This article was downloaded by: [Tilburg University]On: 10 April 2015, At: 22:25Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Click for updates

Journal of Forensic Psychology PracticePublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/wfpp20

Monitoring Risk Behaviors by ManagingSocial Support in the Network of aForensic Psychiatric Patient: A Single-Case AnalysisLydia ter Haar-Pomp MScabc, Carlijn de Beer MScd, Rosalind van derLem PhDe, Marinus Spreen PhDab & Stefan Bogaerts PhDcf

a School of Social Work and Arts Therapies, Stenden University,Leeuwarden, The Netherlandsb Forensisch Psychiatrisch Centrum Dr. S. van Mesdag, Groningen,The Netherlandsc Faculty of Social and Behavioral Sciences, Tilburg University,Tilburg, The Netherlandsd Forensic Psychiatric Ambulant Policlinic het Dok, Dordrecht, TheNetherlandse Forensic Psychiatric Ambulant Policlinic het Dok, Rotterdam, TheNetherlandsf Forensisch Psychiatrisch Centrum De Kijvelanden/het Dok,Rotterdam, The NetherlandsPublished online: 09 Apr 2015.

To cite this article: Lydia ter Haar-Pomp MSc, Carlijn de Beer MSc, Rosalind van der Lem PhD, MarinusSpreen PhD & Stefan Bogaerts PhD (2015) Monitoring Risk Behaviors by Managing Social Support inthe Network of a Forensic Psychiatric Patient: A Single-Case Analysis, Journal of Forensic PsychologyPractice, 15:2, 114-137, DOI: 10.1080/15228932.2015.1007779

To link to this article: http://dx.doi.org/10.1080/15228932.2015.1007779

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,

proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Journal of Forensic Psychology Practice, 15:114–137, 2015Copyright © Taylor & Francis Group, LLCISSN: 1522-8932 print/1522-9092 onlineDOI: 10.1080/15228932.2015.1007779

Monitoring Risk Behaviors by Managing SocialSupport in the Network of a Forensic

Psychiatric Patient: A Single-Case Analysis

LYDIA TER HAAR-POMP, MScSchool of Social Work and Arts Therapies, Stenden University, Leeuwarden, The Netherlands,

Forensisch Psychiatrisch Centrum Dr. S. van Mesdag, Groningen, The Netherlands,and Faculty of Social and Behavioral Sciences, Tilburg University, Tilburg, The Netherlands

CARLIJN DE BEER, MScForensic Psychiatric Ambulant Policlinic het Dok, Dordrecht, The Netherlands

ROSALIND VAN DER LEM, PhDForensic Psychiatric Ambulant Policlinic het Dok, Rotterdam, The Netherlands

MARINUS SPREEN, PhDForensisch Psychiatrisch Centrum Dr. S. van Mesdag, Groningen, The Netherlands,

and School of Social Work and Arts Therapies, Stenden University, Leeuwarden,The Netherlands

STEFAN BOGAERTS, PhDFaculty of Social and Behavioral Sciences, Tilburg University, Tilburg, The Netherlands,

and Forensisch Psychiatrisch Centrum De Kijvelanden/het Dok, Rotterdam,The Netherlands

This prospective case study examines changes over time in the socialsupport network of a forensic psychiatric patient diagnosed withattention deficit hyperactivity disorder (ADHD). The focus is on thefunctional and dysfunctional influences of the patient’s social sup-port dynamics on his risk behavior during mandatory policlinictreatment. A structured Forensic Social Network Analysis interviewwas conducted with the patient and two of his network membersat four time points in his treatment process. The patient’s socialsupporters, their structural network position, and their risk arepooled and labeled through a triad census method. The numberof practical and emotional supporters decreased over time in the

Address correspondence to Lydia ter Haar-Pomp, School of Social Work and ArtsTherapies, Stenden University, P.O. Box 1298, 8900 CG Leeuwarden, The Netherlands. E-mail:lydia.pomp@stenden.com

114

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 115

network of the patient, which also resulted in a decrease of high-riskpractical and emotional supporters. The size and composition offinancial support in the network of the patient remained almost sta-ble. Monitoring and analyzing social support using the triad censusmethod provides valuable insights for individual risk managementpurposes.

KEYWORDS social support, forensic psychiatry, risk manage-ment, triad census method, forensic attention deficit hyperactivitydisorder (ADHD), case study

Social support is an important concept in social network studies and isdescribed as positive mutual social interactions between people that meetbasic needs such as affection, acceptance, and security (Kunst, Winkel, &Bogaerts, 2010). It is common to distinguish several types of social sup-port such as instrumental, practical, informational, or emotional support(Hlebec & Kogovsek, 2013). A key component of social support is thesense of belonging, feelings of acceptance, and being appreciated by others(Gottlieb, 2000; Lindgren, 1990). Social support appears in intimate relationsor relations in which there is a sense of mutual trust; the reciprocity ofsocial support is essential for the support provision (Gleason, Lida, Bolger,& Shrout, 2003; Liang, Krause, & Bennett, 2001; Whittaker, 1992).

Social support has physical and mental health benefits (Hirsch, 1980;Holt-Lunstad, Smith, & Layton, 2010; Kumar & Browne, 2008; Robinson &Garber, 1995; Veiel & Bauman, 1992; Vilhjalmsson, 1994; Warren, Stein, &Grella, 2007). Social support can help individuals to cope with adverse lifeevents (stressors), such as illness, work stress, job loss, death of a loved one,and so forth (Bogaerts, Vanheule, & Desmet, 2006; Cobb, 1976; Cullen, 1994).

Social support is also known for its risk-reducing effect on criminalbehavior. A considerable number of empirical studies in criminology andforensic psychiatry have demonstrated that increased levels of social supportcan result in a decrease of criminal behavior (Colvin, Cullen, & Vander Ven,2002; Cullen, 1994; Nakhaie & Sacco, 2009; Vance, Bowen, Fernandez, &Thompson, 2002). Social support providers can have critical roles in sup-porting or discouraging violent behavior (Andrews & Bonta, 1994; Estroff& Zimmer, 1994; McCarthy & Hagan, 1995; Monahan, 1981; Monahan,Steadman, & Silver, 2001; Shapiro & diZegera, 2010; Warr, 2002). Protectivesocial support is related to self-control and therefore, a lower probability ofgetting involved in antisocial behavior (Cullen, Wright, & Chamlin, 1999).

Several theoretical explanations are applicable to the relationshipbetween social support and criminal behavior. Agnew (1992) argued inhis general strain theory that the presence of conventional social supportmakes it more likely for people to cope with strains in conventional ways,thereby decreasing the likelihood of deviant coping. The social exchange

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

116 L. ter Haar-Pomp et al.

theory highlights the importance of social support for achieving individ-uals’ (criminological) goals (Blau, 1964; Cook & Whitmeyer, 1992). Socialcontrol theories concentrate on the controlling function of social support.In that respect, these theories see delinquency and crime as the result ofweak social bonds (Hirschi, 1969, 1977).The social capital theory (Bourdieu,1981) is based on the social resource hypothesis. The stronger a personis embedded in social capital that provides him with plenty of resources,the better opportunity a person has to achieve personal goals. These goalsare structured by social production function (SPF) theory (Lindenberg, 1996;Ormel, Lindenberg, Steverink, & Verbrugge, 1999). The SPF theory targetstwo important goals; Physical well-being and social well-being. Humans aimto optimize these two goals and use five means to achieve this: stimulation,comfort, status, behavioral confirmation, and affection. The individual goalscome along with both positive and negative behavior. An example of posi-tive behavior is working out with a friend in order to optimize physical andsocial well-being. An example of negative behavior is crime that is commit-ted to fit into or impress a social criminal-oriented group in order to optimizesocial well-being.

