. 58–71). Thousand Oaks, CA: Sage.
Denver Metro Domestic Violence Fatality Review Committee. (2000). When domestic violence kills. Denver, CO: Project Safeguard.
Dobash, R. E., Dobash, R. P., & Cavanagh, K. (2007). Lethal and nonlethal violence against an intimate female partner: Comparing male murderers to nonlethal abusers. Violence Against Women, 13 (4), 329–353.
Douglas, K. S., Cox, D. N., & Webster, C. D. (1999). Violence risk assessment: Science and practice. Legal and Criminological Psychology, 4, 149–184.
Douglas, K., & Kropp, P. R. (2002). A prevention-based paradigm for violence risk assessment: Clinical and research applications. Criminal Justice and Behavior, 2, 617–658.
Douglas, K. S., & Webster, C. D. (1999). Predicting violence in mentally and personality disordered individuals. In R. Roesch, S. D. Hart, & J. R. P. Ogloff (Eds.), Psychology and law: The state of the discipline (pp. 175–239). New York: Plenum.
Dutton, D. G. (1995). The domestic assault of women: Psychological and criminal justice perspec- tives (rev. ed.). Vancouver, BC: UBC Press.
Dutton, D. G., Bodnarchuk, M., Kropp, P. R., Hart, S. D., & Ogloff, J. (1997). Wife assault treatment and criminal recidivism: An eleven-year follow up. International Journal of Offender Therapy and Comparative Criminology, 41 (1), 9–23.
Dutton, D. G., & Hart, S. D. (1992). Evidence for long-term, specific effects of childhood abuse on criminal behavior in men. International Journal of Offender Therapy and Comparative Criminology, 36, 129–137.
Dutton, D. G., & Kropp, P. R. (2000). A review of domestic violence risk instruments. Trauma, Violence & Abuse, 1, 171–181.
Gelles, R. (1998, October). Lethality and risk assessment for family violence cases. Paper presented at the 4th International Conference on Children Exposed to Family Violence, San Diego, CA.
Gondolf, E. W. (1988). Who are those guys? Toward a behavioral typology of batterers. Violence and Victims, 3, 187–203.
Gondolf, E. W. (1998). Assessing woman battering in mental health services. Thousand Oaks, CA: Sage.
Gondolf, E. W. (2001). Batterer intervention systems: Issues, outcomes, and recommendations. Thousand Oaks, CA: Sage.
Gondolf, E. W., & White, R. J. (2001). Batterer program participants who repeatedly reassault: Psychopathic tendencies and other disorders. Journal of Interpersonal Violence, 16, 361–380.
Grann, M., & Wedin, I. (2002). Risk factors for recidivism among spousal assault and spousal homi- cide offenders. Psychology, Crime & Law, 8, 5–23.
Greenland, C. (1985). Dangerousness, mental disorder, and politics. In C. D. Webster, M. H. Ben-Aron, & S. J. Hucker (Eds.), Dangerousness: Probability and prediction, psychiatry and public policy (pp. 25–40). New York: Cambridge University Press.
218 Kropp
Grove, W. M., & Meehl, P. E. (1996). Comparative efficiency of informal (subjective, impres- sionistic) and formal (mechanical, algorithmic) prediction procedures: The clinical-statistical controversy. Psychology, Public Policy, and Law, 2, 293–323.
Hanson, R. K. (2000). Risk assessment: Prepared for the Association for the Treatment of Sexual Abusers. Beaverton, OR: ATSA.
Hanson, R. K., & Morton-Bourgon, K. (2004). Predictors of sexual recidivism: An updated meta- analysis. Ottawa, Canada: Public Safety and Emergency Preparedness.
Hanson, R. K., & Wallace-Capretta, S. (2000). Predicting recidivism among male batterers. User Report 2000 – 06. Ottawa: Department of the Solicitor General of Canada.
Harrell, A., & Smith, B. (1996). Effects of restraining orders on domestic violence victims. In E. Buzawa & C. Buzawa (Eds.), Do arrests and restraining orders work? (pp. 214 –242). Thousand Oaks, CA: Sage.
Hart, B. J. (1992). Program standards for batterer intervention services. Reading, PA: Pennsylvania Coalition Against Domestic Violence.
Hart, S. D. (1998). The role of psychopathy in assessing risk for violence: Conceptual and method- ological issues. Legal and Criminological Psychology, 3, 121–137.
Hart, S. D. (2001). Assessing and managing violence risk. In K. S. Douglas, C. D. Webster, S. D. Hart, D. Eaves, & J. R. P. Ogloff (Eds.), HCR-20 violence risk management companion guide (pp. 13–25). Burnaby, British Columbia: Mental Health, Law, & Policy Institute, Simon Fraser University, and Department of Mental Health Law and Policy, Florida Mental Health Institute, University of South Florida.
