Using a General Case Management Tool With Partner-violent Men on Community Supervision in Iowa

Abstract

Intimate partner violence (IPV) is among the most common acts of violence against women worldwide, making it a major global threat to women’s health and safety. The assessment and management of IPV offenders are therefore vital tasks in criminal justice systems. The current study examined whether the DRAOR, a general case management tool, was useful for supervising 112 male IPV offenders in Iowa, United States. Several risk factors emerged as potentially important treatment targets for partner-violent men, including poor attachment with others, substance abuse, anger/hostility, opportunity/access to victims, and problematic interpersonal relationships. While further research is needed to improve the utility of the DRAOR for predicting IPV recidivism, it assesses several factors that are relevant for supervising IPV cases (e.g., substance abuse, anger/hostility, victim access). This suggests the DRAOR could potentially be used to guide case management in the presence of a validated IPV tool that focuses on static risk factors, such as the ODARA. The use of the DRAOR with IPV offenders may also be warranted if they are found to be generally violent/antisocial rather than as family only offenders.

Keywords

intimate partner violence partner-violent men offender management risk assessment recidivism predictive accuracy

Introduction

Intimate partner violence (IPV) is among the most common acts of violence against women worldwide, making it a major global threat to women’s health and safety. Roughly 35% of women worldwide are victimized by an intimate partner in their lifetime (World Health Organization [WHO], 2017). In the United States, over 1 in 3 (37%) women reported experiencing physical violence, contact sexual violence, and/or stalking by an intimate partner in their lifetime (Smith et al., 2017). Further, IPV is the leading cause of treatment for injuries among women each year in the United States (Anderson et al., 2015). IPV is also linked to a multitude of negative emotional and psychological health effects (e.g., depression, fear, anxiety, post-traumatic stress disorder, eating and sleep disorders; Campbell et al., 2002; Smith et al., 2017; WHO & Pan American Health Organization [PAHO], 2012). Finally, the public health costs associated with physical and sexual IPV against women are also staggering; they were estimated at $5.8 billion U.S. in 1995, which translated to roughly $8.3 billion in 2003 (Max et al., 2004), a number that would be considerably higher today due to inflation and a growing population. These statistics highlight the continuing need for effective IPV assessment and intervention.

DRAOR Development and Implementation

Efforts to improve offender management have resulted in the rise of offender change research, which evaluates changeable factors linked to crime. For instance, dynamic risk factors—classified as stable (more durable) or acute (more volatile; Hanson & Harris, 2000)—have been linked to a multitude of negative community outcomes, including rearrests and hospital readmissions (Olver et al., 2007; Penney et al., 2016). Protective factors, putatively, mitigate recidivism by shielding the offender from the effects of dynamic risk (Ullrich & Coid, 2011). These factors represent internal strengths (e.g., prosocial identity) and external assets (e.g., social support) and have predicted desistance from various type of criminal behavior (de Vries Robbé et al., 2015; Haines et al., 2018; Persson et al., 2017).

The Dynamic Risk Assessment for Offender Re-entry (DRAOR) is a 19-item structured instrument developed to inform case management practice. Specifically, it allows community parole officers to assess their clients at every supervision contact, helping them create individualized case plans and risk management strategies based on real-time changes in offender risk (Serin, 2007, 2015, 2017). The DRAOR evaluates psychosocial and contextual variables across Stable, Acute, and Protective domains. Given their relationship with recidivism, increases or decreases in DRAOR item scores should translate into changes in case planning by officers.

The DRAOR has been administered to over 6,000 male and female community-supervised offenders in Iowa and New Zealand. Total scores have significantly predicted any recidivism (including technical violations; 100% of the effect sizes in 4 studies were significant), general recidivism (excluding technical violations; 89% of the effect sizes in 7 studies were significant, with one of these studies yielding nonsignificant results), violent recidivism (67% of the effect sizes in 3 studies were significant, with one of these studies yielding nonsignificant results), and technical violations (89% of the effect sizes in 6 studies were significant). Significant effects ranged from small to large for any recidivism, general recidivism, and technical violations, and small to moderate for violent recidivism. The evidence for the DRAOR as an accurate predictor of sexual recidivism risk is mixed. Refer to Perley-Robertson (2018) and the supplemental table online for a more detailed overview of these results.

DRAOR With Intimate Partner Violence Offenders

Items assessed by the DRAOR that have been identified in the literature as significant predictors of IPV recidivism include impulse control, attachment with others, interpersonal relationships, substance abuse, anger/hostility, opportunity/access to victims, negative mood, and employment. Perpetrator urgency (the tendency to act impulsively in response to either negative or positive affect) was significantly positively associated with physical IPV in heterosexual couples (Leone et al., 2016). Impulsivity also predicted psychological and physical aggression in a female sample of IPV offenders (Shorey et al., 2011).

Several studies have examined how poor attachment with others and problematic interpersonal relationships influence IPV. Hilton and Harris (2005) identified marital conflict as a risk factor for repeat IPV in their review of factors predicting male-perpetrated wife assault. Marital conflict has also been linked to both the occurrence and severity of violence among male wife assaulters (Aldarondo & Sugarman, 1996). Further, a recent study found that relationship problem severity was significantly associated with male-perpetrated physical and emotional IPV (LaMotte et al., 2018).

In a critical review of 62 studies published between January 1990 and September 2003, Bennett Cattaneo and Goodman (2005) identified common risk factors for repeat male-perpetrated IPV against women. Alcohol and/or drug abuse was the most commonly identified dynamic risk predictor, reported as being significantly associated with reabuse in 8 of the 16 studies assessing this variable (6 positively and 2 negatively associated). Substance abuse has further been identified as a significant predictor of male- and female-perpetrated IPV onset and recidivism in various other studies (Almond et al., 2017; Brem et al., 2018).

Moreover, anger and hostility were moderately associated with male and female IPV perpetration in two meta-analytic reviews, both in terms of occurrence and severity (Birkley & Eckhardt, 2015; Norlander & Eckhardt, 2005). Birkley and Eckhardt (2015) also found a small effect for the relationship between internalizing negative emotions (e.g., anxiety, depression) and IPV perpetration among males and females in their meta-analysis.

In their review of factors predicting male-perpetrated wife assault, Hilton and Harris (2005) reported that, despite separation and/or no-contact orders, many wife assaulters recidivate against the same victim. Almond et al. (2017) also found that women were more likely to be nonviolently reassaulted by their abusers if they “constantly texted, called, followed, stalked, or harassed” them (p. 68). These findings generally suggest that IPV offenders who have access to their previous or current intimate partners are more likely to recidivate.

In Bennett Cattaneo and Goodman’s (2005) review, 4 of the 16 studies that examined unemployment found that male offenders who were unemployed or employed only part-time had higher rates of IPV reoffending. Male unemployment was also predictive of repeat male-to-female IPV in a five-year longitudinal study on a U.S. sample of adult couples (Caetano et al., 2005). In summary, these factors map onto the DRAOR Acute risk factors.