Because of the benefits of social support on physical and mental healthand behavioral outcomes, social support can have an added value in forensicpsychiatric practice. Forensic psychiatry focuses on the assessment and treat-ment of mentally disordered offenders. In forensic psychiatric risk assessmentliterature, social support is often defined as a dynamic risk management fac-tor (Douglas, Hart, Webster, & Belfrage, 2013; Schuringa, Spreen, & Bogaerts,2014). Dynamic risk factors are (forensic) behaviors that are treatable andchangeable and therefore feasible as treatment goals for interventions tominimize, monitor, and control (future) risk. For instance, in the risk assess-ment tool, the Historical Clinical Risk Management-20, Version 3 (HCR-20V3),“Future problems with personal support” is defined as one of the 20 key riskfactors (Douglas et al., 2013).

Although social support is a well-established dynamic risk factor, stud-ies that examine the effect of social support on criminal behavior areinconsistent and range from an adverse to a risk-reducing effect. On theone hand, the availability of social resources, such as stable friendshipsand sufficient social support, provides a buffer against criminal behavior(Odonne-Paolucci, Violato, & Schofield, 2000; Resnick, Ireland, & Borowsky,2004; Sampson & Laub, 1990; Surjadi, van Horn, Bogaerts, & Bullens, 2010;Vance et al., 2002). On the other hand, risk-increasing effects are found whennegative social support is provided (Bolger, Zuckerman, & Kessler, 2000;Buschman et al., 2010; Coyne, Wortman, & Lehman, 1988), for instance, ifsocial supporters lack the skills or insight to be helpful. In forensic psychiatricsettings, this includes, for example, going to a bar with a person with alcoholdependence or advising a patient to not take his prescribed medications.

Despite the buffering effect of social support on the risky behavior offorensic psychiatric patients, there is still limited understanding about how

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 117

social support and the construction of a positive and risk-reducing socialenvironment can help forensic psychiatric patients to prevent and reducecriminal behavior (Monahan et al., 2001). In this article, we describe a caseof an attention deficit hyperactivity disorder (ADHD) male patient who wasmonitored for 10 weeks during mandatory forensic psychiatric care. We focusspecifically on the developments of social support among his network mem-bers in relation to risk management issues. First of all, we will introduce themethod of forensic social network analysis.

FORENSIC SOCIAL NETWORK ANALYSIS

To examine an individual’s social network, the technique of forensic socialnetwork analysis (FSNA) has been developed to disclose and monitorforensic psychiatric patients’ personal social networks (see Pomp, Spreen,Bogaerts, & Völker, 2010). A personal network can be defined as anego-centered network, and consists of a focal actor (ego), ego’s direct rela-tionships (ties) with his network members (alters), and ties between ego’salters (Wasserman & Faust, 1994).

The FSNA method uses a three-way approach to sample information andto analyze changes in personal networks (Bem & Funder, 1978; Monahan,1981). In step 1, the most important network members in relationship to thecrime context are disclosed. Step 2 concentrates on network members thatare important in the current context and in the near future. Step 3 is thecomparison between the current network with the past criminal network.The goal is to analyze and interpret similarities and differences betweenthe different networks in terms of the patient’s specific risk behaviors.The theoretical foundation of FSNA is driven by sociological, criminologi-cal, and psychological theories. The Risk-Need-Responsivity (RNR) model ofAndrews, Bonta, and Hoge (1990) is an important model that is integratedin the FSNA approach. The social network analysis provides insight into aperson’s major risk and need factors according to the patient and some ofhis relevant network members, such as antisocial and procriminal attitudes,values, and beliefs, antisocial peers and isolation from anticriminal others,substance abuse, family/marital relationships, prosocial recreational activi-ties, feelings of personal distress, and physical and mental health (Bonta &Andrews, 2010). After the assessment of the patient’s risk profile and thedisclosure of the role of his social network, possible intervention methodscan be truly judged whether they match with that patient’s risk profile.

Before starting FSNA analysis, the patient and a limited number of sig-nificant persons in the patient’s social network should be interviewed. Forrisk management purposes, the opinions of professionals must be involved.Involving informal network members in risk management procedures hasthe advantage that the behavior of a person in different social contextscan be observed in an informal manner and, where possible, be adjusted

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

118 L. ter Haar-Pomp et al.

by informal or professional network members. Informal network membersare the eyes and ears of professionals (Shapiro & diZegera, 2010). Informalnetwork members can also be used as observers by assessing and monitor-ing whether a patient adequately applies learned skills in an uncontrolledenvironment (Spreen & Pomp, 2009). Another benefit is that informal net-work members may provide collateral information about the patient, becausepatients may withhold relevant forensic information when speaking the truthcould have an adverse effect on, for example, the length of probation.In the FSNA method, informal network members who can disclose essen-tial information about individual risk behaviors of the patient are selected.An important aspect is that the FSNA professional, and not the patient,decides which informal network members are invited for an interview.Network members need to be selected with care, keeping in mind whatis important to know about the specific patient and his social network fac-tors. Informal network members are selected based on their roles, networkposition, and influence on the patient.

In this study, a single case of an adult forensic psychiatric patient isused to illustrate and explore which social support can be used for riskmanagement purposes. We examined social support in relation to positiveand negative behavioral outcomes in order to improve the quality of thepatient’s risk management. We expected that close protective relationshipswith friends, family, or other support groups will reduce the probabilityof criminal behavior. The single case involved an adult man in mandatoryoutpatient treatment. The FSNA procedures consisted of (a) a face-to-facesemistructured interview in which the patient was asked to disclose his socialnetwork during the period of the offense and at the time of the interview,and (b) a face-to-face semistructured interview in which selected social net-works members are asked to add their view about important social networkinformation during the period of the offense and at the time of the interview.After finishing the data collection, the patient’s, and his network members’,data were analyzed with the goal of defining supportive and nonsupportivesocial support issues in the patient’s personal network. The last step of thiscase study comprised reviewing with the patient and his network membersthe FSNA outcomes, with the purpose of reaching agreement on the patient’streatment and risk management plan.

THE PRESENT STUDY

The Institutional Setting

The case study was conducted in the forensic psychiatric outpatient andday treatment center, het Dok, situated in Rotterdam, the Netherlands. Thiscenter has an ADHD unit for forensic patients with ADHD and comorbidpsychiatric disorders. In recent years, dropout and less treatment progress

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 119

were identified as an issue for forensic patients with ADHD. The ADHD-unit het Dok provides forensic psychiatric care on a mandatory or voluntarybasis. The treatment consists of four steps: (a) registration: patients canbe registered by referrers including probation officers, general practition-ers, youth care workers, mental care authorities (in Dutch: GeestelijkeGezondheidsZorg [GGZ]), courts of justice, community health services (inDutch: Gemeentelijke GezondheidsDiensten [GGD]), psychiatrists and otherspecialists; (b) intake: the patient is invited for an intake within threeweeks. During intake, offense characteristics, psychopathology, and staticand dynamic risk factors are inventoried by a psychiatrist, a clinical psy-chologist, and a social worker; (c) advisory meeting: in the presence of thepatient, treatment is discussed and indicated; and (d) treatment: the treat-ment consists of individual therapy or group therapy. Individuals from thedirect environment of the patient may also be involved in the treatment.