Hart, S. D. (2002, September). Violence risk assessment: An anchored narrative approach. Paper presented at the 12th European Conference on Psychology and Law, Leuven, Belgium.
Hart, S. D., Kropp, P. R., & Hare, R. D. (1988). Performance of male psychopaths following condi- tional release from prison. Journal of Consulting and Clinical Psychology, 56, 227–232.
Heilbrun, K. (1997). Prediction versus management models relevant to risk assessment: The impor- tance of legal decision-making context. Law and Human Behavior, 21, 347–359.
Hilton, N. Z., & Harris, G. T. (2005). Predicting wife assault: A critical review and implications for policy and practice. Trauma, Violence, & Abuse, 6 (1), 3–23.
Hilton, N. Z., Harris, G. T., & Rice, M. E. (2001). Predicting violence by serious wife assaulters. Journal of Interpersonal Violence, 16, 408 – 423.
Hilton, N. Z., Harris, G. T., Rawson, K., & Beach, C. A. (2005). Communicating violence risk infor- mation to forensic decision makers. Criminal Justice and Behavior, 32 (1), 97–116.
Hilton, N. Z., Harris, G. T., Rice, M. E., Lang, C., & Cormier, C. A. (2004). A brief actuarial assess- ment for the prediction of wife assault recidivism: The ODARA. Psychological Assessment, 16 (3), 267–275.
Hotaling, G. T., & Sugarman, D. B. (1986). An analysis of risk markers in husband-to-wife violence: The current state of knowledge. Violence and Victims, 1, 101–124.
Jones, A. S., & Gondolf, E. W. (2001). Time-varying risk factors for reassault among batterer pro- gram participants. Journal of Family Violence, 16, 345–359.
Kessler, R. C., Molnar, B. E., Feurer, I. D., & Appelbaum, M. (2001). Patterns and mental health predictors of domestic violence in the United States: Results from the National Comorbidity Survey. International Journal of Law and Psychiatry, 24, 487–508.
Kropp, P. R. (2004a, September). The Brief Spousal Assault Form for the Evaluation of Risk: B-SAFER. Paper presented at the 2nd International Conference—Toward a Safer Society. Edinburgh, Scotland.
Kropp, P. R. (2004b). Some questions about spousal violence risk assessment. Violence Against Women, 10, 696–697.
Kropp, P. R. (2007). The development of the SARA and the B-SAFER. In A. C. Baldry & F. W. Winkel (Eds.), Assessing risk of spousal assault: An international approach. Hauppauge, NY: Nova Science Publishers.
Risk Assessment 219
Kropp, P. R., & Hart, S. D. (2000). The Spousal Assault Risk Assessment (SARA) Guide: Reliability and validity in adult male offenders. Law and Human Behavior, 24, 101–118.
Kropp, P. R., Hart, S. D., & Belfrage, H. (2005). The Brief Spousal Assault Form for the Evaluation of Risk (B-SAFER). Vancouver, BC: Proactive Resolutions.
Kropp, P. R., Hart, S. D., Lyon, D., & LePard, D. (2002). Managing stalkers: Coordinating treat- ment and supervision. In L. Sheridan & J. Boon (Eds.), Stalking and psychosexual obses- sion: Psychological perspectives for prevention, policing and treatment. Chichester, UK: Wiley.
Kropp, P. R., Hart, S. D., Webster, C. D., & Eaves, D. (1994). Manual for the Spousal Assault Risk Assessment Guide. Vancouver: British Columbia Institute on Family Violence.
Kropp, P. R., Hart, S. D., Webster, C. D., & Eaves, D. (1995). Manual for the Spousal Assault Risk Assessment Guide (2nd ed.). Vancouver: British Columbia Institute on Family Violence.
Kropp, P. R., Hart, S. D., Webster, C. D., & Eaves, D. (1999). Spousal Assault Risk Assessment Guide (SARA). Toronto: Multi-Health Systems, Inc.
Kyriacou, D. N., Anglin, D., Taliaferro, E., Stone, S., Tubb, T., Linden, J. A., et al. (1999). Risk factors for injury to women from domestic violence. New England Journal of Medicine, 341, 1882–1898.
Litwack, T. R. (1997). Communications regarding risk. American Psychologist, 52 (11), 1245. Litwack, T. R. (2001). Actuarial versus clinical assessments of dangerousness. Psychology, Public
Policy, and Law, 7, 409–443. Litwack, T. R., & Schlesinger, L. B. (1999). Dangerousness risk assessments: Research, legal and
clinical considerations. In A. K. Hess & I. B. Weiner (Eds.), Handbook of forensic psychology (2nd ed., pp. 171–217). New York: Wiley.
Magdol, L., Moffitt, T. E., Caspi, A., Newman, D. L., Fagan, J., & Silva, P. A. (1997). Gender dif- ferences in partner violence in a birth cohort of 21-year-olds: Bridging the gap between clinical and epidemiological approaches. Journal of Consulting and Clinical Psychology, 65, 68–78.