Most of the research on protective factors and IPV examines victimization rather than perpetration (e.g., Sonis & Langer, 2008). Studies on protective factors in the context of IPV perpetration have measured strengths that are purportedly inverted risk factors as opposed to strengths that exist independently of risk. Being employed—versus being unemployed, a dynamic risk factor—has been found to mitigate repeat IPV perpetration, for instance (Fanslow & Gulliver, 2015). Items in the DRAOR Protective domain are not well represented in the IPV literature as it concerns perpetration specifically. Previous research, however, found that social support, closeness with others, and relationships significantly reduced the likelihood of violence (Abidin et al., 2013; Ullrich & Coid, 2011). These factors map onto the DRAOR Protective social support item.

The Current Study

The DRAOR is currently used statewide in Iowa as a routine case management tool for moderate to high risk community-based adult offenders. Its validity has been tested with samples of general, high risk, and sex offenders; however, the instrument has yet to be validated for use with IPV offenders. Hence, it is important to determine whether the DRAOR can be effectively applied to this population. Examining the utility of the DRAOR in this context will potentially support its use in the development of case plans and intervention strategies for IPV offenders. The current study therefore seeks to validate the DRAOR for use with partner-violent men by addressing three research questions:

Do DRAOR scores differ between IPV and non-IPV offenders? Hypothesis 1 (H₁): As IPV cases would be more homogenous than non-IPV cases with respect to their criminogenic needs, they were expected to score consistently higher on the following IPV-relevant Acute items: substance abuse, anger/hostility, opportunity/access to victims, negative mood, employment, and interpersonal relationships. Given this, we also expected IPV offenders to score higher on Acute subscale. Lastly, IPV offenders were expected to score higher on the Stable items impulse control and attachment with others, as well as lower on the social support item in the Protective domain. No predictions regarding the DRAOR Total, Stable or Protective domains, or the remaining items, were made given the exploratory nature of these analyses. The two groups were also compared on sociodemographic variables, offence-related variables, and baseline risk levels.

Do DRAOR scores accurately predict general, IPV, and violent recidivism, as well as technical violations among IPV offenders? H₂: The DRAOR Total was expected to predict all community outcomes and the Acute subscale was expected to predict IPV recidivism. No predictions regarding the other subscales were made.

Do DRAOR items accurately predict IPV recidivism among IPV offenders? H₃: The Stable items impulse control and attachment with others were expected to predict IPV recidivism, as well as the Acute items substance abuse, anger/hostility, opportunity/access to victims, negative mood, employment, and interpersonal relationships. The Protective item social support was expected to predict IPV desistance. No predictions regarding the other items were made.

Method

Participants

This dataset (N = 510) is from the pilot implementation of the DRAOR in the state of Iowa and has been used in a previous validation study (for additional information, refer to Serin et al., 2020). The sample includes male IPV (n = 112) and non-IPV (n = 398) offenders serving community supervision orders between May 8, 2006, and November 16, 2015. Sociodemographic and offence-related variables are shown in Table 1 (note that the analyses on group differences are discussed in the “Results” section). The majority of IPV and non-IPV offenders identified as White and non-Hispanic, most were single, and the highest level of education for most was high school. Individuals in both groups were most commonly being supervised on probation orders and fewer IPV offenders were convicted of a felony versus misdemeanor offence than non-IPV offenders. In terms of convicting offence type, more IPV offenders were convicted of violent and property crimes, whereas more non-IPV offenders were convicted of drug and public order crimes. Information on victims’ gender was unavailable, as was the nature of the offender-victim relationship at the time of the offence. The average age at the time of the initial DRAOR assessment was 34.6 (SD = 10.6) for IPV offenders and 31.7 (SD = 12.5) for non-IPV offenders.

Table 1.

Differences in Sociodemographic and Offence Variables Between IPV and non-IPV Offenders.

	IPV	Non-IPV	Difference
Variable	n (%)	n (%)	AUC [95% CI]
Race^a			.60 [.53, .66]
White	63 (56.25)	299 (75.31)
Black	49 (43.75)	98 (24.69)
Ethnicity^b			.50 [.44, .56]
Hispanic	4 (3.60)	15 (3.77)
Non-Hispanic	107 (96.40)	383 (97.23)
Marital status^c			.52 [.45, .58]
Single	77 (77.00)	274 (80.35)
Married	23 (23.00)	67 (19.65)
Highest education level^d			.51 [.45, .56]
Less than high school	24 (22.02)	102 (27.06)
High school diploma/GED	82 (75.23)	236 (62.60)
Post-secondary education	3 (2.75)	39 (10.34)
Supervision status			.52 [.47, .58]
Pretrial release	1 (.89)	22 (5.53)
Probation	94 (83.93)	319 (80.15)
Work release	10 (8.93)	34 (8.54)
Parole	7 (6.25)	23 (5.78)
Convicting offence severity^e			.60 [.54, .66]
Misdemeanor	57 (51.35)	123 (31.70)
Felony	54 (48.65)	265 (68.30)
Convicting offence type^f			.55 [.49, .61]
Public order	14 (12.61)	81 (20.45)
Property	23 (20.72)	68 (17.17)
Drug	18 (16.22)	83 (20.96)
Violent	56 (50.45)	164 (41.41)

Note. Unknown cases were excluded from percentages and AUC analyses.

Significant AUCs are bolded.

IPV = Intimate partner violence; AUC = Area under the curve statistic; CI = Confidence interval.

^aOne Asian/Pacific Islander non-IPV offender was removed from percentages and analyses to facilitate comparisons across race.

^bMissing one IPV case.

^cMissing 12 IPV cases and 57 non-IPV cases.

^dMissing three IPV cases and 21 non-IPV cases.

^eMissing one IPV case and 10 non-IPV cases.

^fThose with “other” offence types were excluded from analyses (n = 1 for IPV offenders and n = 2 for non-IPV offenders).

Measures

DRAOR.

The DRAOR is a structured case management scale comprised of six Stable, seven Acute, and six Protective items. The Stable domain addresses criminal orientation and impulsivity concerns, the Acute domain addresses destabilizers and lifestyle stressors, and the Protective domain addresses social support and prosocial identity changes. All items are rated on a 3-point scale (0–2; not a problem to definite problem for risk factors, and not an asset to definite asset for strengths). Stable and Protective domain scores range from 0 to 12 and the Acute domain score ranges from 0 to 14. DRAOR Total scores are calculated by summing the Stable and Acute domains and then subtracting the Protective domain. Scale items are also examined to provide a summary of potential targets for intervention. The DRAOR’s theoretical underpinnings are rooted in an integrated model of offender re-entry that presents a life-course perspective, incorporating crime acquisition and desistance factors (for a depiction and a more detailed description of this model, refer to Serin et al., 2010).

Iowa Risk Assessment.