ADHD is characterized by a pattern of behaviors present in multiplesettings (e.g., school and home), that can result in performance issues insocial, educational, or work settings (American Psychiatric Association, 2013).ADHD symptoms are divided into two categories of inattention and hyper-activity (American Psychiatric Association, 2013). Epidemiological studiesassessed the prevalence of ADHD in adult forensic psychiatry at 25% (e.g.,Kooij, 2009). The overall prevalence of ADHD in the adult population isabout six times lower (around 4.5%) than in adult forensic samples (Henrichs& Bogaerts, 2012; Kessler et al., 2006). The overall ADHD prevalence rate inadult inmates was about 10.5% (Cahill et al., 2012).

Adults with ADHD are often confronted with negative outcomes on arange of long-term life skills. For example, they act impulsively and expe-rience concentration and planning problems (May & Bos, 2000). They areoften looking for tense situations and take risks in order to improve theirconcentration, often use drugs and alcohol, and are restless by nature. Thesetypical characteristics of ADHD can lead to problematic social and intimaterelationships. Therefore, it is important to examine the personal networks ofadults with ADHD with regard to their functional and dysfunctional influ-ences on their individual risk behavior. It is known that individuals withADHD are more likely to misinterpret activities of others and tend to respondinappropriately (Kooij, 2009). Maintaining relationships across the lifespancan be extremely difficult for persons with ADHD. For example, Toner,O’Donoghue, and Houghton (2006) showed that individuals with ADHDhave more marital problems and higher rates of divorce.

For the present case study, a forensic psychiatric male patient wasselected from the FSNA pilot study. In this pilot study, ADHD forensic psy-chiatric male patients from het Dok were interviewed to examine the impactof their personal social support informal networks on their risk behaviorsduring the period of the offense, treatment, and after treatment. This selectedpatient, whose pseudonym is Peter, was the first patient who participated in

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

120 L. ter Haar-Pomp et al.

the study. Taking part in the study was voluntary. Peter agreed, by signingan informed consent form, that his anonymized personal social network datacould be used for scientific research and a scientific report/paper.

METHOD

Patient

Peter is a 31-year-old man who was indicated for cognitive behavior treat-ment in the ADHD-unit. Peter was convicted of attempted murder of hismother. He was sentenced by the court to treatment because of aggressionproblems and poor impulse control during interpersonal conflicts betweenhim and his relatives. The reason for the attempted murder of his mother wasa conflict between the two. In a moment of anger, Peter put his hands aroundher neck. During the intake phase, Peter was diagnosed with ADHD, bor-derline personality disorder, and cannabis dependence. Peter was assessedas having an average intelligence level according to the Wechsler AdultIntelligence Scale (WAIS)-III. Peter also had financial, work, and housingproblems.

Treatment

The focus of treatment was to improve Peter’s impulse control to stop hisaggressive outbursts. Peter had individual therapy sessions with a psycholo-gist to assess the scope and causes of his aggressive behavior. They discussedPeter’s adherence to treatment and medication. Furthermore, Peter receivedsupport for his psychosocial problems. Peter’s girlfriend and his mother wereinvolved in the treatment. They were asked to evaluate the severity and fre-quency of Peter’s aggressive behaviors. They discussed with Peter and hissocial worker deescalation strategies.

As part of the treatment, Peter had started taking ADHD medication;effective medication is a vital aspect of treatment of ADHD (Weiss & Weiss,2004). Without medication, patients can insufficiently profit from the psycho-logical treatment because their attention and concentration fall short; oftenthis leads to treatment dropout.

MEASURES

The Forensic Social Network Analysis Interview

The FSNA interview for forensic psychiatric patients during mandatory poly-clinic treatment entails a face-to-face semistructured interview in which thepatient and network members are asked to describe Peter’s social networkduring the time of the offense and at the time of the interview. During Peter’s

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 121

treatment, three FSNA interviews were conducted at three different times.The first measurement was a retrospective face-to-face FSNA interview withPeter and two of his network members, namely his mother and girlfriend.In this interview, Peter, his girlfriend, and his mother were asked to dis-close social network information during the period of the offense and atthe time of the interview. The second and third interviews were plannedfour and eight weeks later to monitor changes in the network of Peter.The third FSNA interview appointment was canceled by Peter a numberof times and eventually took place after six weeks instead of the prear-ranged four weeks. The interviews were conducted by a trained FSNA socialworker.

FSNA uses data triangulation. The purpose of using different dataresources is the possible unreliable information from the patient. Inaccurateinformation could have an adverse effect on the length and intensity oftreatment. In the FSNA method, this problem is addressed by combiningdata from the patient administrative file, the patient interview, and thenetwork members’ interviews (multiple informant data). The narrative elec-tronic patient file (EPF) of Peter was used to collect personal and forensicpsychiatric characteristics. In general, EPF improves the quality of dossierdocumentation and viewing, recording of prescriptions, messaging, andorganization of accumulated patient data (Miller & Sim, 2004). Researchconfirms the benefits of EPF with regard to quality of care coordination, addi-tional decision support, and patient satisfaction (Zhou et al., 2009). Criminalrecords were used to gain insight into Peter’s criminal history.

To analyze the extent to which social relationships in Peter’s per-sonal network changed during treatment, the following FSNA variableswere collected during the interviews with Peter and his two network mem-bers: network size (the total number of network members), network roles(family/friends, others, victims, and co-offenders), and the total number ofnetwork members who gave social support (0 = no, 1= yes). Categoriesof social support were practical (domestic help), emotional (seeking advicefrom), and financial (borrowing money from others).

The forensic characteristics of Peter’s network members were listed:criminal record, psychiatric problems, drugs, alcohol, financial problems,and problematic lifestyle (e.g., housing problems, conflicts with others, etc.).Peter was requested to give his perception about the nature of the rela-tionships between the network members. The information was scored indichotomous terms of contact and no contact.

Statistical Analysis

To illustrate changes in Peter’s network configurations over time, the visu-alization tool NetDraw was used (Borgatti, Everett, & Freeman, 2002). Datawere analyzed using SPSS 20.0.

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

122 L. ter Haar-Pomp et al.

FIGURE 1 Open Triad Versus Closed Triad.

Structural characteristics of Peter’s personal network, such as the pat-terns of direct relations between the patient and his network members andthe relations between his network members, were examined employing thetriad census method of Kalish and Robins (2006). They introduced a methodof classifying personal networks using a census of triads of different types.In this study, a triad is defined as the relation of the forensic psychiatricpatient with two of his network members (called alters) and the relationshipbetween these two alters. The distribution of the proportion of all types oftriads expresses the personal network structures (Kalish & Robins, 2006).We defined the existence of a tie between two network members as the per-ceived existence of a contact between both persons (the patient has ties withboth network members). This resulted in two possible triads: no tie betweenalters (open triad) or a tie between alters (closed triad). Figure 1 shows thetwo triads.

In this case study, the triads were defined on three variables: (a) theexistence of a tie between two network members in a triad was defined asthe existence of a contact between these persons (Peter had ties with bothnetwork members), (b) whether a network member had one or more riskfactors including criminal record, psychiatric support, drug use, alcoholism,and a problematic way of life, and (c) whether Peter had listed a networkmember as a social supporter (practical, emotional, and financial). Figure 2presents the resulting triad census classified by 20 triads. Network memberswithout forensic risk factors are shown as white; network members withforensic risk factors are shown as grey (these are people who have a criminalrecord, psychiatric support, drug use, alcoholism, or problematic way oflife).