McFarlane, J., Campbell, J. C., & Watson, K. (2002). Intimate partner stalking and femicide: Urgent implications for women’s safety. Behavioral Sciences and the Law, 20, 51–68.
McFarlane, J., Soeken, K., Campbell, J. C., Parker, B., Reel, S., & Silva, C. (1998). Severity of abuse to pregnant women and associated gun access of the perpetrator. Public Health Nursing, 15, 201–206.
Mulvey, E. P., & Lidz, C. W. (1995). Conditional prediction: A model for research on dangerousness to others in a new era. International Journal of Law and Psychiatry, 18 , 129–143.
Palarea, R. E., Zona, M. A., Lane, J. C., & Langhinrichsen-Rohling, J. (1999). The dangerous nature of intimate relationship stalking: Threats, violence, and associated risk factors. Behavioral Sciences and the Law, 17, 269–283.
Quinsey, V. L., Harris, G. T., Rice, G. T., & Cormier, C. A. (1998). Violent offenders: Appraising and managing risk. Washington, DC: American Psychological Association.
Riggs, D. S., Caulfield, M. B., & Street, A. E. (2000). Risk for domestic violence: Factors associated with perpetration and victimization. Journal of Clinical Psychology, 56, 1289–1316.
Rosenfeld, B. D. (1992). Court-ordered treatment of spouse abuse. Clinical Psychology Review, 12 , 205–226.
Saunders, D. G. (1992). Woman battering. In R. T. Ammerman & M. Hersen (Eds.), Assessment of family violence: A clinical and legal sourcebook (pp. 208–235). New York: Wiley.
Schumacher, J. A., Feldbau-Kohn, S., Slep, A. M. S., & Heyman, R. E. (2001). Risk factors for male- to-female partner physical abuse. Aggression and Violent Behavior, 6, 281–352.
Sherman, L. W., Smith, D. A., Schmidt, J. D., & Rogan, D. P. (1992). Crime, punishment and the stake in conformity: Legal and informal control of domestic violence. American Sociological Review, 57, 680 – 690.
Sjöstedt, G., & Grann, M. (2003). Risk assessment: What is being predicted by actuarial prediction instruments? International Journal of Forensic Mental Health, 1 (2), 179–184.
220 Kropp
Slovic, P., Monahan, J., & MacGregor, D. G. (2000). Violence risk assessment and risk communica- tion: The effects of using actual cases, providing instruction, and employing probability versus frequency formats. Law and Human Behavior, 24 (3), 271–296.
Sonkin, D. J. (1987). The assessment of court-mandated male batterers. In D. J. Sonkin (Ed.), Domestic violence on trial: Psychological and legal dimensions of family violence (pp. 174 – 196). New York: Springer Publishing.
Stuart, E. P., & Campbell, J. C. (1989). Assessment of patterns of dangerousness with battered women. Issues in Mental Health Nursing, 10, 245–260.
Tolman, R. M., & Bennett, L. W. (1990). A review of research on men who batter. Journal of Interpersonal Violence, 5, 87–118.
Tweed, R. G., & Dutton, D. G. (1998). A comparison of impulsive and instrumental subgroups of batterers. Violence and Victims, 13, 217–230.
Websdale, N. (1999). Understanding domestic homicide. Boston: Northeastern University Press. Websdale, N., Sheeran, M., & Johnson, B. (1998). Domestic violence fatality reviews: Summarizing
national developments. National Council of Juvenile and Family Court Judges, Reno, NV. Weisz, A. N., Tolman, R. M., & Saunders, D. G. (2000). Assessing the risk of severe domestic vio-
lence: The importance of survivors’ predictions. Journal of Interpersonal Violence, 15, 75–90. Whittemore, K. E., & Kropp, P. R. (2001, August). The victim ’ s impact in spousal assault risk
assessment. Paper presented at annual meeting of the American Psychological Association, San Francisco, CA.
Williams, K., & Houghton, A. B. (2004). Assessing the risk of domestic violence reoffending: A validation study. Law and Human Behavior, 24 (4), 437–455.
Wilson, M., & Daly, M. (1993). Spousal homicide risk and estrangement. Violence and Victims, 8, 3–16.
Acknowledgments. Portions of this article were previously published in Kropp (2004). Some ques- tions about spousal violence risk assessment: Violence Against Women, 10, 676 – 697. Permission to reproduce the material here was provided by Sage Publications, Thousand Oaks, CA.
Correspondence regarding this article should be directed to P. Randall Kropp, PhD, Psychologist, B.C. Forensic Psychiatric Services Commission, Suite 300 –307, West Broadway, Vancouver, BC, Canada, V5Y 2B6. E-mail: [email protected]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