The Iowa Risk Assessment (IRA) is a 13-item actuarial instrument developed by Iowa Department of Corrections (IDOC), is comprised predominately of static risk factors, and is used in routine correctional assessments in Iowa. The IRA has been utilized to predict multiple community outcomes, including new charges and technical violations (Serin et al., 2020). Items include age at classification, age at first adult conviction/juvenile adjudication, prior juvenile commitments, prior probation/parole supervisions, number of prior probation/parole revocations, felony/misdemeanor convictions, misdemeanor conviction history, sex, alcohol usage problems, drug usage problems, number of address changes in the last 12 months, companions, and employment. The IRA is used in the current study as a baseline measure of risk.

Procedure

Community supervision officers with the IDOC who volunteered for the pilot study received a one-day, in-class training session delivered by the developer of the DRAOR. Officers began scoring the DRAOR once they received the training and they completed IRA forms after initial face-to-face contact with offenders. The IRA assessment completed closest to the date of the initial DRAOR evaluation was retained for use in this study. On average, IRA assessments were completed within approximately five months of DRAOR assessments. Completed assessments were recorded in the form of file notes and entered in the Iowa Corrections Offender Network (ICON). While assessments most proximal to recidivism have the highest predictive accuracy (Lloyd et al., 2020), the current dataset contains initial assessments only, hence analyses are based on these DRAOR ratings. Initial assessments were completed between March 2011, and July 2011. IDOC staff retrieved demographic information, DRAOR scores, and state-level criminal records once assessments were completed. Criminal records were used to identify previous, index, and recidivistic IPV convictions and determine offender typology; thus, IPV offenders were those with any prior, index, and/or repeat IPV convictions in the state of Iowa. The earliest IPV conviction dated back to 1990.

Previous and index IPV offences were coded by comparing offenders’ supervision start date to their IPV sentence date. Convictions that occurred before the supervision start date were coded as either history or index offences depending on the type of supervision. For example, convictions that occurred before the supervision start date were coded as previous offences for probationers but were coded as either previous or index offences for parolees, depending on the length of time between sentencing and supervision dates. Convictions that occurred on the same day as the supervision start date were coded as index offences. Offences and subsequent convictions that occurred after the supervision start date were coded as recidivism. Follow-up time was calculated as the time between the offender’s initial DRAOR assessment and the first date of a subsequent new general conviction, IPV conviction, violent conviction, or technical violation, or November 16, 2015, the study end-date.

Outcome Data

For new convictions (i.e., general recidivism), both the presence (yes/no) and type (public order, property, drug, violent, other) of conviction were recorded. In the case of multiple convictions occurring on the same date, the most serious was recorded. IPV and violent recidivism were recorded as dichotomous variables (yes/no). The former includes domestic assault only, while the latter includes assault (domestic and nondomestic), sex offences, child endangerment, harassment, interference with official acts, weapons-related offences, and robbery. Technical violations are the mildest form of recidivism, whereby an offender breaches a condition of their community supervision (i.e., failure to comply with any supervision order). If multiple violations were recorded throughout the follow-up period, the earliest event was retained. All violations were also compiled to indicate the total number of instances an offender violated their conditions. Offence severity was assessed based on whether the first new conviction was a felony or misdemeanor.

The follow-up time ranged from 0 days to approximately 4.5 years. The average follow-up time for IPV offenders was 34.7 months (SD = 21.2) for general recidivism, 45.8 months (SD = 16.7) for IPV recidivism, 42.5 months (SD = 18.5) for violent recidivism, and 21.9 months (SD = 23.7) for technical violations. During this time, 53% of individuals were convicted of a general offence, 23% were convicted of an IPV offence, 36% were convicted of a violent offence, and 71% incurred a technical violation. The average follow-up time for non-IPV offenders was 42.3 months (SD = 18.2) for general recidivism, 51.6 months (SD = 9.3) for violent recidivism, and 23.6 months (SD = 23.9) for technical violations. During this time, 34% of individuals were convicted of a general offence, 7% were convicted of a violent offence, and 63% incurred a technical violation. An average of 8.3 (SD = 11.4) and 7.5 (SD = 10.0) violations were recorded throughout each offender’s supervision period for IPV and non-IPV offenders, respectively.

Data Cleaning

Initial data cleaning was completed by the third author for the original evaluation of the pilot dataset (N = 562; Serin et al., 2020). The author examined all variables for data entry errors and the presence of missing data, which are not a concern for the current study. There was an insufficient number of female IPV offenders to facilitate comparisons across gender, thus female IPV (n = 5) and female non-IPV offenders (n = 47) were removed from the pilot dataset for the current study, resulting in a total of 112 IPV and 398 non-IPV male offenders.

Overview of Analyses

Differences between IPV and non-IPV offenders.

In addition to descriptive statistics, analyses include area under the curve (AUC) statistics from receiver operating characteristic (ROC) curve analyses and 95% confidence intervals (CIs). The AUC is an effect size statistic that is appropriate when one variable is dichotomous and the other is dichotomous, ordinal, or interval (Swets et al., 2000). AUCs are most commonly used to examine relationships to recidivism, but they can be used for other types of dichotomous variables as well. Recent research, for example, has used AUCs to examine the probability that a randomly selected Indigenous offender will have a higher risk score than a randomly selected non-Indigenous offender (Perley-Robertson et al., 2018).

AUCs for the current study were used in this context to predict offender typology (i.e., to determine if DRAOR scores were associated with being an IPV offender) and can be interpreted as the probability that a randomly selected IPV offender will have a higher risk or protective score than a randomly selected non-IPV offender. AUC values can vary between 0 and 1, with .50 indicating no difference in offender typology. For the DRAOR Total, and the Stable and Acute domains, AUCs above .50 indicate that IPV offenders scored higher than non-IPV offenders, while AUCs below .50 indicate that IPV offenders scored lower. The opposite is true for the Protective domain (higher scores represent lower risk, thus AUCs were reversed to ease the interpretation of results). AUCs of .56, .64, and .71 reflect small, moderate, and large effect sizes, respectively, as these values correspond to Cohen’s ds of .20, .50, and .80 (Rice & Harris, 2005). An AUC value is statistically significant if the 95% CI does not include .50. Cohen’s ds were also calculated to measure the magnitude of differences between groups (refer to the interpretation standards above; Cohen, 1992). AUCs were also used to examine differences in sociodemographic variables, offence-related variables, and baseline risk measured by the IRA.

Predictive validity.

AUCs and 95% CIs were used to predict 4 community outcomes: general, IPV, and violent recidivism, as well as technical violations.