Each type of triad has its own risk management interpretation in relationto potential risks. In general, the risk interpretation of a triad depends on thepatient’s individual risk factors. For instance, receiving social support fromdrugs users is more risky for a patient with a drug problem compared to apatient without drug problems (Monahan, 1981). The collective influence ofa network on an individual results not only from direct bonds between theindividual and each network member, but also from relationships between

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 123

Patient

1 2 3 4 5

6 7 8 9 10

11 12 13 14 15

16 17 18 19 20

Network member providing social support - one or more risk factors

Network member providing social support - no risk factors

Network member does not provide social support - one or more risk factors

Network member does not provide social support - no risk factors

FIGURE 2 Twenty FSNA Triads: Social Support (Low- and High-Risk Network Members).

network members (Milardo, 1986). From that perspective, triad 20, with tiesbetween high-risk social supporters, is defined as the most risky network forPeter. Triad 13 is the most desirable for him, because the two social support-ers have a collective protective influence. In relation to Peter’s individual riskmanagement, the most risky triads consist of social supporters with similar ormore severe forensic and psychiatric problems. This includes network mem-bers with aggression problems, poor impulse control (ADHD), personalitydisorders (borderline), cannabis dependence, and financial, work, and hous-ing problems. The influence of a social supporter on Peter’s behavior maydiffer between the three types of social support. For instance, seeking advicefrom a criminal friend may have a different impact on Peter’s behavior thanborrowing money from the same criminal friend. The triads can be labeledhigh-risk, low-risk, and protective to provide the professional guidelines tointerpret network changes in terms of reducing risks. In Peter’s case, fivetriads are labeled as high-risk:

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

124 L. ter Haar-Pomp et al.

Two high-risk social supporters are connected (triad 20). Two high-risk social supporters are not connected (triad 10). One high-risk social supporter is connected with a high-risk network

member who does not provide social support (triad 19). One high-risk social supporter is not connected with the other network

member who does not provide social support (triads 5 and 9).

Three triads are labeled as protective:

Both social supporters have no risk factors and are connected (triad 13). Both social supporters have no risk factors and are not connected (triad 3). One social supporter without risk factors is connected with a low-risk

network member who does not provide social support (triad 12).

The other triads are labeled as low-risk, because we expected they have alimited influence on Peter’s risk behavior. These triads are 1, 2, 4, 6, 7, 8, 11,14, 15, 16, 17, and 18. It is important to note that some of these triads caneasily change in status in being protective or high-risk. For instance, triad18 consists of network members with high-risk factors, but these persons donot provide social support. If they become social support providers, theirlow-risk label would change to that of high-risk.

As an indicator of change, Spearman’s rho was calculated between theproportional distributions of the triad census of all combinations of the foursocial network configurations (crime, at the start of therapy, 4 weeks after,and 10 weeks after starting treatment).

RESULTS

The purpose of the case study was to examine whether the distributionof the different supporting and nonsupporting triads changes within thepersonal network of Peter between the four points of measurement.A triad census was compiled based on the 20 types of supporting andnonsupporting triads, and transformed into a vector of triad proportions toallow comparisons between the four time points. Spearman’s Rho was usedto analyze change within the vector of the triads’ proportions. For each timepoint, the 20 triads were ranked from low to high based on the proportionpresent. A coefficient of plus 1 implies that the frequency distribution ofthe triad types had not changed. A coefficient of 1 is given when nothingchanged between two time points. A negative coefficient indicates that triadtypes with a high proportion at time point 1 have a low proportion at timepoint 2, that is, there has been a change between the two measurements.A coefficient of minus 1 means that the frequency distribution is reversed

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 125

between two time points (the highest-ranked triad at time point 1 is thelowest-ranked triad at time point 2). Finally, a coefficient close to 0 impliesthat there is a change, but rather random.

Changes in Network Size, Network Roles, and Risk Factors

The total persons within Peter’s network varied from 8 at time point 1, to9 at time point 2, to 10 at time point 3, and 9 at time point 4.

At the moment of the offense, Peter’s informal network consisted ofthree family members (mother, sister, and stepfather), one girlfriend, threefriends, and one neighbor. Focusing on the risk factors of the networkmembers, his mother suffered from psychiatric problems, his girlfriend hadfinancial problems, the neighbor had alcohol problems, and his three friendshad criminal records and drug problems. At time point 2, the patient addedhis father to the informal network whose members suffered from psychiatricproblems. At time point 3, the patient added his job coach (no risk factors)to his network. At time point 4, the contact with the job coach had ended.It is important to note that the victim of Peter’s offense, his mother, was partof all four network configurations.

Changes in Peter’s Practical Support Configuration

Figure 3 shows the four practical support networks.Figure 3 shows that the number of individuals that Peter could ask

for practical support increased from seven network members at time point1 to eight network members at time point 2. In Peter’s offense period, five ofseven practical supporters had risk factors; at time point 2, (during treatment)six of eight practical supporters had risk factors. The network at time point3 shows an important change: only two of the eight practical supporters attime point 2 remained listed by Peter as practical supporters at time point3—his girlfriend (risk factor: financial problems) and his job coach (no riskfactors). At the end, time point 4, Peter had no practical supporter left.

The changes in the number of practical supporters also affected thedistribution of the 20 triads in the practical support network over time (seeFigure 3). As mentioned in the statistical analysis section, triad 20, with tiesbetween high-risk social supporters in the network, was defined as the mostrisky for Peter (high-risk triad). The proportion of triad 20 decreased from.28 (time point 2) to .00 (time point 3). Focusing on the quality of his socialsupport, this is a positive finding for risk management: strong social ties withhigh-risk network members can have negative effects if risky behavior is pro-moted. Triad 13 was defined as the most desirable for Peter (protective triad),because the two social supporters have a collective protective influence. Theproportion of triad 13 decreased from .04 (time point 1) to .00 (time point

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

126 L. ter Haar-Pomp et al.

Type High-risk triads Protectivetriads

Low-risk triads

number 5 9 10 19 20 3 12 13 1 2 4 6 7 8 11 14 15 16 17 18Timepoint 1

.00 .18 .07 .00 .29 .00 .00 .04 .00 .00 .00 .07 .18 .00 .00 .00 .00 .00 .18 .00

Timepoint 2

.00 .17 .14 .00 .28 .00 .00 .03 .00 .00 .00 .06 .17 .00 .00 .00 .00 .00 .17 .00

Timepoint 3

.00 .04 .00 .09 .00 .00 .00 .00 .00 .04 .18 .13 .02 .24 .02 .09 .04 .00 .00 .09

Timepoint 4

.00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .25 .00 .00 .33 .03 .17 .00 .00 .00 .22

Patient

Patient

Neighbor

Girlfriend

Girlfriend

Mother

Sister

Stepfather

Time point 1: offense period Time point 2: first interviewStepfather

Sister

Father

Mother

Friend1Friend1

Friend2

Friend3

Friend2

Friend3

Patient

Neighbor

Girlfriend

MotherSister

Father

Stepfather

Friend1

Friend2

Friend3

Coach

Patient

Neighbor

Neighbor

Girlfriend

MotherSister

Father

Stepfather

Friend1

Friend2

Friend3

Time point 3: second interview Time point 4: third interview

FIGURE 3 Practical Support During the Four Measurements.

3). Based on these findings, the treatment team needed to discuss with Peterhow (a) to retain the low proportion of high-risk practical supporters in hisnetwork, (b) to deal with the loss of practical supporters, and (c) to establishstrong practical support ties with protective networks.

To statistically summarize the observed changes, between the four net-work configurations, Spearman’s rho was used as an indicator of change.Spearman’s rho revealed a statistically significant relationship between timepoint 1 and time point 2 (ρ = .999, p < .01) and time points 3 and 4 (ρ =.649, p < .01), implying that ranking of the 20 triad types based on the occur-rence of their frequencies between these time points were almost stable.No correlation was found between the other time points: 1–3, 1–4, 2–3, and2–4. The network configurations between these points were independent ofeach other and the distributions of the triads had changed. For example,between time points 2 and 3, important changes in the triad distributionwere estimated. Peter’s practical support network decreased from eight attime point 2 (first FSNA interview) to two at time point 3 (second FSNA

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 127

interview; Figure 3). These observed changes have implications for Peter’srisk management. Focusing on the available quantity of social support, theobserved decrease is negative for his accessibility to social resources.

Changes in Peter’s Emotional Support Configuration

Figure 4 shows the four emotional support networks.Figure 4 shows that the number of individuals Peter would ask for

emotional support increased from six network members at time point 1 toeight network members at time point 2. At the time of his offense, fourof the six emotional supporters had risk factors. At time point 2 (duringtreatment), six of his eight emotional supporters had risk factors. The net-work configurations at time point 3 and 4 showed an important change for

Type High-risk triads Protectivetriads

Low-risk triads

number 5 9 10 19 20 3 12 13 1 2 4 6 7 8 11 14 15 16 17 18 Timepoint 1

.00 .11 .11 .18 .11 .00 .00 .04 .00 .00 .00 .07 .18 .04 .00 .00 .00 .07 .11 .00

Timepoint 2

.00 .22 .14 .00 .22 .00 .00 .03 .00 .00 .00 .03 .19 .00 .00 .00 .00 .03 .14 .00

Timepoint 3

.00 .00 .00 .00 .00 .00 .00 .00 .02 .00 .31 .00 .00 .29 .04 .16 .00 .00 .00 .18

Timepoint 4

.00 .00 .00 .00 `.00 .00 .00 .00 .00 .00 .22 .00 .00 .36 .03 .17 .00 .00 .00 .22

Patient

Neighbor

Girlfriend

Mother

Sister

Stepfather

Time point 1: offense period

Time point 3: second interview Time point 4: third interview

Time point 2: first interview

Friend1

Friend2

Friend3

Patient

NeighborCoach

Girlfriend

MotherSister

Father

Stepfather

Friend1

Friend2

Friend3

Patient

Neighbor

Girlfriend

Mother

Sister

Father

Stepfather

Friend1

Friend2

Friend3

Patient

Neighbor

Father

Girlfriend

MotherSister

Stepfather

Friend1

Friend2

Friend3

FIGURE 4 Emotional Support During the Four Measurements.

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

128 L. ter Haar-Pomp et al.

Peter; no network member remained as an emotional supporter, as depictedby the lack of triangles in Figure 4 at time points 3 and 4. The changesin the distribution of the 20 triads in the emotional support network areshown below Figure 4. On the positive side, the proportion of the high-risk triads 9, 10, 11, and 20 are decreased at time point 3 to .00. On thenegative side, the only available protective triad at time points 1 and 2,namely triad 13, has disappeared at time point 3. This resulted in impor-tant issues for Peter’s risk management: the amount of emotional supportin his network was decreased (negative condition), which also resulted in adecrease of high-risk emotional support triads (positive condition). Froma risk management perspective, it was a negative finding that Peter nolonger received emotional support from his personal network members.Based on these findings, the treatment team would need to discuss withPeter how he could receive emotional support from protective networkmembers.

To summarize, Spearman’s rho revealed a statistically significant rela-tionship between five of the time points: 1–2 (ρ = .790, p < .01), 1–3 (ρ =–.460, p < .05), 2–4 (ρ =-.448 p < .05), and 3–4 (ρ = .934, p < .01). Betweentime points 1–3 and 2–4, negative coefficients were found. This indicates thatthe values of the two time points vary in opposite directions. For example,at time point 1, triad 4 was not included in the triad census; at time point3, triad 4 was counted 14 times (proportion of .31). Another example: timepoint 2 showed for triad 20 a network proportion of .22; at time point 4,triad 20 was no longer present. No correlation was found between timepoints 1 and 4. The network configurations were independent of each otherand the distributions of the triads had changed between time points 1 and 4.

Changes in Peter’s Financial Support Configuration

Figure 5 shows the four financial support networks.Figure 5 shows that the number of individuals Peter would ask for finan-

cial support remained stable, as depicted by three connected social supportproviders at time point 1 and two connected social supports at time points 2,3, and 4. At time point 1, the three financial support providers were girlfriend(risk factor: financial problems), mother (risk factor: psychiatric problems),and stepfather (no risk factors). At time point 2, the mother was not listedany longer as a financial support provider. The general composition of thenetwork triads turned out to be rather stable (see Figure 5). For example,the network proportion of high-risk triad 20 at time points 2, 3, and 4 was.00. The high-risk triads 10 and 19 were present in all four financial supportnetworks. The protective triad 12 was rather stable over time. From a riskmanagement perspective, it was important that the social worker discussedwith Peter and his girlfriend how to deal with financial difficulties, becausethe girlfriend was an important financial supporter over time, but she had

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 129

Type High-risk triads Protectivetriads

Low-risk triads

number 5 9 10 19 20 3 12 13 1 2 4 6 7 8 11 14 15 16 17 18 Timepoint 1

.00 .14 .00 .14 .04 .00 .04 .00 .00 .00 .14 .11 .00 .00 .00 .00 .07 .04 .07 .21

Timepoint 2

.00 .06 .00 .11 .00 .00 .03 .00 .00 .00 .11 .11 .00 .31 .00 .06 .03 .06 .03 .11

Timepoint 3

.02 .04 .00 .07 .00 .00 .02 .00 .02 .02 .24 .07 .00 .24 .00 .04 .02 .04 .02 .11

Timepoint 4

.00 .06 .00 .11 .00 .00 .03 .00 .00 .00 .11 .11 .00 .31 .00 .06 .03 .06 .03 .11

Time point 1: offense period

Time point 3: second interview Time point 4: third interview

Time point 2: first interview

Patient

Neighbor

Girlfriend

Mother

Sister

Stepfather

Friend1

Friend2

Friend3

Patient

Neighbor

Father

Girlfriend

MotherSister

Stepfather

Friend1

Friend2

Friend3

Patient

NeighborCoach

Girlfriend

MotherSister

Father

Stepfather

Friend1

Friend2

Friend3

Patient

Neighbor

Girlfriend

MotherSister

Father

Stepfather

Friend1

Friend2

Friend3

FIGURE 5 Financial Support During the Four Measurements.

similar financial problems to Peter. It is likely that she lacked the appropriateskills herself to be helpful to Peter in the area of financial management.

The stability in financial support is reflected by Spearman’s rho.Statistically significant relationships between all defined time points werefound: 1–2 (ρ = .670, p < .01), 1–3 (ρ = .591, p < .01), 1–4 (ρ = .670, p <

.01), 2–3 (ρ = .929, p < .01), 2–4 (ρ = 1.00, p < .01), and 3–4 (ρ = .929,p < .01).

DISCUSSION

The recognition that social networks play an important role in an offender’srisk management is widely accepted. However, social relationships andpatterns have been only partially taken into account in clinical forensic

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

130 L. ter Haar-Pomp et al.

psychiatric practice of risk management. The FSNA method was introducedas a monitoring method to measure and follow changes over time in pos-itive and negative social support on the individual level. This study is thefirst conducted in an outpatient forensic psychiatric setting. Earlier FSNAresearch focused on social networks of forensic psychiatric patients dur-ing their incarceration in a forensic psychiatric hospital setting (Haar-Pomp,Spreen, Bogaerts, & Volker, 2014). In this study, the social support net-work of a forensic psychiatric patient with ADHD, borderline personalitydisorder, and cannabis dependence was examined and monitored over time(prospective). The FSNA triad census method was introduced in which socialsupporters, their structural network position, and their risk were combinedand labeled. The triads are labeled high-risk, low-risk, and protective, andprovide the professional guidelines to interpret network changes in termsof reducing risks. The overall goal of this study was to assess the role ofnetwork members in supporting or discouraging the patient with regard toliving a crime-free life.