Results

Differences Between IPV and Non-IPV Offenders

There were no differences between groups on ethnicity, marital status, highest education level, supervision status, or convicting offence type. However, IPV offenders were more likely than non-IPV offenders to be classified as Black versus White and to be convicted of a misdemeanor versus felony offence (AUCs = .60, small effects; refer to Table 1). When examining differences on the DRAOR, IPV offenders were rated as higher risk on the Total, domains, and individual items (except for the responsive to advice item in the Protective domain; refer to Table 2). Significant comparisons include the Stable item attachment with others (AUC = .57 and d = 0.26, small effects), the Acute domain (AUC = 0.65 and d = 0.52, moderate effects), Acute items substance abuse (AUC = 0.57 and d = 0.27, small effects), anger/hostility (AUC = 0.72 and d = 0.92, large effects), opportunity/access to victims (AUC = 0.59 and d = 0.39, small effects), and interpersonal relationships (AUC = 0.60 and d = 0.39, small effects), as well as the DRAOR Total (AUC = 0.58 and d = 0.32, small effects). IPV offenders also scored significantly higher on the IRA (AUC = 0.59 and d = 0.32, small effects).

Table 2.

Differences in DRAOR Scores Between IPV and Non-IPV Offenders.

	IPV^a	Non-IPV^b	Difference
Risk/Protective Factor	M (SD)	M (SD)	AUC [95% CI]	d
DRAOR stable	5.19 (2.73)	4.67 (2.65)	.55 [.49, .61]	.19
Peer associations	1.03 (0.58)	0.96 (0.54)	.53 [.47, .59]	.12
Attitudes towards authority	0.69 (0.63)	0.65 (0.61)	.51 [.45, .57]	.06
Impulse control	1.04 (0.68)	0.91 (0.64)	.55 [.49, .61]	.20
Problem solving	0.96 (0.57)	0.88 (0.60)	.54 [.48, .60]	.14
Sense of entitlement	0.68 (0.67)	0.63 (0.61)	.52 [.45, .58]	.08
Attachment with others	0.79 (0.59)	0.64 (0.55)	.57 [.51, .63]	.26
DRAOR acute	6.57 (3.00)	5.07 (2.73)	.65 [.59, .71]	52
Substance abuse	0.96 (0.83)	0.75 (0.73)	.57 [.51, .63]	.27
Anger/hostility	1.09 (0.74)	0.48 (0.58)	.72 [.66, .77]	.92
Opportunity/access to victims	0.90 (0.76)	0.63 (0.60)	.59 [.53, .66]	.39
Negative mood	0.71 (0.69)	0.68 (0.64)	.51 [.45, .57]	.05
Employment	0.93 (0.89)	0.86 (0.83)	.52 [.46, .58]	.08
Interpersonal relationships	1.11 (0.68)	0.86 (0.60)	.60 [.54, .66]	.39
Living situation	0.88 (0.66)	0.81 (0.64)	.53 [.47, .59]	.11
DRAOR protective	6.43 (2.72)	6.76 (2.84)	.53 [.47, .59]	.12
Responsive to advice	1.17 (0.61)	1.17 (0.55)	.50 [.44, .56]	.00
Prosocial identity	1.04 (0.50)	1.10 (0.55)	.53 [.47, .59]	.11
High expectations	1.12 (0.64)	1.16 (0.65)	.52 [.46, .58]	.06
Costs/benefits	1.05 (0.55)	1.17 (0.60)	.55 [.49, .61]	.21
Social support	1.14 (0.64)	1.17 (0.62)	.51 [.45, .57]	.05
Social control	0.91 (0.61)	1.00 (0.60)	.54 [.48, .60]	.15
DRAOR total	5.33 (7.40)	2.98 (7.37)	.58 [.52, .64]	.32
IRA	14.13 (5.40)	12.21 (6.59)	.59 [.53, .65]	.32

Note. Significant AUCs are bolded.

DRAOR = Dynamic Risk Assessment for Offender Re-entry; IPV = Intimate partner violence; AUC = Area under the curve statistic; CI = Confidence interval; d = Cohen’s d; IRA = Iowa Risk Assessment.

^aN = 112.

^bN = 398.

Predictive Validity

Predictive validity results for IPV offenders (n = 112) across the four community outcomes are presented in Table 3. The DRAOR Acute domain significantly predicted general recidivism (AUC = .62, a small effect), with higher scores associated with an increased likelihood of recidivism. None of the other DRAOR measures predicted general recidivism. In terms of IPV recidivism, neither the DRAOR Total nor the domains produced significant AUCs. The DRAOR Total and Stable scores were significantly associated with higher rates of violent recidivism (AUCs = .65, moderate effects), whereas Protective scores were significantly associated with lower rates of violent recidivism (AUC = .65, a moderate effect). The Acute subscale did not predict violent recidivism. The DRAOR Total, Stable, and Acute significantly predicted technical violations (AUCs = .65, moderate effects), with higher scores associated with offenders’ increased likelihood of violating their conditions. Conversely, higher scores on the DRAOR Protective domain were significantly associated with offenders’ decreased likelihood of violating their conditions (AUC = .62, a small effect). Apart from the sense of entitlement item from the Stable domain (AUC = .64, a moderate effect), none of the items significantly predicted IPV recidivism.

Table 3.

Predictive Validity of the DRAOR (AUCs) Across Outcomes for IPV Sample.

	AUC [95% CI]
Risk/Protective Measure	General Recidivism	IPV Recidivism	Violent Recidivism	Technical Violations
Stable	.60 [.49, .70]	.62 [.50, .73]	.65 [.54, .75]	.65 [.54, .76]
Acute	.62 [.51, .72]	.55 [.43, .68]	.60 [.49, .71]	.65 [.54, .76]
Protective	.59 [.48, .69]	.58 [.45, .71]	.65 [.54, .76]	.62 [.51, .73]
Total	.61 [.50, .71]	.60 [.48, .72]	.65 [.55, .76]	.65 [.54, .76]

Note. Significant AUCs are bolded.

DRAOR = Dynamic Risk Assessment for Offender Re-entry; AUC = Area under the curve statistic; IPV = Intimate partner violence; CI = Confidence interval.

Discussion

The current study investigated the predictive validity of the DRAOR for IPV offenders to determine whether it can be used to inform case planning and intervention strategies. This research is the first to examine the DRAOR in this context, providing information regarding its generalizability to IPV cases.

Differences Between IPV and Non-IPV Offenders

Preliminary analyses revealed that IPV offenders scored significantly higher than non-IPV offenders on the following items: attachment with others, substance abuse, anger/hostility, opportunity/access to victims, and interpersonal relationships. This partially supports our hypothesis and is consistent with previous research on areas that are particularly problematic for IPV offenders (e.g., Bennett Cattaneo & Goodman, 2005; Birkley & Eckhardt, 2015; Hilton & Harris, 2005). One caveat to these findings, however, is that effect sizes were generally small in magnitude (except for the anger/hostility item). Further, the lower CI limits for attachment with others and substance abuse were .51, meaning the effects only just reached statistical significance. Caution is therefore warranted when interpreting these results.