In the present study, social support was measured by first asking theforensic psychiatric patient “Peter” to identify his informal support networkand then asking him to rate his network members in terms of practical, emo-tional, and financial support and his satisfaction with the support received.

This study shows the benefits of interviewing the patient and hisnetwork members repeatedly: important social support dynamics wereuncovered. We found that social network changes for one person over ashort period of time can differ between the three types of social support.Peter’s social network at the time of his offense was characterized by asocial support network with a high proportion of risk triads. The size ofthe practical and emotional support networks were significantly decreasedduring treatment. The number of financial supporters remained almost sta-ble. The triad census method provides better insights in the meaning ofthe observed decrease in social support in the context of individual riskmanagement. For example, a decrease in practical and emotional support(negative condition) resulted in a lower proportion of the high-risk triads inPeter’s social support network (positive condition). Peter’s therapists men-tioned two possible explanations for the decrease in social support fromhis personal network members. First, treatment intensity and contact withhet Dok may have reduced the need for social support from other sources.In other words, there is a shift from informal to formal network contacts thatcan be temporarily positive. However, over time, formal network membersmust be replaced by new, informal network members. Second, Peter hadstarted taking ADHD medication. As a result, Peter showed less impulsivebehavior. Unfortunately, the medication also had negative side effects. Peterexperienced more emptiness and depression. Peter indicated that he neededdistance between himself and his important informal social supporters. If thistriggers a shift toward seeking social support from formal network members,

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 131

it can positively influence the treatment outcomes. It is known that positiverelationships between patients and their mental health care professionals canpositively influence the patient’s motivation, his compliance with the rules,and treatment outcomes (Skeem, Encandela, & Eno Louden, 2003; Skeem,Eno Louden, Manchak, Vidal, & Haddad, 2009). However, the finding that apatient relies less on his personal social network during treatment requiresattention. The patient’s informal social supporter network will be part ofhis life after supervision and must play a significant role in the success oftreatment in the long run (Shapiro & diZegera, 2012). This means that bothformal and informal relationships must be taken into account to understandthe influence of the social support system on the patient’s (risk) behavior.

From a risk management perspective, it is also interesting that the vic-tim of Peter’s offense, his mother, was still part of his personal networkduring treatment. Another important finding was that the network size ofPeter’s total personal network remained almost stable. The stability of thetotal network size showed that not being mentioned as a social supporter ata certain period did not mean that these network members were no longerpart of Peter’s personal network. This raises an interesting question thatfuture research could address: whether network members who disappearedfrom the social support network returned into the social support networkat a later point in treatment or probation. Before a patient can rely on hissocial support system, there should be a sufficient amount of network mem-bers that qualify for giving various kinds of social support. For instance,it is nearly impossible, and certainly undesirable, that a single member ina network should offer all kinds of social support (Walker, Wasserman, &Wellman, 1994).

This study has focused on one forensic psychiatric patient across ashort period (10 weeks). This patient has his own unique risk profile. Otherpatients with different mental health problems or offenses may have otherprofiles with their own relevant social support factors and related triads. Eachcase requires a thorough study of possible positive and negative social sup-port factors. Subsequently, such a study requires time and the expertise ofthe appointed professional. The limited research period for this study makesit hard to establish if the decrease of (high-risk) practical and emotional sup-porters will be permanent. Future research should examine a larger numberof ambulant forensic psychiatric patients with multiple measurements over alonger period of time.

In forensic psychiatry, it is essential to compare social relationships dur-ing treatment with relationships at the time of the crime because patientsmay use their social networks to create new risk contexts (circumstancesin which the likelihood of an offence increases; Ward & Beech, 2004). It isbeneficial to focus on the preventive effect of certain types of social rela-tions (in our case protective triads), that might actually protect society fromcrime (Colvin et al., 2002). Involvement of network members can contribute

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

132 L. ter Haar-Pomp et al.

to better social control, support, and functioning of the patient. The higherthe density of relationships between personal network members, the morelikely those persons are capable of monitoring and adjusting the behavior ofthe patient (Kadushin, 2012). Further, not only will the patient benefit fromprotective social support, but it will also influence the well-being of family,friend, and community. Notably, Skeem, Eno Louden, Manchak, Vidal, andHaddad (2009) found that the quality of relationships is as important as asocial support as a buffer against criminogenic strains that lead to antisocialbehavior. They found evidence that a high-quality relationship is crucial foreffective social control.

We advise forensic mental health professionals to frequently check ifthere are significant changes in the patient’s network, such as a lack ofpersonal support and the involvement and influence of high-risk networkmembers. To estimate the influence of social network members on thepatient’s (risk) behavior, it is important to ask patients for reasons why peo-ple are mentioned as social supporters and why network members are nolonger part of the network. It is also essential that professionals apply effec-tive risk management interventions especially aimed at increasing supportivesocial support during treatment. This may contribute to a better quality offorensic psychiatric treatment.

ACKNOWLEDGMENTS

We are grateful to het Dok for allowing us to perform this research. We alsothank Peter, who gave permission for us to use the data for scientificresearch. The research was conducted independently; the executive boardof het Dok did not play a role in the research design, the data collection, theanalysis, or the interpretation of the data.

REFERENCES

Agnew, R. (1992). Foundation for a general strain theory of crime and delinquency.Criminology, 30(1), 47–87. Retrieved from http://dx.doi.org/10.1111/j.1745-9125.1992.tb01093.x

American Psychiatric Association. (2013). Diagnostic and statistical manual ofmental disorders (5th ed.). Arlington, VA: Author.

Andrews, D. A., & Bonta, J. (1994). The psychology of criminal conduct. Cincinnati,OH: Anderson.

Andrews, D. A., Bonta, J., & Hoge, R. D. (1990). Classification for effective reha-bilitation: Rediscovering psychology. Criminal Justice and Behavior, 17 , 19–52.Retrieved from http://dx.doi.org/10.1177/0093854890017001004

Bem, D. J., & Funder, D. C. (1978). Predicting more of the people more of the time:Assessing the personality of situations. Psychological Review, 85(6), 485–501.

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 133

Blau, P. M. (1964). Exchange and power in social life. New York, NY: Wiley.Bogaerts, S., Vanheule, S., & Desmet, M. (2006). Feelings of subjective emotional

loneliness: An exploration of attachment. Social Behavior and Personality, 34,797–812. Retrieved from http://dx.doi.org/10.2224/sbp.2006.34.7.797

Bolger, N., Zuckerman, A., & Kessler, R. C. (2000). Invisible support and adjustmentto stress. Journal of Personality and Social Psychology, 79, 953–961. Retrievedfrom http://dx.doi.org/10.1037/0022-3514.79.6.953

Bonta, J., & Andrews, D. (2010).Viewing offender assessment and rehabilitationthrough the lens of the risk-need-responsivity model. In F. McNeill, P. Raynor, &C. Trotter (Eds.), Offender supervision: New directions in theory, research andpractice (pp. 241–256). New York, NY: Willan.

Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). Ucinet for Windows: Softwarefor social network analysis. Harvard, MA: Analytic Technologies.

Bourdieu, P. (1981). Le capital social. Actes de la Recherche en Sciences Sociales, 31,2–3.

Buschman, J., Bogaerts, S., Foulger, S., Wilcox, D., Sosnowski, D., & Cushman,B. (2010). Sexual history disclosure polygraph examinations with cybercrimeoffences: A first Dutch explorative study. International Journal of OffenderTherapy and Comparative Criminology, 54, 395–411. Retrieved from http://dx.doi.org/10.1177/0306624X09334942

Cahill, B., Coolidge, F., Segal, D., Klebe, K., Marle, P. & Overmann, K. (2012).Prevalence of ADHD and its subtypes in male and female adult prison inmates.Behavioral Sciences & the Law, 30, 154–166. Retrieved from http://dx.doi.org/10.1002/bsl.2004

Cobb, S. (1976). Social support as a moderator of life stress. Psychosomatic Medicine,38, 300–314.

Colvin, M., Cullen, F. T., & Vander Ven, T. (2002). Coercion, social support, andcrime: An emerging theoretical consensus. Criminology, 40(1), 19–42. Retrievedfrom http://dx.doi.org/10.1111/j.1745-9125.2002.tb00948.x

Cook, K. S., & Whitmeyer, J. M. (1992). Two approaches to social structure: Exchangetheory and network analysis. Annual Review of Sociology, 18, 109–127.Retrieved from http://dx.doi.org/10.1146/annurev.so.18.080192.000545

Coyne, J. C., Wortman, C. B., & Lehman, D. R. (1988). The other side of sup-port. Emotional over involvement and miscarried helping. In B. H. Gottlieb(Ed.), Marshaling social support: Formats, processes and effects (pp. 305–330).Newbury Park, CA: Sage.

Cullen, F. T. (1994). Social support as an organizing concept for criminology:Presidential address to the academy of criminal justice sciences.Justice Quarterly, 11, 527–559. Retrieved from http://dx.doi.org/10.1080/07418829400092421

Cullen, F. T., Wright, J. P, & Chamlin, M. B. (1999). Social support and social reform:A progressive crime control agenda. Crime Delinquency, 45, 188–207. Retrievedfrom http://dx.doi.org/10.1177/0011128799045002002

Douglas, K. S., Hart, S. D., Webster, C. D., & Belfrage, H. (2013). HCR-20V3: Assessingrisk of violence: User guide. Burnaby, Canada: Mental Health, Law, and PolicyInstitute, Simon Fraser University.

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

134 L. ter Haar-Pomp et al.

Estroff, S. E., & Zimmer, C. (1994). Social networks, social support and violenceamong persons with severe, persistent mental illness. In J. Monahan & H. J.Steadman (Eds.), Violence and mental disorder: Developments in risk assessment(pp. 249–295). Chicago, IL: University of Chicago Press.

Gleason, M. E. J., Lida, M., Bolger, N., & Shrout, P. E. (2003). Daily supportiveequity in close relationships. Personality and Social Psychology Bulletin, 29,1036–1045. Retrieved from http://dx.doi.org/10.1177/0146167203253473

Gottlieb, B. H. (2000). Selecting and planning support interventions. In S. Cohen,L. G. Underwood, & B. H. Gottlieb (Eds.), Social support measurement andintervention: A guide for health and social scientists (pp. 195–220). New York,NY: Oxford University Press.

Haar-Pomp, L. ter, Spreen, M., Bogaerts, S., & Volker, B. (2014). The personalsocial networks of personality disordered forensic psychiatric patients. Journalof Social Work. doi:10.1177/1468017314537425

Henrichs, J., & Bogaerts, S. (2012). Predictors and correlates of posttraumaticstress disorder in forensic psychiatric outpatients in the Netherlands. Journalof Traumatic Stress, 25, 315–322. Retrieved from http://dx.doi.org/10.1002/jts.21706

Hirsch, B. J. (1980). Natural support systems and coping with major life changes.American Journal of Community Psychology, 8(2), 159–172. Retrieved fromhttp://dx.doi.org/10.1007/BF00912658

Hirschi, T. (1969). Causes of delinquency. Berkeley, CA: University of CaliforniaPress.

Hirschi, T. (1977). Causes and prevention of juvenile delinquency. SociologicalInquiry, 47 , 322–341. Retrieved from http://dx.doi.org/10.1111/j.1475-682X.1977.tb00804.x

Hlebec, V., & Kogovsek, T. (2013). Different approaches to measure ego-centeredsocial support networks: A meta-analysis. Quality & Quantity: InternationalJournal of Methodology, 47(6), 3435–3455. Retrieved from http://dx.doi.org/10.1007/s11135-012-9731-2

Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortalityrisk: A meta-analytic review. PloS Med, 7(7). Retrieved from http://dx.doi.org/10.1371/journal.pmed.1000316

Kadushin, C. (2012). Understanding social networks: Theories, concepts and findings.New York, NY: Oxford University Press.

Kalish, Y., & Robins, G. L. (2006). Psychological predispositions and network struc-ture: The relationship between individual predispositions, structural holes andnetwork closure. Social Networks, 28, 56–84. Retrieved from http://dx.doi.org/10.1016/j.socnet.2005.04.004

Kessler, R., Adler, L., Barkley, R., Biederman, J., Connors, K., Demler, O., . . .

Zaslavsky, A. (2006). The prevalence and correlates of adult ADHD in the UnitedStates: Results from the National Comorbidity Survey Replication. AmericanJournal of Psychiatry, 163, 716–723. Retrieved from http://dx.doi.org/10.1176/appi.ajp.163.4.716

Kooij, S. J. J. (2009). ADHD bij volwassenen. Diagnostiek en behandeling. [ADHDin adults. Diagnostics and treatment]. Amersfoort, The Netherlands: DrukkerijWilco.

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 135

Kumar, S., & Browne, M. A. O. (2008). Usefulness of the construct of social networkto explain mental health service utilization by the Maori population in NewZealand. Transcultural Psychiatry, 45, 439–454. Retrieved from http://dx.doi.org/10.1177/1363461508094675

Kunst, M., Winkel, F. W., & Bogaerts, S. (2010). Prevalence and predictors of posttrau-matic stress disorder among victims of violence applying for state compensation.Journal of Interpersonal Violence, 25, 1631–1654. Retrieved from http://dx.doi.org/10.1177/0886260509354591

Liang, J., Krause, N., & Bennett, J. (2001). Social exchange and well-being: Is givingbetter than receiving? Psychology and Aging, 16 , 511–523. Retrieved from http://dx.doi.org/10.1037/0882-7974.16.3.511

Lindenberg, S. M. (1996). Continuities in the theory of social productions func-tions. In H. Ganzenboom and S. M. Lindenberg (Eds.), Verklarende sociologie:Opstellen voor Reinhard Wippler (pp.169–184). Amsterdam, The Netherlands:Thesis.

Lindgren, C. L. (1990). Burnout and social support in family caregivers. WesternJournal of Nursing Research, 12(4), 469–487. Retrieved from http://dx.doi.org/10.1177/019394599001200404

May, B., & Bos, J. (2000). Personality characteristics of ADHD adults assessed withthe Millon Clinical Multiaxial Inventory-II: Evidence of four distinct subtypes.Journal of Personality Assessment, 75, 237–248. Retrieved from http://dx.doi.org/10.1207/S15327752JPA7502_5

McCarthy, B., & Hagan, J. (1995). Getting into street crime: The structure and processof criminal embeddedness. Social Science Research, 24, 63–69. Retrieved fromhttp://dx.doi.org/10.1006/ssre.1995.1003

Milardo, R. M. (1986). Personal choice and social constraint in close relationships:Applications of network analysis. In V. D. Derlega & B. A. Winstead (Eds.),Friendship and social interaction (pp. 145–166). New York, NY: Springer Verlag.