In addition to their higher item scores, IPV offenders received higher Acute and Total scores than non-IPV offenders, indicating they pose greater acute and overall risk. Contrary to our expectations, IPV and non-IPV offenders did not differ on the Stable item impulse control, the Acute items negative mood or employment, or the Protective item social support. Recent research found that negative affect (angry, irritable, hostile, frustrated, mad, upset, sad, depressed, and/or anxious) predicted IPV perpetration among male university students (Shorey et al., 2015); however, few studies have examined this relationship in offender samples and those that have are dated (e.g., Jacobson et al., 1996; Palmer et al., 1992). The null results for negative mood in the current study may indicate that male offenders are more predisposed to display, report, and/or express anger than affective states such as depression or anxiety. It may also be that supervision officers are less inclined to inquire about negative affective states, especially those who adopt a more surveillance than change agent approach to supervision. The link between unemployment and IPV recidivism has been more commonly studied and supported (e.g., Caetano et al., 2005; Edwards, 2015), but research is still mixed (Bennett Cattaneo & Goodman, 2005). Despite the null findings in the current study, the literature suggests that negative mood and unemployment may be important treatment targets for IPV offenders. Finally, while research on social support as a protective factor for repeat IPV perpetration is lacking, this relationship might be worth exploring given the link between social support and violence desistance (Ullrich & Coid, 2011).

In terms of sociodemographic factors, IPV offenders were more likely than non-IPV offenders to identify as Black versus White. Future research should therefore examine whether certain risk factors are more salient for one group than the other. IPV offenders also scored higher on the IRA than non-IPV offenders. This might suggest that IPV offenders received higher ratings on more DRAOR items because they had higher baseline risk levels. However, out of 4 overlapping items on the 2 scales (alcohol/drug abuse [substance abuse], address changes [living situation], companions [peer association], and employment), IPV offenders only scored significantly higher on DRAOR substance abuse, which is relevant for IPV risk. If IPV offenders scored higher on more DRAOR items simply because they had higher baseline risk levels, we would at least expect them to receive higher ratings on the DRAOR items that are also reflected in the baseline risk measure. As this was not the case, it seems more likely that group differences were driven by the nature of their crimes rather than baseline risk. For instance, IPV offenders were more likely to be convicted of misdemeanor versus felony offences, paradoxically suggesting their index crimes were less serious.

Predictive Validity

None of the DRAOR measures predicted IPV recidivism among IPV offenders and only the Acute domain predicted general recidivism. Results were more promising for the other community outcomes, with the Total score and all but the Acute domain predicting violent recidivism, and the DRAOR Total score and all domains predicting technical violations (note that the Protective domain predicted desistance from these outcomes). These findings were partially consistent with our hypothesis. In terms of effect size magnitude, the DRAOR Total predicted violent recidivism and technical violations with moderate strength. Similar AUCs have been reported for other case management tools, such as the Ohio Risk Assessment System (ORAS; Latessa et al., 2010). Namely, AUCs for predicting re-arrest during a one-year follow-up period ranged from moderate to large across its six versions (AUCs = .65 to .71; Latessa et al., 2010). Therefore, the use of the DRAOR as a risk/case management tool for violence and technical violations still seems warranted despite the lower AUCs.

A more detailed exploration of the DRAOR’s predictive validity for IPV recidivism showed that only the sense of entitlement item in the Stable domain was a significant predictor. It is important to note that this finding is consistent with what would be expected by chance. Specifically, the Type 1 error rate is .05, which means that 5 out of every 100 analyses (i.e., roughly 1 in 20) will be significant due to chance. The one DRAOR item (out of 19) that significantly predicted IPV recidivism will therefore not be interpreted as meaningful.

The DRAOR appears to be particularly useful for predicting technical violations among IPV offenders, but not for predicting their general or IPV recidivism. The DRAOR was also an accurate predictor of violent recidivism among IPV offenders, but unexpectedly, the Acute domain did not emerge as a significant predictor in this context. This is surprising because many of the Acute items are identified in the IPV literature as significant predictors of partner abuse. Given that all IPV offences are also classified as violent offences, the Acute items should theoretically have predicted violent recidivism. However, this DRAOR domain was the only one that failed to do so. Certain Acute items may therefore be important predictors of violence in intimate partner relationships, but not violence in non-intimate partner relationships. This hypothesis is unlikely, though, as anger/hostility, opportunity/access to victims, and negative mood (risk factors identified in the literature as IPV correlates) predicted violence against non-intimate partners in a study on 4,116 violent offenders (Lowenkamp et al., 2016). A more plausible explanation for these null findings is that there was insufficient statistical power in the current study due to the small sample size and low occurrence of new violence convictions.

The literature has identified a multitude of risk factors that are relevant for IPV cases (e.g., marital conflict, substance abuse, victim access; see the “Introduction” for an overview of this research). A recent review on risk assessments for repeat partner abuse also found that predictive accuracy was higher for IPV-specific tools than general tools (Svalin & Levander, 2020). Despite the growing body of research on IPV risk factors, however, there is limited research on whether strengths apply differently to IPV versus non-IPV cases. Namely, characteristics that promote desistance (e.g., stable employment, prosocial identity, sobriety, maturation; Maruna, 2010) have primarily been studied in general offenders rather than violent offenders. The extent to which strengths are differentially related to desistance for IPV versus non-IPV offenders therefore remains unclear.

Strengths and Limitations

One important strength of the current study is that all supervision officers received formal training on the DRAOR, giving confidence that the scale was implemented with fidelity. This is important because poor implementation fidelity has been shown to reduce predictive accuracy (Chadwick, 2014; Flores et al., 2006). The use of field data is a second strength in the methodology of this research; field studies have high ecological validity, meaning that results can be generalized to real-life settings. However, several studies have found weaker predictive performance for assessment instruments in field versus lab settings (Edens & Boccaccini, 2017), thus the generally low predictive accuracy of the DRAOR in the current study is perhaps not unduly surprising. The small sample of IPV offenders may have also contributed to these results. Namely, Vergouwe et al. (2005) recommend 100 events and 100 non-events for adequate statistical power in binary logistic regression, which likely applies to AUCs as well (L. M. Helmus, personal communication, January 25, 2018), and the current sample included only 112 offenders.

Beyond limitations related to the nature of the data and the small sample used in this research, some important measurement issues should be acknowledged. First, IPV offenders may have been underrepresented if any of their IPV offences were undetected, plea bargained down, or occurred outside of Iowa. The former measurement issues are inherent to all risk assessment research, however, and it is not uncommon for studies of this nature to be restricted to state-level data or its equivalent (e.g., Gerth et al., 2017; McRee Lauch et al., 2017; Renauer & Henning, 2005). In terms of sample characteristics, victims’ unknown gender identity represents a second limitation; risk and criminogenic needs may vary greatly among IPV offenders in homosexual versus heterosexual relationships (Messinger, 2014; Whitehead et al., 2020). It is also important to know the nature of the offender-victim relationship to determine if this predicts reoffending, however, these data were unavailable as well. A fourth measurement issue is the use of initial versus proximal DRAOR scores. Dynamic risk and protective factors show changes in risk across time. Scores that are proximal in time to recidivism should therefore be more closely related to repeat offences than distal scores. Lloyd et al.’s (2020) research supported this notion, which could account for the generally nonsignificant predictive validity findings in the present study. Unfortunately, in this study, the use of the most proximal DRAOR assessment would have further reduced sample size.