Miller, R. H., & Sim, I. (2004). Physician’s use of electronic medical records: Barriersand solutions. Health Affairs, 23(2), 116–126. Retrieved from http://dx.doi.org/10.1377/hlthaff.23.2.116

Monahan, J. (1981). Predicting violent behavior: An assessment of clinical techniques.London, UK: Sage Library of Social Research.

Monahan, J., Steadman, H. J., Silver, E. (2001). Rethinking risk assessment:The MacArthur Study of mental disorder and violence. Oxford, NY: OxfordUniversity Press.

Nakhaie, M. R., & Sacco, V. F. (2009). Social capital, individual disorders and propertyoffences. International Journal of Law and Psychiatry, 32, 392–399. Retrievedfrom http://dx.doi.org/10.1016/j.ijlp.2009.09.009

Odonne-Paolucci, E. O., Violato, C., & Schofield, M. A. (2000). A review of mar-ital and family variables as they relate to adult criminal recidivism. Calgary,Canada: National Foundation for Family Research and Education.

Ormel, J., Lindenberg, S., Steverink, N., & Verbrugge, L. M. (1999). Subjective well-being and social production functions. Social Indicator Research, 46 , 61–90.Retrieved from http://dx.doi.org/10.1023/A:1006907811502

Pomp, L., Spreen, M., Bogaerts, S., & Völker, B. (2010). The role of personal socialnetworks in risk assessment and management of forensic psychiatric patients.

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

136 L. ter Haar-Pomp et al.

Journal of Forensic Psychology and Practice, 4, 267–284. Retrieved from http://dx.doi.org/10.1080/15228932.2010.481232

Resnick, M. D., Ireland, M., & Borowski, I. (2004). Youth violence perpetrations:What protects? What predicts? Findings from the National Longitudinal Study ofAdolescent Health. Journal of Adolescent Health, 35, 424.e1–424.e10. Retrievedfrom http://dx.doi.org/10.1016/j.jadohealth.2004.01.011

Robinson, N. S., & Garber, J. (1995). Social support and psychopathology across thelife span. In D. Cicchetti & D. J. Cohan (Eds.), Developmental psychopathology:Vol. 2. Risk, disorder, and adaptation (pp. 162–209). New York, NY: Wiley.

Sampson, R. J., & Laub, J. H. (1990). Crime and deviance over the life course: Thesalience of adult social bonds. American Sociological Review, 55, 609–627.

Schuringa, E., Spreen, M., & Bogaerts, S. (2014). Inter-rater and test–retest reliability,internal consistency and factorial structure of the Dutch Instrument for ForensicTreatment Evaluation. Journal of Forensic Psychology Practice, 14(2), 127–144.Retrieved from http://dx.doi.org/10.1080/15228932.2014.897536

Shapiro, C., & diZegera, M. (2010). It’s relational: Integrating family into communitycorrections. In F. McNeill, P. Raynor, & C. Trotter (Eds.), Offender supervision:New directions in theory, research and practice (pp. 241–256). New York, NY:Willan.

Skeem, J., Encandela, J., & Eno Louden, J. (2003). Perspectives on probation andmandated mental health treatment in specialized and traditional probationdepartments. Behavioral Sciences & the Law, 21, 429–458. Retrieved from http://dx.doi.org/10.1002/bsl.547

Skeem, J., Eno Louden, J., Manchak, S., Vidal, S., & Haddad, E. (2009). Socialnetworks and social control of probationers with co-occurring mental and sub-stance abuse problems. Law and Human Behavior, 33(2), 122–135. Retrievedfrom http://dx.doi.org/10.1007/s10979-008-9140-1

Spreen, M., & Pomp, L. (2009). Landelijke implementatie Forensische SocialeNetwerk Analyse (FSNA): Procesevaluatie [Nationwide implementation ForensicSocial Network Analysis (FSNA): Process evaluation]. Utrecht, The Netherlands:Expertise Center for Forensic Psychiatry.

Surjadi, B., van Horn, J., Bogaerts, S., & Bullens, R. (2010). Internet offending: Sexualand non-sexual functions within a Dutch sample. Journal of Sexual Aggression,16(1), 47–58. Retrieved from http://dx.doi.org/10.1080/13552600903470054

Toner, M., O’Donoghue, T., & Houghton, S. (2006). Living in chaos and striving forcontrol: How adults with attention deficit hyperactivity disorder deal with theirdisorder. International Journal of Disability, Development and Education, 53,247–262. Retrieved from http://dx.doi.org/10.1080/10349120600716190

Vance, J. E., Bowen, N. K., Fernandez, G., & Thomson, S. (2002). Risk and protectivefactors as predictors of outcome in adolescents with psychiatric disorder andaggression. Journal of the American Academy of Child & Adolescent Psychiatry,41(1), 36–43. Retrieved from http://dx.doi.org/10.1097/00004583-200201000-00009

Veiel, H. O. F., & Bauman, U. (1992). The meaning and measurement of socialsupport. New York, NY: Hemisphere.

Vilhjalmsson, R. (1994). Effects of social support on self-assessed health in adoles-cence. Journal of Youth and Adolescence, 23, 437–52. Retrieved from http://dx.doi.org/10.1007/BF01538038

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15

Monitoring Risk Behaviors 137

Walker, M., Wasserman, S., & Wellman, B. (1994). Statistical models for social supportnetwork. In S. Wasserman & J. Galaskiewicz (Eds.), Advances in social networkanalysis: Research in the social and behavioral sciences (pp. 53–79). ThousandOaks, CA: Sage. Retrieved from http://dx.doi.org/10.4135/9781452243528.n3

Ward, T., & Beech, A. (2004). The etiology of risk: A preliminary model. SexualAbuse: A Journal of Research and Treatment, 16(4), 271–284. Retrieved fromhttp://dx.doi.org/10.1177/107906320401600402

Warr, M. (2002). Companions in crime: The social aspects of criminal conduct.Cambridge, UK: Cambridge University Press.

Warren, J. I., Stein, J. A., & Grella, C. (2007). Role of social support and self-efficacy in treatment outcomes among clients with cooccuring disorders. DrugAlcohol Dependence, 89, 267–274. Retrieved from http://dx.doi.org/10.1016/j.drugalcdep.2007.01.009

Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and application.Cambridge, UK: Cambridge University Press.

Weiss, M. D., & Weiss, J. R. (2004). A guide to the treatment of adults with ADHD.The Journal of Clinical Psychiatry, 65(3), 27–37.

Whittaker, J. K. (1992). Enhancing social support for high risk youth and their fami-lies following residential care. In J. D. Van der Ploeg, P. M. Van den Bergh, M.Klomp, E. J. Knorth, & M. Smit (Eds.), Vulnerable youth in residential care. Part1: Social competence, social support and social climate (pp. 81–100). Leuven,Belgium: Garant.

Zhou, L., Soran, C. S., Jenter, C. A., Volk, A. L., Orav, E. J., Bates, D. W., & Simon,S. R. (2009). The relationship between electronic health record use and qualityof care over time. Journal of the American Medical Informatics Association, 16 ,457–464. Retrieved from http://dx.doi.org/10.1197/jamia.M3128

Dow

nloa

ded

by [

Tilb

urg

Uni

vers

ity]

at 2

2:25

10

Apr

il 20

15