The unknown inter-rater reliability of the DRAOR reflects an additional shortcoming of the current research, as well as past research, given that the scale’s inter-rater reliability has yet to be examined. Future research should evaluate this psychometric property of the DRAOR to determine if its dynamic risk and protective factors can be reliably assessed across evaluators. In addition to the unknown inter-rater reliability of the DRAOR, the extent to which case management correlated with risk level is unknown. Higher risk offenders were likely supervised more closely, which has important implications for predictive accuracy. Namely, reducing offender risk through treatment or intervention would decrease the potential association between high baseline DRAOR scores and recidivism. The results of the current study must also be interpreted with caution given that AUCs do not account for time at risk (Helmus & Babchishin, 2017) and this study used variable follow-up periods. It is important to note, however, that a more in-depth examination of the current data found similar predictive accuracy results for the DRAOR Total and domains using Cox regression survival analyses, which account for time at risk (Perley-Robertson, 2018). Lastly, IPV offenders in this study were compared to those convicted of a range of non-IPV crimes (59% had a non-violent index conviction). This is problematic in that group differences could be related to violent tendencies in general rather than partner violence specifically. To remedy this, future research should compare partner-violent men with violent men who targeted stranger/non-partner victims.

Conclusions and Recommendations

Risk factors that emerged as potentially important treatment targets for IPV offenders were attachment with others, substance abuse, anger/hostility, opportunity/access to victims, and interpersonal relationships. Supervision officers should also focus their attention on IPV offenders’ Acute needs in general; these destabilizers and lifestyle stressors were particularly problematic for IPV versus non-IPV offenders. Despite the relatively poor predictive validity results in the current study, the presence of risk-relevant factors for IPV in the DRAOR (e.g., substance abuse, anger, victim access) suggests it could potentially be used to guide case management when paired with a specialized IPV measure that does not focus on offender change, such as the Ontario Domestic Assault Risk Assessment (ODARA).

The ODARA appears to be one of the most promising IPV risk scales, with AUCs for IPV recidivism ranging from .66 to .72 in recent research (Jung & Buro, 2017; Olver & Jung, 2017). The ODARA, however, focuses on static risk factors, thus cannot be used for ongoing case management purposes. Holtzworth-Munroe and Stuart’s (1994) batterer typology work also suggests that general measures of risk may not be as suitable for family only offenders (i.e., those who do not typically engage in violence outside the family); however, if the IPV offender is classified as generally violent/antisocial, the DRAOR may be more appropriate as these individuals are more criminally diverse, reflecting Central 8 risk factors (Andrews & Bonta, 2010). The DRAOR may therefore be useful as a supplement to static tools like the ODARA, or if used with IPV offenders who are identified as generally violent/antisocial versus family only.

Supplemental Material

Supplemental Material - Using a General Case Management Tool With Partner-violent Men on Community Supervision in Iowa

Supplemental Material for Using a General Case Management Tool With Partner-violent Men on Community Supervision in Iowa by Bronwen Perley-Robertson, Ralph C. Serin, Nick Chadwick, in Journal of Interpersonal Violence

Footnotes

Authors’ Note

The dataset used in this project was also used in previous works (Perley-Robertson et al., 2018; Serin et al., 2020).

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Ralph C. Serin is the author and holds the copyright of the Dynamic Risk Assessment for Offender Re-entry (DRAOR) tool.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the Social Sciences and Humanities Research Council of Canada.

ORCID iD

Bronwen Perley-Robertson

Author Biographies

Bronwen Perley-Robertson, MA, is a PhD student in Forensic Psychology at Carleton University. Her research focuses on risk assessment in community contexts and the predictive accuracy of risk tools. She has explored these topics in specialized populations, including Indigenous and intimate partner violence offenders, as well as general offenders. Her other research evaluates the role strength-based factors play in promoting crime desistance.

Ralph C. Serin, PhD, is professor of Forensic Psychology and Director of the Criminal Justice Decision-Making Laboratory at Carleton University. Current research topics include correctional programming (evaluation, measurement of offender change); risk assessment (dynamic risk, sentinel events); crime desistance (trajectories, protective factors); violent offenders (acute dynamic risks, typologies, management); and parole and case management decision-making (strategies, competencies, and factors influencing accuracy).

Nick Chadwick, PhD, is a recent graduate of the Criminal Justice Psychology Program at Carleton University in Ottawa, Canada. He has conducted research on the use and implementation of evidence-based practices in community supervision, the utility of assessing changes in dynamic risk and protective factors in the prediction of recidivism, and effective correctional programming.

References

Abidin

, Davoren

, Naughton

, Gibbons

, Nulty

, & Kennedy

H. G.

(2013). Susceptibility (risk and protective) factors for in-patient violence and self-harm: Prospective study of structured professional judgement instruments START and SAPROF, DUNDRUM-3 and DUNDRUM-4 in forensic mental health services. BMC Psychiatry , 13, Article 197. https://doi.org/10.1186/1471-244X-13-197

Aldarondo

, & Sugarman

D. B.

(1996). Risk marker analysis of the cessation and persistence of wife assault. Journal of Consulting and Clinical Psychology , 64(5), 1010–1019. https://doi.org/10.1037/0022-006X.64.5.1010

Almond

, McManus

, Brian

, & Merrington

D. P.

(2017). Exploration of the risk factors contained within the UK’s existing domestic abuse risk assessment tool (DASH): Do these risk factors have individual predictive validity regarding recidivism? Journal of Aggression, Conflict and Peace Research , 9(1), 58–68. https://doi.org/10.1108/JACPR-01-2016-0211

Anderson

J. C.

, Stockman

J. K.

, Sabri

, Campbell

D. W.

, & Campbell

J. C.

(2015). Injury outcomes in African American and African Caribbean women: The role of intimate partner violence. Journal of Emergency Nursing , 41(1), 36–42. https://doi.org/10.1016/j.jen.2014.01.015

Andrews

D. A.

, & Bonta

(2010). The psychology of criminal conduct (5th ed.). Anderson Publishing Co.

Bennett Cattaneo

, & Goodman

L. A.

(2005). Risk factors for reabuse in intimate partner violence: A cross-disciplinary critical review. Trauma, Violence, & Abuse , 6 (2), 141–175. https://doi.org/10.1177/1524838005275088

Birkley

E. L.

, & Eckhardt

C. I.

(2015). Anger, hostility, internalizing negative emotions, and intimate partner violence perpetration: A meta-analytic review. Clinical Psychology Review , 37, 40–56. https://doi.org/10.1016/j.cpr.2015.01.002

Brem

M. J.

, Florimbio

A. R.

, Elmquist

, Shorey

R. C.

, & Stuart

G. L.

(2018). Antisocial traits, distress tolerance, and alcohol problems as predictors of intimate partner violence in men arrested for domestic violence. Psychology of Violence , 8(1), 132–139. https://doi.org/10.1037/vio0000088

Caetano

, McGrath

, Ramisetty-Mikler

, & Field

C. A.

(2005). Drinking, alcohol problems and the five-year recurrence and incidence of male to female and female to male partner violence. Alcoholism: Clinical and Experimental Research , 29(1), 98–106. https://doi.org/10.1097/01.ALC.0000150015.84381.63

10.

Campbell

, Snow Jones

, Dienneman

, Kub

, Schollenberger

, O’Campo

, Gielen

A. C.

, & Wynn

(2002). Intimate partner violence and physical health consequences. Archives of Internal Medicine , 162(10), 1157–1163. https://www.ncbi.nlm.nih.gov/pubmed/12020187

11.

Chadwick

(2014). Validating the Dynamic Risk Assessment for Offender Re-entry (DRAOR) in a sample of U.S. probationers and parolees [Master’s thesis, Carleton University]. Carleton University Research Virtual Environment. https://curve.carleton.ca/f43f2b17-42cf-4792-8ddc-351682dee8bf

12.

Cohen

(1992). A power primer. Psychological Bulletin , 112(1), 155–159. https://doi.org/10.1037/0033-2909.112.1.155

13.

de Vries Robbé

, de Vogel

, Douglas

K. S.

, & Nijman

H. L. I.

(2015). Changes in dynamic risk and protective factors for violence during inpatient forensic psychiatric treatment: Predicting reductions in postdischarge community recidivism. Law and Human Behavior , 39(1), 53–61. https://doi.org/10.1037/lhb0000089

14.

Edens

J. F.

, & Boccaccini

M. T.

(2017). Taking forensic mental health assessment “out of the lab” and into “the real world:” Introduction to the special issue on the field utility of forensic assessment instruments and procedures. Psychological Assessment , 29(6), 599–610. https://doi.org/10.1037/pas0000475

15.

Edwards

K. M.

(2015). Intimate partner violence and the rural–urban–suburban divide: Myth or reality? A critical review of the literature. Trauma, Violence, & Abuse , 16(3), 359–373. https://doi.org/10.1177/1524838014557289

16.

Fanslow

, & Gulliver

(2015). Exploring risk and protective factors for recent and past intimate partner violence against New Zealand women. Violence and Victims , 30(6), 960–983. https://doi.org/10.1891/0886-6708.VV-D-14-00010

17.

Flores

A. W.

, Lowenkamp

C. T.

, Holsinger

A. M.

, & Latessa

E. J.

(2006). Predicting outcome with the level of service inventory-revised: The importance of implementation integrity. Journal of Criminal Justice , 34(5), 523–529. https://doi.org/10.1016/j.jcrimjus.2006.09.007

18.

Gerth

, Rossegger

, Bauch

, & Endrass

(2017). Assessing the discrimination and calibration of the Ontario Domestic Assault Risk Assessment in Switzerland. Partner Abuse , 8(2), 168–189. https://doi.org/10.1891/1946-6560.8.2.168

19.

Haines

, Khan

, Noblett

, Omodunbi

, Sadiq

, Zaman

, & Whittington

(2018). Assessing protective factors for violence risk in U.K. general mental health services using the Structured Assessment of Protective Factors. International Journal of Offender Therapy and Comparative Criminology , 62(1), 3965–3983. https://doi.org/10.1177/0306624X17749449

20.

Hanson

R. K.

, & Harris

A. J. R.

(2000). Where should we intervene?: Dynamic predictors of sexual offense recidivism. Criminal Justice and Behavior , 27(1), 6–35. https://doi.org/10.1177/0093854800027001002

21.

Helmus

L. M.

, & Babchishin

K. M.

(2017). Primer on risk assessment and the statistics used to evaluate its accuracy. Criminal Justice and Behavior , 44, 8–25. https://doi.org/10.1177/0093854816678898

22.

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. https://doi.org/10.1177/1524838004272463

23.

Holtzworth-Munroe

, & Stuart

G. L.

(1994). Typologies of male batterers: Three subtypes and the differences among them. Psychological Bulletin , 116(3), 476–497. https://doi.org/10.1037/0033-2909.116.3.476

24.

Jacobson

N. S.

, Gottman

J. M.

, Gortner

, Berns

, & Shortt

J. W.

(1996). Psychological factors in the longitudinal course of battering: When do the couples split up? when does the abuse decrease? Violence and Victims , 11(4), 371–392. https://www.ncbi.nlm.nih.gov/pubmed/9210278

25.

Jung

, & Buro

(2017). Appraising risk for intimate partner violence in a police context. Criminal Justice and Behavior , 44(2), 240–260. https://doi.org/10.1177/0093854816667974

26.

LaMotte

A. D.

, Meis

L. A.

, Winters

J. J.

, Barry

R. A.

, & Murphy

C. M.

(2018). Relationship problems among men in treatment for engaging in intimate partner violence. Journal of Family Violence , 33(1), 75–82. https://doi.org/10.1007/s10896-017-9920-9

27.

Latessa

E. J.

, Lemke

, Makarios

, Smith

, & Lowenkamp

C. T.

(2010). The creation and validation of the Ohio Risk Assessment System (ORAS). Federal Probation , 74(1), 16–22. https://www.uscourts.gov/sites/default/files/74_1_2_0.pdf

28.

Leone

R. M.

, Crane

C. A.

, Parrott

D. J.

, & Eckhardt

C. I.

(2016). Problematic drinking, impulsivity, and physical IPV perpetration: A dyadic analysis. Psychology of Addictive Behaviors , 30(3), 356–366. https://doi.org/10.1037/adb0000159

29.

Lloyd

C. D.

, Hanson

R. K.

, Richards

D. K.

, & Serin

R. C.

(2020). Reassessment improves prediction of criminal recidivism: A prospective study of 3,421 individuals in New Zealand. Psychological Assessment . https://dx-doi-org.web.bisu.edu.cn/10.1037/pas0000813

30.

Lowenkamp

C. T.

, Johnson

J. L.

, Trevino

, & Serin

R. C.

(2016). Enhancing community supervision through the application of dynamic risk assessment. Federal Probation , 80(2), 16–20. http://www.uscourts.gov/sites/default/files/80_2_3_0.pdf

31.

Maruna

(2010). Understanding desistance from crime [Unpublished report]. National Offender Management Service.

32.

Max

, Rice

D. P.

, Finkelstein

, Bardwell

R. A.

, & Leadbetter

(2004). The economic toll of intimate partner violence against women in the United States. Violence and Victims , 19(3), 259–272. https://doi.org/10.1891/vivi.19.3.259.65767

33.

McRee Lauch

, Hart

K. J.

, & Bresler

(2017). Predictors of treatment completion and recidivism among intimate partner violence offenders. Journal of Aggression, Maltreatment & Trauma , 26(5), 543–557. https://doi.org/10.1080/10926771.2017.1299824

34.

Messinger

A. M.

(2014). Marking 35 years of research on same-sex intimate partner violence: Lessons and new directions. In D.

Peterson

& Panfil

(Eds.), Handbook of LGBT communities, crime, and justice (pp. 65–85). Springer Science+Business Media. https://doi.org/10.1007/978-1-4614-9188-0_4

35.

Norlander

, & Eckhardt

(2005). Anger, hostility, and male perpetrators of intimate partner violence: A meta-analytic review. Clinical Psychology Review , 25(2), 119–152. https://doi.org/10.1016/j.cpr.2004.10.001

36.

Olver

M. E.

, & Jung

(2017). Incremental prediction of intimate partner violence: An examination of three risk measures. Law and Human Behavior , 41(5), 440–453. https://doi.org/10.1037/lhb0000251

37.

Olver

M. E.

, Wong

S. C. P.

, Nicholaichuk

, & Gordon

(2007). The validity and reliability of the Violence Risk Scale-Sexual Offender Version: Assessing sex offender risk and evaluating therapeutic chance. Psychological Assessment , 19(3), 318–329. https://doi.org/10.1037/1040-3590.19.3.318

38.

Palmer

S. E.

, Brown

R. A.

, & Barrera

M. E.

(1992). Group treatment program for abusive husbands: Long-term evaluation. American Journal of Orthopsychiatry , 62(2), 276–283. https://doi.org/10.1037/h0079336

39.

Penney

S. R.

, Marshall

L. A.

, & Simpson

A. I. F.

(2016). The assessment of dynamic risk among forensic psychiatric patients transitioning to the community. Law and Human Behavior , 40(4), 374–386. https://doi.org/10.1037/lhb0000183

40.

Perley-Robertson

(2018). Predictive validity of the Dynamic Risk Assessment for Offender Re-Entry among intimate partner violence offenders (Master’s thesis). Carleton University, Carleton University Research Virtual Environment. https://curve.carleton.ca/71561148-616b-4e95-bec3-e9b83301be3c

41.

Perley-Robertson

, Helmus

L. M.

, & Forth

(2018). Predictive accuracy of static risk factors for Canadian Indigenous offenders compared to non-Indigenous offenders: Implications for risk assessment scales. Psychology, Crime & Law , 25(3), 248–278. https://doi.org/10.1080/1068316X.2018.1519827

42.

Persson

, Belfrage

, Fredriksson

, & Kristiansson

(2017). Violence during imprisonment, forensic psychiatric care, and probation: Correlations and predictive validity of the risk assessment instruments COVR, LSI-R, HCR-20^V3, and SAPROF. The International Journal of Forensic Mental Health , 16(2), 117–129. https://doi.org/10.1080/14999013.2016.1266420

43.

Renauer

, & Henning

(2005). Investigating intersections between gender and intimate partner violence recidivism. Journal of Offender Rehabilitation , 41(4), 99–124. https://doi.org/10.1300/J076v41n04_05

44.

Rice

M. E.

, & Harris

G. T.

(2005). Comparing effect sizes in follow-up studies: ROC area, Cohen’s d, and r. Law and Human Behavior , 29(5), 615–620. https://doi.org/10.1007/s10979-005-6832-7

45.

Serin

R. C.

(2007, 2015, 2017). The Dynamic Risk Assessment Scale for Offender Re-entry (DRAOR): Guidelines for case planning and risk management using structured assessment of dynamic risk and protective factors [Unpublished user-manual].

46.

Serin

R. C.

, Chadwick

, & Prell

(2020). Dynamic risk and protective factors [Unpublished manuscript].

47.

Serin

R. C.

, Lloyd

, & Hanby

L. J.

(2010). Enhancing offender re-entry: An integrated model for understanding offender re-entry. European Journal of Probation , 2(2), 53–75. https://doi.org/10.1177/206622031000200205

48.

Shorey

R. C.

, Brasfield

, Febres

, & Stuart

G. L.

(2011). The association between impulsivity, trait anger, and the perpetration of intimate partner and general violence among women arrested for domestic violence. Journal of Interpersonal Violence , 26(13), 2681–2697. https://doi.org/10.1177/0886260510388289

49.

Shorey

R. C.

, McNulty

J. K.

, Moore

T. M.

, & Stuart

G. L.

(2015). Emotion regulation moderates the association between proximal negative affect and intimate partner violence perpetration. Prevention Science , 16(6), 873–880. https://doi.org/10.1007/s11121-015-0568-5

50.

Smith

S. G.

, Chen

, Basile

K. C.

, Gilbert

L. K.

, Merrick

M. T.

, Patel

, Walling

, & Jain

(2017). The National Intimate Partner and Sexual Violence Survey (NISVS) : 2010–2012 state report. Centers for Disease Control and Prevention. https://www.cdc.gov/violenceprevention/pdf/NISVS-StateReportBook.pdf

51.

Sonis

, & Langer

(2008). Risk and protective factors for recurrent intimate partner violence in a cohort of low-income inner-city women. Journal of Family Violence , 23(7), 529–538. https://doi.org/10.1007/s10896-008-9158-7

52.

Svalin

, & Levander

(2020). The predictive validity of intimate partner violence risk assessments conducted by practitioners in different settings: A review of the literature. Journal of Police and Criminal Psychology , 35(2), 115–130. https://doi.org/10.1007/s11896-019-09343-4

53.

Swets

J. A.

, Dawes

R. M.

, & Monahan

(2000). Psychological science can improve diagnostic decisions. Psychological Science in the Public Interest , 1(1), 1–26. https://doi.org/10.1111/1529-1006.001

54.

Ullrich

, & Coid

(2011). Protective factors for violence among released prisoners: Effects over time and interactions with static risk. Journal of Consulting and Clinical Psychology , 79(3), 381–390. https://doi.org/10.1037/a0023613

55.

Vergouwe

, Steyerberg

E. W.

, Eijkemans

M. J. C.

, & Habbema

J. D. F.

(2005). Substantial effective sample sizes were required for external validation studies of predictive logistic regression models. Journal of Clinical Epidemiology , 58(5), 475–483. https://doi.org/10.1016/j.jclinepi.2004.06.017

56.

Whitehead

, Dawson

, & Hotton

(2020). Same-sex intimate partner violence in Canada: Prevalence, characteristics, and types of incidents reported to police services. Journal of Interpersonal Violence . https://doi.org/10.1177/0886260519897342

57.

World Health Organization. (2017). Intimate partner and sexual violence against women: Fact sheet . http://www.who.int/mediacentre/factsheets/fs239/en/

58.

World Health Organization & Pan American Health Organization. (2012). Intimate partner violence: Understanding and addressing violence against women (Report No. WHO/RHR/12.36). http://apps.who.int/iris/bitstream/10665/77432/1/WHO_RHR_12.36_eng.pdf

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB