THE HETEROGENEITY OF SEXUAL ASSAULT RESPONSE TEAMS (SARTS): A
NATIONAL STUDY OF SART IMPLEMENTATION AND ITS IMPLICATIONS FOR SART
EFFECTIVENESS
By
Megan R. Greeson

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Psychology—Doctor of Philosophy
2013

 
 

ABSTRACT
THE HETEROGENEITY OF SEXUAL ASSAULT RESPONSE TEAMS (SARTS): A
NATIONAL STUDY OF SART IMPLEMENTATION AND ITS IMPLICATIONS FOR SART
EFFECTIVENESS
By
Megan R. Greeson
Sexual Assault Response Teams (SARTs) are community-level interventions that seek to
improve the response to sexual assault by increasing coordination and building positive
relationships among the systems that respond to sexual assault, particularly the legal, medical,
and mental health/advocacy systems (Greeson & Campbell, 2013; National Sexual Violence
Resource Center [NSVRC], 2011; Oregon Attorney General’s Sexual Assault Task Force
[OAGSATF], 2009; Pennsylvania Coalition Against Rape [PCAR], 2002; Zajac, 2006, 2009).
Ultimately, SARTs seek to improve the processing of sexual assault cases in the legal system and
improve sexual assault victims’ experiences of seeking help post-assault. Despite these common
goals, prior descriptive research (using convenience sampling) has illustrated that SARTs are
heterogeneous with respect to how they are structured. More specifically SARTs vary according
to the breadth of different stakeholder groups that are a part of the team (membership breadth),
and their implementation of formal structures (formalization) and collaborative activities (case
review, multidisciplinary cross-training, policy and protocol development and review, and
program evaluation) to organization their work. Diffusion of Innovation Theory (Rogers, 2005)
suggests that such variation in how SARTs are structured may create variation in SARTs’
effectiveness at improving victims’ help-seeking experiences and legal outcomes. As of yet,
there have been no studies with rigorous, representative samples that have examined how SARTs

 
 

are structured, and no studies have examined whether differences in SARTs’ structure are related
to differences in SARTs’ effectiveness.
Therefore, this national study of n=172 SARTs was conducted in pursuit of three aims:
(1) to use random sampling methods to obtain a nationally representative sample of SARTs to
assess their structural characteristics (2) to use cluster analysis to empirically derive subgroups of
SARTs with different profiles based on their implementation of formal structures and
collaborative activities; (3) to examine the relationship between SARTs’ structural characteristics
and their effectiveness at achieving improvements in legal outcomes and victims’ help-seeking
experiences. Interviews were conducted with SART leaders, who reported on their SARTs’
structural characteristics and their perceptions of their SARTs’ effectiveness. Descriptive
findings indicated that SARTs do vary in how they are structured. Cluster analysis was used to
identify three subgroups of SARTs that varied in their implementation of formal
structures/resources (formalization) and collaborative activities to organize their efforts: “Low
Adopters” (characterized by low formalization and less institutionalization of collaborative
activities), “High Adopters except Program Evaluation” (characterized by high formalization and
institutionalization of collaborative activities, but no program evaluation), and “High Adopters
plus Program Evaluation” (characterized by high formalization and institutionalization of
collaborative activities, as well as program evaluation). Findings indicated that SARTs in the
“High Adopters plus Program Evaluation” group tended to be perceived as more effective than
SARTs in the Low Adopters group. Other contextual features of SARTs and the communities
they served were also examined as predictors of effectiveness. Implications for future research,
policy, and practice for SARTs are discussed.

 

ACKNOWLEDGMENTS
Many people deserve recognition and thanks for the influence they have had on this
dissertation and my graduate career.
First, a very special thank you to my husband Danny for sharing every step of this
journey with me. Words aren’t enough. I can’t wait to take the next leap to Chicago together.
Thank you to my parents for teaching me to love books and learning. Thank you for
supporting and encouraging all through my life. To my family, thank you for keeping me
grounded, centered, and well fed. A special thanks to my beloved aunts for the many trips to the
MSU Dairy Store! I am grateful to my siblings, who make me want to live up to being a good
big sis and role model.
Thank you to Jonathan Livingston, Adrienne Adams, and Cris Sullivan, who were critical
in helping me find and pursue a career in community psychology.
To my committee—Deb Bybee, Bill Davidson, and Angela Kennedy. Thank you for your
support, guidance, and wisdom through the process of completing this dissertation and going on
the job market. Thank you for the many opportunities you have provided, your patience while I
fretted over getting a job, and for your confidence in me.
To the stats group—Deb, Angie, Adrienne, and Marisa. Thank you for your collegiality,
support, and the laughs we’ve had working together over the years. In the midst of academia, our
group was integral in helping me to stay grounded in who I am and helping me to see the joy and
friendship that can blossom in academia.

iv
 

Special thanks to Deb, who helped me see that I loved and was actually pretty good at
stats and data analysis. What a joy that has been. I am grateful for your exceptional kindness and
your unique ability to make me feel valued as a colleague.
To those special friends who really shared this experience with me--Adrienne, Debra,
Erin, and Lauren. Thank you for pushing me to think differently, for listening, for your gentle
advice, for keeping me laughing, for your strong values, for the cheerleading, and for just plain
being there whether taking walks, sitting on cars, or hanging by the fountain. You are the best
and it wouldn’t have been the same without you.
Thank you to the SARTs that were willing to share their time and insight with me and my
team and the National Sexual Violence Resource Center for their interest in the project.
Thank you to the members of the MSU National SART project team for their time, effort,
and patience.
Finally, to my chair, Becki Campbell—an outstanding mentor. I am so grateful that you
took me on as a green, first-year student and continued to work with me all throughout graduate
school. Thank you for your time and energy, your genuine interest in helping me identify and
pursue my own goals, and the multitude of doors you have opened and helped me to open. I am
so appreciative that you took a chance on me and agreed to undertake a big, difficult grant
together. Thank you for the many ways you went above and beyond the employer and advisor
role, and for your patience with my many quirks and limitations. You have done so much for me.

v
 

Funding Note:

This research was supported by a grant from the National Institute of Justice, 2010-WG-BX0010 awarded to Dr. Rebecca Campbell, PI. The opinions or points of view expressed in this
document are those of the authors and do not reflect the official position of the U.S. Department
of Justice.

vi
 

TABLE OF CONTENTS
LIST OF TABLES ................................................................................................................... ix 
LIST OF FIGURES .................................................................................................................. x 
OVERVIEW ............................................................................................................................. 1 
LITERATURE REVIEW ......................................................................................................... 3 
Sexual Assault as a Social Problem ................................................................................. 3 
The Role of Community Systems in Responding to Sexual Assault............................. 6 
Under-Utilization of Community Systems by Rape Survivors ..................................... 7 
Problems with the Response of Formal Systems to Sexual Assault Survivors............ 8 
Problematic Interactions between Formal Systems in the Response to Rape........... 12 
Sexual Assault Response Teams .................................................................................... 15 
Membership ................................................................................................................16 
Collaborative activities ..............................................................................................17 
Use of formal structures ............................................................................................18 
Effectiveness of SARTs ................................................................................................... 19 
Relationships between systems .................................................................................19 
Victims’ experiences ..................................................................................................20 
Legal outcomes ...........................................................................................................21 
Diffusion of Innovation Theory ..................................................................................... 22 
DOI, indigenous interventions, and SARTs ............................................................24 
Structural Factors Related to Collaborative Effectiveness ......................................... 26 
Overview of scholarship on collaboration ...............................................................26 
Membership ................................................................................................................28 
Activities......................................................................................................................30 
Use of formal structures ............................................................................................30 
Summary.....................................................................................................................32 
CURRENT STUDY................................................................................................................ 32 
METHOD ............................................................................................................................... 38 
Sampling .......................................................................................................................... 38 
Procedures ....................................................................................................................... 39
Recruitment ............................................................................................................... 39
Interviewer training ...................................................................................................41 
Piloting ........................................................................................................................43 
Interviewing procedures............................................................................................43 
Data storage and human subjects protection ..........................................................44 
Measurement ................................................................................................................... 45 
Contextual variables ..................................................................................................46 
Membership breadth (independent variable) .........................................................46 
vii
 

Formalization (independent variable)......................................................................46 
Use of collaborative activities (independent variables) ..........................................47 
Perceived effectiveness variables (dependent variables) ........................................47 
Data Analyses .................................................................................................................. 49 
Cluster analysis ..........................................................................................................49 
Structural equation modeling ...................................................................................53 
RESULTS ............................................................................................................................... 56 
Descriptives: Participant and SART Characteristics .................................................. 56 
Participants.................................................................................................................56 
SART contextual characteristics ..............................................................................57 
Structural characteristics of SARTs ........................................................................57 
SARTs’ perceived effectiveness ..................................................................................... 57 
Clustering SARTs on their use of Formal Structures and Collaborative Activities 58 
Cluster one ..................................................................................................................61 
Cluster two ..................................................................................................................61 
Cluster three ...............................................................................................................62 
Predicting SARTs’ Perceived Effectiveness ................................................................. 63 
Improving victims’ help-seeking experiences..........................................................68 
Improving victims’ participation in the criminal justice system ...........................68 
Improvements related to police processing of sexual assault cases and
improvements related to the prosecution of sexual assault cases ..........................70 
DISCUSSION ......................................................................................................................... 71 
Aim One: SARTs’ Structural Characteristics ............................................................. 72 
Aim Two: Clustering SARTs on their Formalization and Collaborative Activities. 72 
Aim Three: Predicting SARTs’ Perceived Effectiveness ............................................ 73 
Cluster membership and perceived effectiveness ...................................................73 
Membership and effectiveness ..................................................................................75 
Contextual features and effectiveness ......................................................................76 
Limitations and Implications for Future Research ..................................................... 80 
Implications for Policy and Practice ............................................................................. 83 
Structure .....................................................................................................................83 
Evaluation ...................................................................................................................85 
Conclusion ....................................................................................................................... 86 
APPENDICES ........................................................................................................................ 88 
APPENDIX A: Study 1: SART Leader Interview ........................................................ 89 
APPENDIX B: Perceived Legal Effectiveness Items and their Corresponding
Subscales ........................................................................................................................ 139 
APPENDIX C: Comparison of Three and Four Cluster Solution Results ............... 141 
REFERENCES ..................................................................................................................... 146 

viii
 

LIST OF TABLES
Table 1: Proportion of U.S. SARTs Utilizing Recommended Activities and Infrastructure .. 18 
Table 2: Bivariate Correlations among the Clustering Variables ........................................... 58 
Table 3: Cluster Analysis of SARTs’ Use of Formal Structures and Collaborative Activities
Results………………………………………………………………………………………..60 
Table 4: Correlations among Predictors of SARTs’ Perceived Effectiveness from the Final
Path Model .............................................................................................................................. 65 
Table 5: Correlations among Residuals of the Perceived Effectiveness (Dependent) Variables
from the Final Path Model ...................................................................................................... 66 
Table 6: Predictors of SARTs’ Perceived Effectiveness from the Final Path Model ............. 67 
Table 7: Perceived Legal Effectiveness Items and their Corresponding Subscales ............. 139 
Table 8: Results of Three Cluster Solution (Chosen) ........................................................... 142 
Table 9: Results of Four Cluster Solution (Not Chosen) ...................................................... 143 
 

ix
 

LIST OF FIGURES
Figure 1: Sampling Flow Chart............................................................................................... 42 
Figure 2: Cluster Analysis ...................................................................................................... 50 
Figure 3: Model of SART Structure and Perceived Effectiveness with Control Variables.... 53 
Figure 4: Revised Full Path Model ......................................................................................... 55 
 

x
 

OVERVIEW
Over the past two decades, improving the community response to rape has received
substantial attention from researchers, practitioners, and policy makers. Key stakeholders in the
response to sexual assault include the legal system (police and prosecutors who investigate and
prosecute cases), the medical system (doctors and nurses who collect forensic evidence and
attend to victims’ healthcare needs), and the mental health/advocacy systems (counselors and
advocates who address mental health needs, and provide support as victims navigate the legal
and medical systems; Martin, 2005). Historically, these systems have failed to hold rapists
accountable and provide victims with services in a sensitive manner (see Campbell, 2008 and
Seidman & Pokorak, 2011 for reviews). Through local, grassroots efforts, communities
throughout the United States have developed two alternatives to address these problems: Sexual
Assault Nurse Examiner Programs (SANEs) and Sexual Assault Response Teams (SARTs).
In the 1970s, the first SANE programs were developed (Ledray, 2001). SANE programs
are staffed by nurses who have received extensive specialized training in attending to rape
victims’ emotional needs; providing comprehensive medical services; and collecting,
documenting, and handling medical/forensic evidence which can then be used in legal cases
against assailants (Ledray, 2001; Littel, 2001). Research on SANEs has demonstrated that they
are effective at improving rape victims’ experiences with the medical system, providing more
comprehensive medical services, and improving the quality of medical/forensic collection for use
in the legal system; in some communities, they have also been able to contribute to
improvements in local prosecution rates (see Campbell, Patterson, & Lichty, 2005 for review).
SARTs emerged in the 1970s in tandem with SANE programs (Zajac, 2006).

1
 

SARTs are community-level interventions that seek to improve the response to sexual assault by
increasing coordination and building positive relationships among the systems that respond to
sexual assault, particularly the legal, medical, and mental health/advocacy systems (Greeson &
Campbell, 2013; National Sexual Violence Resource Center [NSVRC], 2011; Oregon Attorney
General’s Sexual Assault Task Force [OAGSATF], 2009; Pennsylvania Coalition Against Rape
[PCAR], 2002; Zajac, 2006, 2009). SANEs were involved in the development of some of the
first SART teams, and now, the majority of Sexual Assault Nurse Examiner programs participate
in a Sexual Assault Response Team (Logan, Cole, & Capillo, 2007; Plichta, Vandecar-Burin,
Odor, Reams, & Zhang, 2006).The underlying philosophy of SARTs is that by learning from one
another’s expertise, developing trust, encouraging communication, increasing accountability, and
creating a coordinated service system, the legal, medical, and mental health systems can more
effectively meet victims’ needs and hold rapists accountable.
While the body of research on SANEs has grown and continued to demonstrate their
utility (see Campbell et al., 2005 for review), there have been fewer studies of SARTs’
effectiveness and the results of these studies have been less clear (see Greeson & Campbell, 2013
for recent review). The limited research that exists shows that some SARTs have the potential to
improve victims’ help-seeking experiences, and impact some legal outcomes like arrest rates
(e.g., Noble, Brannon-Patel, & Tysoe, 2001; Nugent-Borakove, Fanflik, Troutman, Johnson,
Burgess, & Lewis-O’Connor, 2006). An examination of practitioner writings on the SART
model (i.e., what constitutes a SART and how it should be implemented) as well as descriptive
research on SARTs provides some insight into these mixed findings. Among practitioner
writings on SARTs, there is a lack of consensus regarding how exactly SARTs should be
structured and function. Moreover, descriptive research confirms that in practice, SARTs are
2
 

quite heterogeneous (Zajac, 2006, 2009). It seems plausible that due to differences in how they
are implemented, some SARTs are more successful than others. However, research has yet to
examine how differences in the implementation of SART teams may relate to differences in their
effectiveness.
The current study seeks to advance our understanding of SARTs by furthering our
knowledge on the heterogeneity of SARTs and exploring the relationship between how SARTs
are structured and function and their effectiveness. In order to situate the work of SARTs, the
literature review will begin with an introduction to the prevalence of sexual assault and its effects
on victims’ well-being. This will be followed by an overview of the community response to rape,
and a review of the problems that have been associated with the traditional uncoordinated
response to sexual assault. Then, a comprehensive review of the research on Sexual Assault
Response Teams will be presented. Next, diffusion of innovation theory and its relevance to
scholarship on SARTs will be discussed. The review will conclude with a discussion of the
broader literature base on structural factors that are related to the effectiveness of collaborations,
again applied to the context of SARTs.
LITERATURE REVIEW
Sexual Assault as a Social Problem
Sexual assault is defined as sexual contact that occurs in the absence of consent, meaning
the contact is perpetrated through the use of force, threats of force, or when the victim is unable
to consent to sexual activity (for example, due to age, mental disability, or intoxication; Krug,
Dahlberg, Mercy, Zwi, & Lozano, 2002). The term rape refers specifically to assaults that
involve penetration or attempted penetration of the victim’s body (Krug et al., 2002). Rapists and
their victims come from all social groups; most often, sexual assault is perpetrated by men
3
 

against women and children (Tjaden & Thoennes, 2006; World Health Organization [WHO],
1

2003) .
Rape is an extensive problem that has affected the lives of millions of women in the
United States (Post, Biroscak, & Barboza, 2011). National random samples of the general
population (obtained by random digit dialing) have found that 11-18% of adult women had been
raped in their lifetimes and 0.3% to 0.9% had been raped in the past year (Basile, Chen, Black, &
Saltzman, 2007; Kilpatrick et al., 2007; National Victim Center, 1992; Tjaden & Thoennes,
2006). These studies may actually underestimate the magnitude of rape as random digit dialing
limits the inclusion of certain high-risk groups in the sample (e.g. homeless women) and the
2

researchers utilized narrower definitions of victimization (Raphael & Logan, 2008). In fact, in a
nationally representative sample of college women, Koss et al. found a much higher prevalence
rate: 28% of their sample had been raped (completed or attempted) since the age of 14 (Koss,
Gidcyz, & Wisniewski, 1987). Despite the inconsistencies in this area of research, it is clear that
rape is not an uncommon experience (see Raphael & Logan, 2008 for a recent critique of the
rape prevalence literature).

                                                            
1

As such, the female pronoun will be used to refer to victims throughout this document.

2

The National Violence Against Women Study by Tjaden and Thoennes (2006) and the
National Women’s Study (National Victim’s Center, 1992) asked about rape which involved
force or threats of harm (which is unlikely to capture assaults that occurred when the victim was
unable to consent, for example, due to intoxication from alcohol or drugs). The National
Women’s Study (National Victim’s Center, 1992), the Injury Control and Risk Survey (Basile et
al., 2007) and Kilpatrick and colleagues’ study of drug-facilitated, incapacitated, and forcible
rape (Kilpatrick et al., 2007) captured completed penetration, but not attempted rape.
4
 

The pervasiveness of this problem is particularly troubling when one considers the effects
it can have on people’s lives. Fundamentally, rape is a violation of a person’s body and will, and
3

such a traumatic incident can profoundly impact survivors’ emotional and physical well-being.
Immediately after the assault, survivors may have injuries and be at risk for sexually transmitted
infections and unwanted pregnancy (Basile & Smith, 2011; El-Mouehly, 2004; S. Martin &
Macy, 2009; S. Martin, Macy, & Young, 2011; Resnick, Acierno, Holmes, Dammeyer, &
Kilpatrick, 2000). In addition, many survivors experience significant psychological distress postassault. Feelings of fear, agitation, confusion, shock, shame, self-blame, emotional detachment,
and social withdrawal are common (Basile & Smith, 2011; Bryant-Davis, Chung, & Tillman,
2009; Jordan, Campbell, & Follingstad, 2010; Yuan, Koss, & Stone, 2006).
In the long term, rape victims are at heightened risk for a variety of negative outcomes.
Mental health problems associated with sexual assault include depression, posttraumatic stress
disorder, generalized anxiety problems, suicidality, substance abuse, poor self-esteem, stress, and
somatic complaints (Basile & Smith, 2011; Booth, Mengeling, Torner & Sadler; Bryant-Davis et
al., 2009; Campbell, Dworkin, & Cabral, 2009; El-Mouehly, 2004; Hanson et al., 2008; Jordan et
al., 2010; Kilpatrick & Acierno, 2003; Kimerling & Calhoun, 1994; Koss, Bailey, Yuan, Herrera,
& Lichter, 2003; S. Martin & Macy, 2009; S. Martin, Macy, & Young, 2011; Palm & Follette,
2008; Yuan et al., 2006; Zinzow et al., 2010a; Zinzow et al., 2010b). Long-term physical health
sequelae include headaches, chronic pain, pregnancy complications, and gastrointestinal,
                                                            
3

The terms “victim” and “survivor” will be used interchangeable throughout this document, as
is often done in this literature. Victim denotes the crime that was committed against them, and
survivor acknowledges the strengths and the agency of the people who have lived through such
experiences.
5
 

gynecological, sleeping, and sexual problems. More generally, when compared to non-victims,
rape survivors report a greater number of physical health symptoms and poorer perceptions of
their health (Amstadter, McCauley, Ruggiero, Resnick, & Kilpatrick, 2010; Cloutier, S. Martin,
& Poole, 2002; Golding, 1999; Jordan et al., 2010; Kimerling & Calhoun, 1994; S. Martin &
Macy, 2006; Martin, Macy, & Young. 2011; Palm & Follette, 2008; Resnick and Acierno et al.,
2000). These findings show that rape is not only a crime, but also a serious public health issue.
The scope of the problem suggests that both prevention of rape and working to mitigate the
negative consequences of rape are important aims.
The Role of Community Systems in Responding to Sexual Assault
Social systems—in particular, the legal, medical and mental health/advocacy systems—
can play an important role in responding to victims’ needs. The legal system is responsible for
apprehending the suspect and investigating and prosecuting the crime (P. Martin, 2005; Seidman
& Pokorak, 2011; Spohn & Tellis, 2012). The medical system provides forensic and medical
services. Nurses and/or physicians are responsible for collecting and documenting forensic
evidence of the assault from the survivors’ body and clothing (e.g. documenting injuries,
swabbing for DNA)—evidence which can then be utilized in the criminal justice process
(American College of Emergency Physicians, 1999; Department of Justice, 2004; P. Martin,
2005). In addition, nurses and doctors attend to victims’ medical needs, such as addressing any
injuries the survivor may have sustained and providing prophylactic medication to prevent STIs
and pregnancy (American College of Emergency Physicians, 1999; Department of Justice, 2004;
P. Martin, 2005). Finally, mental health providers, including rape crisis center advocates and
counselors, provide crisis intervention, counseling, and support groups to help survivors cope
with the emotional effects of the assault (Campbell & P. Martin, 2001; Macy, Giatinna, Sangster,
6
 

Crosby, & Montijo, 2009; P. Martin, 2005). Advocates may also provide informational and
emotional support to survivors to help them navigate the legal and medical systems (Campbell &
P. Martin, 2001; Macy et al., 2009; P. Martin, 2005).
Under-Utilization of Community Systems by Rape Survivors
Although these social systems are available resources for survivors, they are vastly
under-utilized. Studies have found that a substantial proportion of victims do not seek help from
formal systems in response to the assault (see Ullman, 2010 for a review). Approximately 6-40%
of adult rape survivors report the assault to police, 10-43% have contact with the medical system,
and 27-60% access mental health services (Ahrens, Cabral, & Abeling, 2009; Amstadter,
McCauley, Ruggiero, Resnick, & Kilpatrick, 2008; Brecklin & Ullman, 2010; Campbell, 2008,
Campbell & Raja, 2005; Campbell, Wasco, Ahrens, Sefl, & Barnes, 2001; Chen & Ullman,
2010; Filipas & Ullman, 2001; George, Winfield & Blazer, 1992; Kilpatrick, Resnick, Ruggiero,
Conoscenti, & McCauley, 2007; Resnick and Holmes et al., 2000; Tjaden & Thoennes, 2006;
Ullman, 1996, 2007; Ullman, Filipas, Townsend, & Starzynski, 2007; Wolitzky-Taylor et al.,
2011). However, these systems are not bypassed by sexual assault victims without reason.
Barriers to help-seeking include issues of accessibility, such as lack of awareness of services,
limited program hours, and lack of transportation (Logan, Evans, Stevenson, & Jordan, 2005). In
addition, many victims believe that their experiences do not qualify them for the services that are
available. For example, they may feel that it has been too long since the assault, or they believe
the incident is not “serious” enough (Logan et al., 2005; Patterson, Greeson, & Campbell, 2008;
Tjaden & Thoennes, 2006; Walsh, Banyard, Moynihan, Ward, & Chon, 2010). Finally, many
survivors have negative perceptions of formal systems and believe that the systems may not be
helpful to them or that contact with these systems will actually be hurtful, and therefore do not
7
 

seek help (Logan et al., 2005; Patterson et al., 2008; Tjaden & Thoennes, 2006; Ullman, 2010;
Walsh et al., 2010; Wolitzky-Taylor et al., 2011). These barriers to help-seeking are problematic
because they prevent survivors from receiving services that may facilitate their emotional and
physical recovery; in addition, they prevent women from reporting the assault which contributes
to the under-reporting of rape cases.
Problems with the Response of Formal Systems to Sexual Assault Survivors
The negative perceptions of social systems that deter survivors from reporting and
seeking mental and physical health care are not unwarranted. Historically, the community
response to sexual assault has been inadequate. To provide a brief overview, key problems that
will be reviewed include: gaps in service provision in the medical system, problems with
forensic evidence collection, low conviction rates, and negative treatment of survivors by legal
and medical systems personnel, which exacerbates the trauma that survivors have experienced.
In the medical system, many survivors who seek help from a hospital emergency
department do not receive comprehensive services: approximately 29-60% do not receive
information about risk of pregnancy, 39-80% do not receive emergency contraception, 31-61%
do not receive information about STIs, and 13-51% do not receive some form of prophylactic
medication (Amey & Bishai, 2002; Campbell 2005; Campbell & Bybee, 1997; Campbell & Raja,
2005; Campbell et al., 2001; Monroe et al., 2005; National Victim Center, 1992; Rambow,
Adkinson, Frost, & Peterson, 1992; Rovi & Shimoni, 2002). Clearly, many victims do not
receive appropriate health care post-assault (see Campbell & Patterson, 2011 for recent review).
In addition to these gaps in health care for rape survivors, there are serious problems with
traditional emergency department (ED) forensic evidence collection. Typically, ED personnel are
not adequately trained in evidence collection and do not conduct enough forensic evidence
8
 

exams to maintain proficiency; as a result ED staff often neglect to collect all types of available
evidence and fail to maintain the chain of evidence (Littel, 2001; P. Martin, 2005; Plichta et al.,
2006). Some staff may even deny survivors forensic exams altogether. In a series of studies,
Campbell and colleagues found that 14-30% of survivors who seek medical help do not receive a
forensic exam (Campbell, 2005; Campbell & Bybee, 1997; Campbell et al., 2001). Failure to
provide evidence collection, as well as inadequate evidence collection and maintenance, result in
less evidence for the legal system to use in prosecuting and convicting rapists.
These problems with forensic evidence collection may contribute to another troublesome
aspect of the community response to sexual assault: the staggeringly high rate of attrition of rape
cases in the legal system. Research has demonstrated that the majority of rape cases are closed/
dropped in the earliest stages of the criminal justice process (e.g., Campbell et al., 2001; Frazier
4

& Haney, 1996; Lonsway & Archambault, 2012; Tjaden & Thoennes, 2006) . In a communitybased study of women seeking help from an urban Veteran’s Administration hospital, over half
of rape victims who reported the assault stated that they were discouraged from reporting by
legal personnel and approximately half stated that legal personnel refused to take the report of
the assault (Campbell & Raja, 2005). Another study found that police frequently fail to take
action on rape cases that are brought to their attention. Campbell (2005) interviewed rape victims
who sought help from an urban emergency department and the police officers who responded
when they reported the assault. Based on the accounts of the officers themselves, approximately
                                                            
4

Lonsway and Archambault concluded (based on Uniform Crime Report data) that “a fraction”
of reported rape cases result in arrest, and that this pattern is similar to other violent crimes.
However, they also concluded that the proportion of rape cases that result in arrest is getting
smaller over time, while the proportion of other violent crimes that result in arrest has remained
fairly steady.
9
 

half of rape case reports are not filed (45%) and only one-fourth (27%) are investigated
(Campbell, 2005). Given these findings, it is perhaps not surprising that approximately twothirds of reported cases are closed by police and never referred to the prosecutor’s office for
consideration of filing charges (see Campbell, 2008 for review). Evidence illustrates that legally
relevant (e.g., forensic evidence) as well as legally irrelevant factors (e.g., victim characteristics
such as risk-taking behavior) influence police officers’ decision-making regarding which cases to
refer/not refer for prosecution (Spohn & Tellis, 2012).
Case attrition is not just a problem that stems from the actions of police officers; many
cases are also dropped in the prosecution phase of the criminal justice system process. As in the
case of police, prosecutors attune to both legally-pertinent factors (e.g., victim willingness to
prosecute) as well as legally-extraneous factors (victim and perpetrator race), when deciding
which cases to pursue (Spohn & Tellis, 2012). Recent reviews have concluded that
approximately 33-40% of reported cases result in the prosecutor’s office filing charges and 716% result in incarceration of the offender (Campbell, 2008; Seidman & Pokorak, 2011). Given
that less than half of cases are reported by survivors, only a fraction of rape cases result in the
rapist being held criminally accountable. This is a pressing matter, as the majority of rapists are
repeat offenders (Lisak & Miller, 2002), and those that are not held criminally accountable
remain at large, free to perpetrate again.
Thus, survivors who turn to the legal and medical systems have a very high probability of
not attaining the services and legal outcomes they seek. For many victims, in addition to not
being helpful, legal and medical system personnel are actually hurtful to many survivors.
Research by Ullman and colleagues has shown that rape victims who disclose the assault to a
formal community system (including the legal and medical systems) receive a higher number of
10
 

negative reactions to disclosure (e.g., victim-blaming; Filipas and Ullman, 2001; Starzynski,
Ullman, Filipas, & Townsend, 2005; Ullman, 1996). More specifically, survivors who have had
contact with the legal and medical systems describe a variety of upsetting experiences, such as
system personnel treating them in a cold or impersonal manner, failing to express empathy,
treating them with disbelief, suggesting that they were to blame for the assault, and denying them
services (Campbell, 2005, 2008; Campbell & Raja, 1999; Logan et al., 2005; Madigan &
Gamble, 1991; P. Martin, 2005; P. Martin & Powell, 1994). The evidence reveals that such
negative experiences, termed “secondary victimization,” are quite common. Results from
community samples indicate that the majority of rape victims who seek medical and or legal
assistance experience some form of secondary victimization (Campbell 2005; Campbell et al.,
2001; Campbell & Raja, 2005), and this has also been corroborated by legal and medical
personnel themselves (Campbell, 2005). The pervasiveness of secondary victimization indicates
that this is not an idiosyncratic problem caused by a few personnel, but is instead systemic in
nature.
These experiences are not benign. Negative interactions with legal and medical system
personnel (which typically take place in the first few days after the assault) exacerbate the
trauma that rape survivors have already experienced (Campbell, 2008; Campbell et al., 2001).
Many survivors report that denial of services and victim-blaming, disbelieving, and cold
treatment by legal and medical system personnel contribute to distressing emotions such as
feeling bad about themselves, self-blame, depression, anxiety, and feeling violated (Campbell,
2005; Campbell & Raja, 2005). Studies of the effects of survivors’ disclosure to a variety of
recipients—including legal, medical, and mental health system personnel—have found that
receiving more negative victim-blaming reactions to disclosure is associated with more severe
11
 

posttraumatic stress symptoms (Ullman, 2010; Ullman & Filipas, 2001; Ullman et al., 2007).
Negative interactions specifically from legal and medical system personnel are also associated
with increased posttraumatic stress, as well as poorer physical and psychological health
(Campbell et al., 2001). In short, many rape victims reach out to community systems and, instead
of receiving help, are actually harmed.
In summary, there are a variety of inter-related problems with the formal response to
sexual assault including: (1) low reporting and help-seeking rates; (2) gaps in medical service
delivery and forensic evidence collection; (3) case attrition in the legal system; and (4)
exacerbation of survivors’ trauma by legal and medical systems personnel.
Problematic Interactions between Formal Systems in the Response to Rape
In addition to these problems regarding the interactions between survivors and systems,
there are also problems regarding the relationships between the legal, medical, and mental
health/advocacy systems. Bronfenbrenner’s ecological theory (1979) posits that individuals are
influenced not only by the settings they directly interact with, but also the interactions between
those settings. Applied to the case of rape, this means that the legal, medical, and medical
systems—as well as the interactions between these systems—impact survivors and their cases.
Historically, relationships among the legal, medical, and mental health systems constitute
another problematic aspect of the community response to rape, as relationships are often nonexistent or even outright negative. In part, this likely due to the fact that these systems have
fundamental differences, from different organizational structures, different funding streams,
different policies and procedures for governing their work, to different goals and philosophies for
working with victims (P. Martin, 2005). The criminal justice system’s priority is public safety,
meaning that they want victims to get medical/forensic evidence collected, and to report and
12
 

participate in the prosecution of their rapist (Martin, 2005). On the other hand, the burden of
proof in the U.S. criminal justice system requires evidence the crime was committed by the
accused rapist beyond a reasonable doubt and it is the job of police and prosecutors to help
determine which cases are/are not substantiated by sufficient evidence (P. Martin, 2005).
Consequently, there is also a long-standing pattern of devaluing victims in cases where criminal
justice system personnel perceived the victim to lack credibility and in cases that lack hard
evidence (P. Martin, 2005, Spohn & Tellis, 2002). Furthermore, prosecutors, who are elected
officials, may be reluctant to pursue cases that are not sure to result in conviction. Hospitals and
emergency rooms provide critical medical care to patients, may be reluctant to care for rape
victims without serious injuries when they have other patients to attend to, and may also be
hesitant to engage in the forensic aspects of rape cases for fear of getting entangled in the legal
system (P. Martin, 2005). Rape victim advocates and counselors provide support to all victims,
and believe that attending to victims’ well-being and restoring choice and control their lives is of
utmost importance (P. Martin, 2005). This may put them at odds with the legal system when a
victim does not want to participate in prosecution or at odds with the medical system when they
need to advocate for doctors/nurses to provide the victim with services (P. Martin, 2005).
Perhaps not surprisingly, research illustrates that—historically—these very different
groups have not worked well together to address sexual assault. In many communities, the
response to rape is fragmented, with systems responding to survivors in isolation from one
another. Responders from one system lack knowledge of how other systems respond to survivors
(Campbell & Ahrens, 1998). Role blurring and/or conflicts across systems regarding who should
do what when responding to rape are common (P. Martin, DiNitto, Byington, & Maxwell, 1992).
Negative relationships may even exist among stakeholder groups, and cause interference with
13
 

one another’s work. For example, P. Martin (2005) describes instances in which nurses and
prosecutors, who devalued the work of victim advocates, excluded advocates from the medical
forensic exam and the pretrial interview with the prosecutor—preventing the advocate from
providing advocacy services to the survivor.
These problematic interactions between systems result in missed opportunities for
helping survivors access comprehensive services. Survivors, rather than reaching out for help
once to an interconnected web of community responders, must identify all resources and seek out
help multiple times from each system individually (Campbell & Ahrens, 1998). In a coordinated
service-system, one system takes responsibility for helping survivors access the other systems. In
the uncoordinated model, victims who must reach out for assistance repeatedly may fatigue and
give up, in essence falling through the proverbial cracks of an uncoordinated system.
This lack of coordination between systems also creates missed opportunities for diverse
sexual assault responders to assist one another. Systems are dependent upon one another for their
work. For example, police depend on the medical system for forensic evidence collection and the
prosecutor to charge cases, the prosecutors rely on police to build good cases, and, to some
extent, rape victim advocates rely on the medical and legal systems for access to survivors in
order to provide them with advocacy services (Martin, 2005). Such dependency requires an
understanding of how these systems can and should work together, and what each group can
contribute to the other groups in order to be effective. However, in uncoordinated communities
there is a fundamental lack of opportunities for the responders to “get to know each other, learn
from each other, and support each other” (Campbell & Ahrens, 1998, pp. 562). This may create a
variety of problems. If police are not aware that medical/forensic personnel can conduct
examinations of suspects for DNA evidence, they will not include this type of evidence
14
 

collection in their investigations. If prosecutors lack an understanding of evidence collection
techniques employed by medical/forensic personnel they are limited in their utilization of
forensic findings. If medical and legal system personnel resist the presence of rape advocates,
they lose an opportunity to facilitate survivors’ emotional recovery, and potentially their longterm participation in the legal system. It is likely that such missed opportunities for survivors and
systems contribute to the many problems reviewed before: low reporting and help-seeking rates,
gaps in medical service delivery and forensic evidence collection, case attrition in the legal
system, and negative treatment of survivors by legal and medical systems personnel.
Sexual Assault Response Teams
In recognition of these problems, communities through the United States have developed
an alternative model for responding to sexual assault. Sexual Assault Response Teams (SARTs)
are community-level interventions that seek to improve the response to sexual assault by
increasing coordination and building positive relationships among the systems that respond to
5

sexual assault, particularly the legal, medical, and mental health/advocacy systems . Typically,
SARTs primarily focus on improving victims’ post-assault help-seeking experiences (e.g., by
addressing barriers to seeking help, improving the quality of service delivery, and decreasing
revictimization of survivors by system personnel); and improving legal outcomes (e.g.,
increasing reporting to the police, improving prosecution/conviction rates, improving forensic

                                                            
5

Note that some organizations use the term “Sexual Assault Response Team” to refer to a
medical team of nurses and/or doctors that respond within a hospital to rape victims. In this
document, the term SART will be limited to multidisciplinary teams that include stakeholder
groups from multiple systems.
15

 

evidence, increasing victim participation in the criminal justice system; PCAR, 2002;
OAGSATF, 2009; Zajac, 2009).
Creating collaboration between the key groups that respond to sexual assault—
particularly the legal, medical, and advocacy/mental health systems is a lofty, but highly
desirable goal. Since the inception of the first SARTs in the 1970s, SARTs have diffused rapidly
and widely to hundreds of communities (Zajac, 2006, 2009). While the commonality among
SARTs is multidisciplinary coordination and relationship-building, there is considerable
evidence that precisely how these teams are structured and function varies from community to
community. Key domains in which SARTs may differ include their membership, activities, and
their use of formal structures and resources to organize their team.
Membership. The representation of various stakeholder groups may vary across SARTs.
Most guidelines recommend that SARTs include (at a minimum) the four primary immediate
responders to sexual assault: medical/forensic examiners, police, prosecutors, and rape crisis
center victim advocates (Department of Justice [DOJ], 2004; Florida Coalition Against Sexual
Violence [FCASV], 2007; Kentucky Association of Sexual Assault Programs [KASAP], 2002;
Ledray, 2001; OAGSATF, 2009; PCAR, 2002). However, descriptive research on SARTs
suggests that not all SARTs engage each of these groups. In two studies in 2005 and 2008-2009,
the National Sexual Violence Center (NSVRC) utilized convenience sampling to survey SARTs
regarding their operations (Zajac, 2006, 2009). Results showed that 86-87% of teams that
responded to the survey included victim advocates, 84-85% included law enforcement, 73%77% included medical/forensic examiners, and 72-76% included prosecutors (Zajac, 2006,
2009). Approximately two-thirds of teams included all four of these stakeholder groups (Zajac,
2009).
16
 

SARTs may also differ in the degree to which they involve other stakeholders in their
efforts to address sexual violence. Some SART development guidelines recommend the
inclusion of crime lab personnel, dispatch, corrections, judges, sex offender treatment personnel,
and/or victim advocates from the prosecutor’s office as key players in the response to sexual
assault (DOJ, 2004; FCASV, 2002; OAGSATF, 2009; PCAR, 2002). In addition, other service
organizations that are not geared toward providing services to sexual assault victims, but may
have contact with them may be invited to participate in a SART (including mental health, public
health, and social services agencies, schools, religious/faith institutions). Additionally,
organizations that serve specific populations within the community who are affected by sexual
assault (such as college students, LGBTQ, military, tribal, people with disabilities) may become
members of a SART (DOJ, 2004; FCASV, 2002; OAGSATF, 2009; PCAR, 2002). In sum,
SARTs’ membership may vary based on which stakeholders and how many stakeholders belong
to their team.
Collaborative activities. Similar to their membership, SARTs exhibit heterogeneity in
the collaborative activities in which they engage. SART guidelines recommend a variety of
activities. The most commonly recommended activities include: case review (in which the
response to individual cases is systematically reviewed by a multidisciplinary group);
development of policies and protocols that delineate the desired response to sexual assault in
their community; cross-disciplinary trainings to increase stakeholders’ knowledge of one
another’s roles and limitations and improve their ability to respond to sexual assault effectively;
and evaluation to assess the implementation and outcomes of their team’s efforts (FCASV, 2007;
KASAP, 2002; NSVRC, 2011; OAGSATF, 2009; PCAR, 2002). SARTs may participate in

17
 

none, some, or all of these activities. Evidence from the NSVRC studies shows that SARTs
implementation of these “best-practices” varies from community to community. See Table 1.
Use of formal structures. In a similar vein, SARTs vary in the degree to which they
utilize formal structures and resources to organize their multidisciplinary teamwork (their
formalization). Although guidelines recommend the implementation of formal structures
(including, but not limited to, regular meetings, committee structure, SART leader, facilitator, or
coordinator, mission statement (e.g., FCASV, 2007; KASAP, 2002; OAGSATF, 2009), SARTs
range from very informal partnerships to highly formalized collaborative groups (NSVRC,
2011). See Table 1 for descriptive statistics on SARTs’ formalization.
Table 1:
Proportion of U.S. SARTs Utilizing Recommended Activities and Infrastructure
SART feature

NSVRC 2005

NSVRC 2008-2009
Activities

Regular case review

23%
43-57%

Policies and protocols
Holds trainings for SART team members

25%
a

52%

n/a

64%
Formal structure

Holds regular collaborative meetings

82%

81%

Team facilitator or coordinator

60%

82%

Paid staff

44%

49%

Formal funding

65%

61%

Note. Activities and formal structures that are discussed in the text but not in the table were not
a
assessed by these studies. This represents the proportion of teams that had protocols
delineating the response of various stakeholder groups (e.g., 57% had protocols re: the response
of rape victim advocates).
18
 

In summary, the way in which SARTs are structured varies a great deal from community
to community. While this section has provided insight into how SARTs are implemented in
practice, the next question that must be asked is, are these interventions actually effective at
achieving their aims?
Effectiveness of SARTs
Although SARTs have been lauded by many practitioners and policy-makers (e.g., DOJ,
2004; Ledray, 2001; Malefyt, Little, & Walker, 1998; PCAR 2002), research and evaluation of
SARTs has been unable to keep up with their rapid diffusion and the vast majority of SARTs
have not been formally evaluated (Zajac, 2009). The published research that does exist has
focused on SARTs’ effectiveness in three domains: improvements in the relationships among the
systems that respond to sexual assault, improvements in victims’ help-seeking experiences, and
improvements in legal cases outcomes. No published literature has reported on SARTs’
effectiveness at preventing perpetration or improving public awareness of sexual assault or
services for victims. See Greeson and Campbell (2013) for recent review of the effectiveness of
SARTs.
Relationships between systems. The extant literature on improving cross-system
6

relationships suggests that SARTs have the potential to improve communication and the quality
of relationships. In an evaluation of pilot SANE-SART programs in Illinois, stakeholders
reported that SART meetings facilitated information-sharing across systems, such as follow-up
about individual cases and updates about policy changes. The same evaluation also documented
                                                            
6

Pre-post studies that examined the impact of SANE and SART in communities where they
were implemented concurrently were excluded from this review, as it is impossible to tell
whether their findings are attributable to the SANE or SART component of the intervention.
19
 

improved relationships: SART members reported that SART meetings enhanced their
understanding of other members’ perspectives, and enabled them to make decisions collectively.
In Noble and colleagues’ study of a funding initiative to support 22 SARTs in California, SART
leaders noted that a benefit of the SART program was increased contact among team members,
which led to improved relationships (Noble et al., 2001). While these findings are encouraging,
the question remains whether SARTs are able to translate these improvements in relationships
into improved services for victims and increased offender accountability in the legal system.
Victims’ experiences. The few studies that have examined victims’ experiences in
communities with SARTs have indicated that these interventions can have a have a positive
impact on survivors’ help-seeking experiences. Campbell (1998) conducted a national study of
rape victim advocates and asked them to report on their most recent case. Cluster analysis was
used to capture victims’ experiences with the legal and medical systems, as reported by the
advocates. Survivors who received care in a community with a coordinated response to sexual
assault (all but one of which had a SART) were more likely to receive services and had the
highest fit between the services they desired and the services they received. In a follow-up
qualitative study with rape victim advocates, Campbell and Ahrens (1998) found that advocates
in communities with a SART believed that SARTs improved communication between survivors
and community systems (Campbell & Ahrens, 1998). In a more recent quasi-experimental study
comparing cases processed in a community with a SANE-SART, a SANE but no SART (SANEonly), and no-SANE/no-SART, Nugent-Borakove and colleagues found that survivors in the
SANE-SART community were offered more services than survivors in the comparison
communities (Burgess, Lewis-O’Connor, Nugent-Borakove, & Fanflik, 2006; Lewis-O’Connor,
2009; Nugent-Borakove et al., 2006). Finally, in Noble and colleagues’ study, leaders from seven
20
 

SARTs noted that a benefit of SARTs was preventing secondary trauma to victims during their
interactions with community systems (Noble et al., 2001). Thus, the extant literature, though
sparse, suggests that SARTs have the potential to improve victims’ experiences with community
systems.
Legal outcomes. Findings regarding SARTs’ impact on legal case outcomes are mixed.
Campbell’s national study (1998) found that survivors who received care in highly coordinated
communities (all but one of which had a SART) were more likely than survivors who received
care in an uncoordinated community to be classified in the cluster with the highest prosecution
rates. Nugent-Borakove’s study also found a positive impact of SART on proximal legal
outcomes; more specifically, quality of evidence collection, victim participation in the legal
system, and arrest and charging rates were significantly higher in the SANE-SART community
(Nugent-Borakove et al., 2006). On the other hand, in the same study, there were no significant
differences between the communities on more distal legal outcomes, specifically conviction rates
and sentence lengths, after controlling for site differences in case characteristics (NugentBorakove et al., 2006). Another study by Wilson and Klein (2005) examined the impact of a
SART on a variety of legal outcomes; they were unable to identify any significant differences
between SART and non-SART cases. However, the utility of this study is restricted by key
methodological limitations. The SART they examined is highly unusual in several ways (e.g., it
operates at the state rather than community level, victims choose whether to receive services
through a SART or non-SART model, and only victims who agree to cooperate with the criminal
justice system receive services) and may not be generalizable to other SARTs. In addition, the
authors acknowledged that their analyses were hampered by low statistical power.

21
 

In answer to the question—are these interventions effective at achieving their aims?—it
appears that SARTs are a “promising practice.” In particular, many SARTs have been able to
achieve improvements in relationships among stakeholders, victims’ help-seeking experiences,
and proximal legal outcomes (e.g., evidence collection, referral rates). However, descriptive
research shows that SARTs vary a great deal in practice and studies of their effectiveness have
not captured this heterogeneity. This is an important omission, as variation in SART
implementation may be related to variation in their effectiveness: some types of SARTs may be
more able to achieve positive effects, while other SART variants may be less beneficial. To
examine this issue of heterogeneity in practice in more detail, the next section will provide an
overview of diffusion of innovation theory (Rogers, 2003) and its applications to SARTs.
Diffusion of Innovation Theory
Diffusion of innovation (DOI) theory explores the adoption and implementation of
“innovations,” which are defined as ideas, practices, or objects that are perceived as new by
potential users (Rogers, 2003, pp. 12). It has been utilized in both the natural and social sciences,
and has been applied to a variety of types of innovations including social fads, technological
advances, and social programs/interventions. Within community psychology, DOI has been
applied to understand the processes of adoption and implementation within the context of
community-based interventions (Mayer & Davidson, 2000; Emshoff et al., 2003; Hazel &
Onaga, 2003; Rodríguez, Baumann, & Schwartz, 2010). More specifically, DOI has traditionally
been applied to the diffusion of evidence-based interventions, in which an intervention model
was empirically tested, and then disseminated.
According to DOI, as innovations are adopted by new users, a process termed
“reinvention” may occur. Reinvention occurs when an innovation is modified by those who
22
 

adopt and implement it (Rogers, 2003). In the diffusion of interventions into community settings,
common examples of reinvention include discarding or altering components of the intervention
or adding to the intervention (Mayer & Davidson, 2000). There are many incentives for adopters
to engage in reinvention of interventions, such as creating local buy-in and ownership as well as
ensuring the intervention is appropriate to the local population and the setting within which it is
implemented (Dearing, 2009; Dearing, Maibach & Buller, 2006; Emshoff et al., 2003; Mayer &
Davidson, 2000). As such, modifying interventions during implementation in community
settings is quite common (Dearing, 2009; Durlak & Dupre, 2008; Mayer & Davidson, 2000).
Through the process of reinvention, new forms of the intervention emerge and are put into
practice.
In the past, there has been a lively debate in the literature regarding whether adopters
should focus on achieving complete “fidelity” to the original model (i.e., implementing the
intervention as similarly to the model that was originally tested, without reinvention) or engage
in “adaptation” of the intervention (i.e., strategic reinvention of the intervention to maximize fit
with the environment in which it is being implemented; Dearing et al., 2006; Mayer & Davidson,
2000). The process of reinvention may have important consequences regarding the effectiveness
of community-located interventions. On the one hand, if an intervention has been evaluated and
concluded to be effective, changing parts of the intervention during implementation in a new
setting may dilute its effectiveness (Durlak & Dupre, 2008; Hazel & Onaga, 2003; Mayer &
Davidson, 2000). On the other hand, an intervention may have differential effects in diverse
contexts, and adapting the intervention so that it is more appropriate to the local context (e.g., the
local target population, the organizational setting implementing the intervention, etc.) may
enhance the effectiveness of the intervention (or even be necessary to achieve positive results),
23
 

and may also promote local ownership and sustainability (Dearing, 2009; Durlak & Dupre, 2008;
Hazel & Onaga, 2003; Mayer & Davidson, 2000; Rodríguez, et al., 2010).
Over time, researchers have come to agree that adaptation and fidelity are not necessarily
incongruent; rather, a “planned adaptation” approach, which maintains fidelity to the “core
components” of the intervention, and utilizes thoughtful adaptation of other aspects of the
intervention is optimal (Dearing, 2009; Durlak & Dupre, 2003; Emshoff, 2008; Hazel & Onaga,
2003; Lee, Atschul, & Mowbray, 2008; Rodríguez et al., 2010). From this vantage point, there
are certain “core components” of an intervention model that must be implemented in order for
the intervention to achieve positive results (e.g., dosage; see Emshoff et al., 2003). On the other
hand, other aspects of an intervention (e.g., specific language in handouts provided to
participants; see Emshoff et al., 2003), may not be as critical to success and therefore can be
adapted by the local community without deleterious effects; hopefully, such adaptations will
actually increase the intervention’s effectiveness and sustainability through optimal
intervention/setting coupling.
These ideas have important implications for community practice. When implementing an
intervention, communities need to balance their desires to adapt the intervention model to their
local context with the need to maintain the integrity of the “core components” of the intervention
that are integral to its success. This of course means that communities need information
regarding which aspects of an intervention are core components that should not be altered and
which are open to adaptation (Dearing et al., 2006; Durlak & Dupre, 2003; Lee et al., 2008).
How then, does this relate to SARTs?
DOI, indigenous interventions, and SARTs. Traditionally, DOI has been applied to the
academic model of intervention development, in which an intervention model is developed by
24
 

academic experts (often in a university setting), empirically tested, and then disseminated to new
settings. When the original intervention is transferred to the new settings, reinvention and
adaptation occurs; ultimately, this results in different versions of the intervention being put into
practice. This process is fairly different from the development of “indigenous interventions” like
SARTs (Miller and Shinn, 2005). Indigenous interventions are developed by communities (rather
than academics) in the very settings where they were intended to be implemented, and often are
not empirically tested before being put into practice. This fits the development of SARTs. Rather
than academic experts creating an evidenced-based SART intervention model at the outset,
several communities recognized a need, and developed the first coordinated SART teams. Their
ideas spread, and new communities invented and reinvented new versions of SARTs. While the
development and dissemination of indigenous interventions (like SARTs), is quite different from
the traditional academic model of intervention dissemination, both types of interventions end up
in the same place: different versions of the intervention are put into practice in different settings.
DOI suggests that when different versions of an intervention are being put into practice in
different settings, (1) this has implications for effectiveness, and (2) it is critical to understand
which components of an intervention are critical for effectiveness and which can vary from
setting to setting without detracting from the intervention’s effectiveness. While DOI has
traditionally been applied to evidence-based interventions that were purposefully disseminated to
new settings, it seems that these implications are likely to also hold true for communitydeveloped interventions like SARTs, given that in both situations, different versions of an
intervention are being put into practice in different settings.
Indeed, this need to identify the “core components” of an intervention that are required
for effectiveness may be particularly strong for indigenous interventions like SARTs that
25
 

developed without the guidance of research evidence. A review of practitioner writings on what
constitutes a SART and which aspects of a SART must be implemented reveals a lack of
consensus in the field (Greeson, Parker, & Campbell, in progress). In practice, there is a great
deal of variation in SARTs’ membership, goals, activities, and infrastructure. DOI suggests that
attending to the implementation of the intervention is a critical part of understanding effective
utilization of the intervention in communities. Different SARTs models may be more or less
effective in different contexts. In order to more fully understand the effectiveness of SARTs and
inform community practice, it is necessary to learn about the different SART models that are in
practice, the differential effectiveness of these different models, and which aspects of SARTs
represent “core components” that are necessary to achieve positive results.
Of course, this begs the question of which features of SARTs are critical, and which
SART models (meaning different combinations of different SART features) are most effective.
While DOI tells us that differences in SARTs may impact their successes, this theory cannot tell
us which permutations of SARTs may be more or less effective. Which features of SARTs relate
to their ability to achieve positive collaboration among the multidisciplinary players and their
ability to translate positive collaboration into improvements in community-level outcomes? This
issue has not been examined in the SART literature. However, the broader scholarship on
effectiveness of collaborative interventions may provide valuable insight into SARTs.
Structural Factors Related to Collaborative Effectiveness
Overview of scholarship on collaboration. The purpose of this section is to review the
academic literature on factors that predict collaborative effectiveness and to apply this research
to the context of SARTs. Before doing so, it is necessary to limit the scope of the literature that
will be discussed. Currently, there are thousands of published manuscripts on the effectiveness of
26
 

collaborations; review articles and meta-analyses alone have identified hundreds of factors that
have been associated with collaborative effectiveness (for reviews see B. Berkowitz, 2001;
Bronstein, 2003; Butterfoss Goodman, & Wandersman, 1993; Foster-Fishman, S. Berkowitz,
Lounsbury, Jacobson, & Allen, 2001; Kreuter, Lezin, & Young, 2000; Mattesich, Murray-Close,
& Monsey, 2001; Roussos & Fawcett, 2000; Zakocs & Edwards, 2006). These factors span
multiple levels of analysis (e.g., characteristics of individual members, member organizations,
the collaborative itself, the broader community context) and encompass such diverse aspects of
collaborations as capacity building efforts, structure, process, and climate.
This review will focus specifically on structural factors that relate to collaborative
effectiveness, in particular, the members that are part of the collaboration, the specific change
strategies or activities that the collaboration engages, and the organization of the team (i.e., the
use of formal structures), as these issues are most pertinent to the SART field at this time. It is
evident from descriptive research that SARTs do vary across these dimensions, and therefore this
represents a fruitful area to explore how these differences in SARTs relate to their effectiveness.
Furthermore, SART guidelines and trainings focus a great deal on practitioners’ beliefs regarding
these aspects of SART operations, and some even go so far as to provide recommendations for
structuring SARTs (e.g., adopting formal structures, implementing various activities) as though
they are “best practices” despite a dearth of empirical evidence. For these reasons, this review of
the collaborative literature will focus on how structural factors, specifically membership,
activities, and use of formal structures and resources relate to collaborative effectiveness.
To enhance the relevance of this literature to the context of SARTs, this review will only
capture current literature (published in the past twenty years) on multidisciplinary community
collaboratives that seek to address or prevent social problems (e.g., for profit and international
27
 

collaboratives were excluded). More specifically, the review will focus on empirical literature on
domestic violence coordinating councils (DVCC’s), which seek to coordinate the efforts of
diverse stakeholder groups in the response to domestic violence and in many ways are highly
similar in purpose and structure to SARTs (see Allen, Watt, & Hess, 2008). This will be
supplemented with insights from reviews of the broader literature on collaboration. Given the
similarity between the purpose and functioning of DVCC’s and SARTs, these studies are most
likely to be relevant to the context of SARTs.
Membership. Collaboratives bring together representatives from diverse stakeholder
groups who are believed to have a common purpose or focus. As such, it is not surprising that
membership of different stakeholder groups may influence collaborative success. This is
important for understanding SARTs, as descriptive research illustrates that SARTs vary in their
membership. Reviews of the broader collaboration literature suggest that having diverse
membership (e.g., greater breadth of stakeholder groups involved) and appropriate membership
coverage (meaning that members are representative of critical stakeholder groups and/or
populations in the community based on the target issue) are positively associated with
collaborative success (Butterfoss et al., 1993; Foster-Fishman et al., 2001; Kreuter et al., 2000;
Mattesich et al. 2001; Roussos & Fawcett, 2000).
Studies of domestic violence coordinating councils (DVCC’s) have also examined the
influence of membership diversity and coverage. In a study of 43 DVCC’s in one state, Allen
(2005, 2006) found that councils with broader active membership (i.e., larger number of
stakeholder groups represented) were perceived to be more effective at achieving their goals,
including criminal justice reform as well as human service, social service, and educational
reform. In another study of domestic violence coordinating councils (DVCCs) conducted by
28
 

Allen and colleagues (2010) in a different state, quantitative and qualitative methods supported
prior findings from the collaboration literature. Surveys of DVCC members showed that broader
active membership within a council contributed to perceived effectiveness at achieving
institutionalized changed via promoting perceived social capital among members. In addition,
qualitative interviews with stakeholders from three purposively sampled “successful” councils
revealed that council members believed that broad membership, particularly from key
stakeholder groups, was critical to promoting social capital and increasing institutionalized
change (Allen et al., 2010).
In another study of DVCCs, Nowell (2006) examined several aspects of membership
diversity as a predictor of the members’ perceptions of their council’s effectiveness at promoting
coordination. Unlike prior studies, Nowell did not find a relationship between stakeholder
breadth and members’ perceptions of their success at achieving their goals (perhaps because she
measured the breadth of stakeholder groups represented by all members, rather than active
members). Other aspects of membership diversity were also examined. Results indicate that
gender diversity (more balanced proportions of men and women members) was positively
associated with coordination effectiveness, while greater sectorial diversity (e.g., for-profit vs.
non-profit) was negatively associated with effectiveness at increasing coordination. Other
aspects of diversity such as length of membership in the council, and levels of management
represented) were not predictive of perceived effectiveness. Although there have been many
ways of examining the representation of various groups among collaborative members, the
pattern across studies indicates that a broad active membership that includes key groups is
optimal This suggests that in order to understand the impact of variation in SARTs membership
on their effectiveness, it is important to examine both the number of stakeholder groups that are
29
 

represented by active SART members, as well as whether critical stakeholder groups are
represented (in the case of SARTs—victim advocates, police, prosecutors, and medical/forensic
examiners).
Activities. SARTs also vary in the particular activities or change strategies they utilize to
achieve their goals. Although the collaboration literature does not speak directly to the impact of
specific activities that SART utilize, the literature provides some insight into the types of
activities that are likely to be the most beneficial. The literature indicates that collaboratives are
most successful when their strategies for change are innovative (i.e., addresses an unmet need)
and fit the local context and the types of change they hope to facilitate (Foster-Fishman et al.,
2001). While this literature has focused very little on the impact of specific activities, there is
some evidence that that can help us to think about the work of SARTs. In Allen and colleagues’
qualitative work with purposively sampled “exemplar” councils (2010), members reported that
development of trainings and protocols through the multidisciplinary group helped contributed to
their ability to create change in their communities. Similarly another qualitative study found that
council members believed that facilitating training, particularly cross-training between
stakeholder groups was particularly important to improving the community response to domestic
violence (Clark, Burt, Schulte, & Maguire, 1996). Given the similarities between DVCCs and
SARTs, such activities should also be considered in future empirical work on SARTs.
Use of formal structures. Additionally, SARTs differ in the extent to which their
teamwork is formalized. While there has been a push toward more formalization of SARTs, this
has yet to be supported by empirical research on SARTs. Reviews of the collaboration literature
do typically suggest that a more formalized collaborative structure, with specified roles and
responsibilities, official statements of purpose (e.g., mission statements, written goals and
30
 

objectives), and formalized structures and processes for organizing group work and
communication (e.g., meeting agendas and minutes, subcommittees, procedures for conflict
resolution) is most effective (Butterfoss et al., 1993; Foster-Fishman et al., 2001; Mattesich et al.,
2001; Kreuter et al., 2000; Zackocs & Edwards, 2006). However, within studies of DVCC’s,
findings are less clear, as two studies have shown a positive impact of formality on council
effectiveness while one was unable to replicate this finding. In Allen and colleagues most recent
study (2010), formalized council structure contributed to perceived effectiveness via an increase
in social capital. Similarly, a case study of several DVCC’s conducted by Clark and colleagues
(1996) concluded that councils needed to be structured and well-organized in order to engage
stakeholders effectively. However, Allen’s (2005) study of DVCCs measured formalization of
structure and did not find a significant association with perceived effectiveness. She argued that
the contribution of formalization may depend on the influence of other council factors, such as
council size; this implies a potential moderating effect such that for certain councils (e.g., large
councils) formalization may have a positive impact while for others (e.g., small councils) it may
not. Overall, the lack of consistency across studies suggests that the impact of formalization on
collaborative success warrants further study within the context of SARTs.
A review by Mattesich and colleagues (2001) concluded that in addition to formalized
roles and procedures, a flexible, adaptive structure that can be altered in response to changing
circumstances without compromising its ability to meet the group’s needs is optimal. In Allen’s
qualitative interviews with purposefully sampled exemplar councils, members believed the
adaptability of their council structure as facilitating their successes (Allen et al. 2010). In
addition, incorporating various resources (e.g., a coordinator position, funding, paid staff) into
the council structure may be beneficial to councils. Foster-Fishman and colleagues in their
31
 

review (2001) concluded that “sufficient resources” was an indicator of collaboration’s ability to
achieve their goals, and similarly Roussos and Fawcett (2000) argued that staff and financial
support help sustain collaborative partnerships. It does not appear that the influence of such
resources has been studies within the context of DVCCs. Taken together, studies of DVCC’s and
reviews of the collaboration literature have identified various aspects of a collaboration’s use of
formal structures and resources that may be relevant to the effectiveness collaboratives, and in
turn, SARTs.
Summary. In sum, the broader collaboration literature and studies specific to DVCC’s
reveal that collaborative membership, activities, and utilization of formal structures and
resources that are salient to collaborations’ effectiveness. These constructs are likely to be
important in the context of SART and merit further study.
CURRENT STUDY
The legal, medical, and mental health/advocacy systems are community resources that
are available to meet sexual assault victims’ post-assault needs. Historically, there have many
problems associated with the way these systems respond to sexual assault. In recognition of these
problems, communities have developed SARTs to promote cross-disciplinary relationships and
coordination in the hopes that this would lead to improvements in community-level outcomes,
particularly victims’ help-seeking experiences and prosecution rates. While SARTs are widely
utilized and considered by many to be a “best-practice,” there is limited empirical evidence
regarding their impact on communities.
The few studies that have examined the effectiveness of SARTs—while limited in
number and mixed in methodological quality—suggest that the SART model holds promise.
Some SARTs demonstrate improved cross-disciplinary communication and relationships,
32
 

including a better understanding of one another’s roles. Research findings also illustrate that
SARTs can contribute to improvements in victims’ help-seeking experiences, such as increased
referrals to needed services, and reductions in secondary victimization by system personnel.
Additionally, studies show SARTs may be able to improve proximal legal outcomes (e.g.,
referral rates), but it is less clear whether they can ultimately improve conviction rates.
This body of research has two noteworthy limitations. First, descriptive research on
SARTs have utilized convenience sampling, which is problematic because it likely that nonrandom methods have over- or under-represented certain types of SART (e.g., it is likely they
over-sampled highly formalized SARTs, which are more likely to be in contact with the
organizations that solicited participation). As such, researchers, practitioners, and policy-makers
do not have a representative picture of the SART landscape. Such information is necessary to
help stakeholders decided how best to meet the needs of U.S. SARTs and advance the field as a
whole. A second limitation is that the literature has yet to examine how variation in SARTs’
structure may relate to their effectiveness. This is important because descriptive research
demonstrates that SARTs vary in their structure (specifically their membership breadth, and
implementation of formal structures, and collaborative activities) and these differences may
translate into differences in effectiveness. Indeed, DOI theory suggests the importance of
identifying which aspects of an intervention are “core components” that are critical to
effectiveness and should not vary across different settings.
In light of these needs, a national study was conducted of 172 U.S. SARTs. The current
study examined how variations in SART structure related to SART perceived effectiveness at
improving legal outcomes and improving victims’ help-seeking experiences within a random
sample of U.S. SARTs. Consistent with prior studies of DVCC’s, this study captured SART
33
 

leaders’ perceptions of SARTs’ effectiveness at improving victim and legal outcomes. Although
it would have been desirable to collect archival data to document changes in legal outcomes,
such as prosecution rates, this was not a feasible strategy in a national scale study of nearly two
hundred communities. For example, in a recent study, it took two years of data collection to
obtain such information in only two counties (see Campbell, Greeson, Bybee, & Kennedy,
2011). Collecting self-report data from survivors to assess changes to help-seeking experiences
would also be valid indicators of SART effectiveness. Again, prior work suggests this strategy
would not have been feasible in a study of this scope. A recent study in this literature required
one year of data collection to obtain interviews with twenty rape survivors in one community
(see Campbell, Bybee, Ford, Patterson, & Ferrell, 2009). Therefore, the current study asked
SART members to report on their perceptions of their SART’s effectiveness, which is a common
strategy in studies that are interested in comparing collaboratives to understand the factors that
are predictive of (perceived) collaborative effectiveness (Allen, 2005; Allen et al., 2010; Nowell,
2006). Consistent with prior studies of DVCCs, SART leaders were asked to act as key
informants, by providing information about both their perceptions of the SARTs’ effectiveness,
and the structural characteristics of their SART (Allen, 2005; Allen et al., 2010; Nowell, 2006).
To understand the influence of SARTs’ structure on their perceived effectiveness, two
aspects of SARTs’ structure were examined: SART membership breadth, and SARTs’ use of
formal structures (formalization) and collaborative activities (specifically case review,
multidisciplinary cross-trainings, policy and protocol development and review, and formal
program evaluation). Prior descriptive research (which utilized convenience sampling) indicates
that SARTs vary on these dimensions. Furthermore, research on DVCCs indicates that
formalization and membership breadth are related to perceived effectiveness. While prior
34
 

research has not examined the influence of specific collaborative activities like cross-disciplinary
training or policy and protocol development, SART manuals indicate that practitioners in the
field believes such activities should be implemented, and therefore it was important to examine
this third feature of SARTs structure.
SARTs’ use of formal structures and collaborative activities were captured in a unique
way. Prior studies have typically examined the relationship between different characteristics of a
collaboration and its effectiveness in isolation from one another (i.e., different characteristics are
treated as individual predictors of effectiveness). However, collaboration is an interactive,
dynamic process and it seemed unlikely that SARTs’ utilization of formal structures to organize
their team (formalization) and their utilization of formal collaborative activities to structure their
group process would operate in isolation from one another. Rather, it seemed more plausible that
each of these aspects of SARTs’ operations are interdependent and that a SART’s profile across
these characteristics is more meaningful for SART effectiveness than examining each
characteristic individually. Therefore, in order to capture SARTs’ heterogeneity in their use of
formal structures and collaborative activities, cluster analysis was used to identify clusters or
subgroups of SARTs with similar profiles. Cluster analysis is well-suited to addressing patterns
of heterogeneity. Cluster analysis allows an analyst to identify subgroups of cases within a
sample that are similar to one another, and different from cases in the other groups (Luke, 2005).
In this study, cluster analysis was used to identify clusters (or subgroups) of SARTs, such that
SARTs in one cluster have similar profiles on the variables of interest, but different profiles from
SARTs in the other clusters. In this way, the current study identified subgroups of SARTs that
have similar profiles based on their use of formal structures and resources (formalization) and
their use of key collaborative activities (specifically case review, multidisciplinary cross35
 

trainings, policy and protocol development and review, and formal program evaluation). Thus,
cluster analysis was a more holistic approach to assessing the influence of these factors on
collaborative effectiveness.
In summary, the current study had three aims. This first aim of the study was to use
random sampling methods to obtain a nationally representative sample of SARTs to assess their
structural characteristics. Three specific characteristics were examined: (a) the breadth of
membership across different sexual assault stakeholder groups (b) their formalization, or use of
formal resources and structures to organize their work; and (c) their utilization of various
collaborative activities. The second aim of the study was to use cluster analysis to empirically
derive groups of SARTs with similar profiles based on their formalization and use of various
collaborative activities. It was anticipated that through cluster analysis, this study would identify
meaningful clusters that represent differences in SARTs’ use of formal structures and
collaborative activities. The third aim of the study was to examine the relationship between
SARTs’ structural characteristics and perceived effectiveness at achieving improvements in legal
outcomes and victims’ help-seeking experiences. This (third) aim had two associated research
questions:
1. Is SART cluster membership (or SART type based on profile across their use of formal
structures and collaborative processes) associated with (a) perceived effectiveness at
improving legal outcomes and (b) perceived effectiveness at improving victims’ helpseeking experiences?
2. Is membership breadth associated with perceived effectiveness (a) improving legal
outcomes and (b) improving victims’ help-seeking experiences?

36
 

Given the exploratory nature of this study, specific hypotheses were not developed. However, it
was anticipated that SART clusters would be related to perceived effectiveness.

37
 

METHOD
Sampling
The target population of this study was U.S. Sexual Assault Response Teams. At the
inception of this project, there was no national list of U.S. SARTs to use as a sampling frame.
Therefore, the initial task in this project was to develop a sampling frame through the use of five
strategies. First, the research team was granted access to two national registries of organizations
that are likely to participate in SARTs: the International Association of Forensic Nurses’ (IAFN)
registry of Sexual Assault Nurse Examiner programs and SANE-SART.com’s registry of SANESART member organizations. Between these two registries 785 non-duplicative organizations
were identified. By phone and email, the research team was able to successfully contact 99%
(781) of these organizations (i.e., the team was able to determine whether a SART existed in
their community or was able to verify from another source that the organization was no longer in
existence). Second, the research team contacted subscribers to the National Sexual Violence
Resource Center’s SART list-serv. Individuals subscribed to this list-serv who could be matched
to a SART that was already identified in the first step of sampling frame development
(contacting members of the IAFN and SANE-SART registries) were not contacted. Of the
remaining 126 subscribers, over 98% were successfully reached by the research team. Third,
state sexual assault coalitions were contacted to find out if they had a list of SARTs in their state.
Forty-six of 51 possible coalitions were reached (90% success rate). Of these coalitions, 18
(39%) did not have a list of SARTs, 19 (41%) had a list of SARTs in their state, and 9 (20%) did
not have a formal list of SARTs, but identified SARTs they were aware of in their state for the
research team. Fourth, the research team conducted a search of the academic literature and
practitioner SART manuals to identify published material on existing SARTs. Finally, a Google
38
 

search was conducted to identify SARTs that were promoted on the Internet. As expected, many
SARTs were identified by multiple strategies. After removing duplicate entries, a total of 858
SARTs were identified. (During the interviewing process, an additional six SARTs were
identified that should have been included in the initial sampling frame; these were added to the
sampling frame and also had the opportunity to be selected for interviewing). Thus, a total of 864
unique SARTs comprised the study’s sampling frame.
To be eligible for the interviews, the SART had to engage in multidisciplinary
coordination in the response to sexual assault, meet regularly, and respond to adult sexual assault
victims. SARTs that only served child sexual assault victims were excluded from the study. The
target sample size for the study was 170 SARTs; ultimately interviews were completed with 172
teams. A power analysis revealed that this sample size would ensure reasonable statistical power
(greater than 0.80; as recommended by Cohen, 1992) to detect a medium-sized effect of SART
clusters on perceived effectiveness (assuming a reasonably small number of clusters and
reasonably evenly distributed cluster sizes). Simple random sampling was used to select SARTs
for recruitment. See recruitment procedures (below) for information on response rates.
Procedures
Recruitment. Trained graduate research assistants were responsible for contacting SART
key informants to verify the SART was eligible for the study and to recruit them to participate in
a structured phone interview. Similar to prior studies of domestic violence coordinating councils,
the leader or coordinator of the sampled SART was selected to participate on behalf of their team
as a “key informant.” Leaders were presumed to be well placed to provide information about
their team’s structure and functioning. However, when the SART leader was either unavailable
(e.g., one was on medical leave), had become a member of the SART so recently that they felt
39
 

they would not be a good informant, or was unwilling to participate (because they were the
leader of multiple independent SARTs and had already been interviewed about a different
SART), the most recent leader or a long-term member of the SART was asked to participate. In
96% of the interviews, the participant was the current or former leader of the SART.
Recruitment contacts were made by phone, and when an email address was available, by
email. To increase the likelihood of successfully contacting the SART key informant,
recruitment attempts were made at different times of the day and different times of the week.
Recruitment calls, messages, and emails focused on establishing the validity of the project,
showing the relevance of the project to their work, the confidentiality of their participation, and
how the study’s findings would be used to inform resources and practice for SARTs. Initially,
recruitment numbers were low, so an additional recruitment strategy was added. Non-responders
(who had been contacted six or more times without responding) received a recruitment mailing
with a $5 incentive, a recruitment letter signed by the research team and the National Sexual
Violence Resource Center, and a letter of support from the Director of the National Institute of
Justice7. SART key informants who agreed to participate were scheduled for an appointment
with an interviewer and received a reminder phone call or email the day before the interview.
Recruitment efforts were closely monitored throughout the study by the project director.
When SARTs refused to participate or were ineligible for the study, more SARTs were
sampled (using simple random sampling) in order to ensure the target sample size was reached.
Out of a sampling frame of 864 teams, 268 were recruited to participate in the study. Of the 268
that were recruited, 81 were successfully reached and deemed to be ineligible, because they did
                                                            
7

After that point, all interview participants received the $5 incentive, either in the mailing or
after the interview.  
40
 

8

not serve adult victims, the SART did not exist, or the SART had stopped meeting . Of the
remaining 187 teams (268 recruited – 81 reached and ineligible to participate = 187), 11 were not
reached successfully (6%; either they were never spoken to, or never able to schedule and
complete an interview), four refused to participate (2%), and interviews were completed with
172 teams (92%). Thus, by defining the response rate as: [number of interviews
completed]/number of potential interviews that were eligible and/or may have been eligible], the
response rate in the current study was 92% (see Figure 1 on next page).
Interviewer training. The project director was responsible for interviewer training. To
help the interviewers understand the context in which SARTs operate, interviewers read SART
manuals and were trained in the community response to rape. Interviewers were also trained on
the purpose of the study and each of the sections of the interview, administration of the interview
(including probing and answer questions), building rapport with participants, and ethical
considerations. In addition, interviewers were trained to administer the interview consistently,
while still engaging on a personal level with the participants. More specifically, interviewers
were trained to read the introduction, questions, and transitions word for word. However, they
                                                            
8

Sampling frame development was designed conservatively, in order to reduce the likelihood of
non-coverage bias (i.e., sampling error due to not including SARTs that should have been part of
the sampling frame). Due to these conservative procedures, it was expected that some teams that
were identified in sampling frame development would not be eligible for interviews. As one
example, the research team began screening out teams that only served children after sampling
frame development had been completed. In addition, three of the five sampling frame
development methods (contacting state coalitions, searching the academic and practitioner
literatures for references to SARTs, and Google searching) did not involve contacting people
who were potential members of the actual SART, and were therefore prone to identifying teams
that had been in existence at one point, but may not have been in existence when the sampling
frame was being developed.
41
 

were also trained to engage with the participant on a more personal level (e.g., by indicating that
they are listening, thanking the participant, acknowledging frustrations, etc.) without using
language that might bias participants’ answers (e.g., that sounds great, that’s very unusual). This
strategy balanced standardization with the need to build rapport with participants so that they
would be motivated to complete the entire interview thoughtfully and honestly.
Figure 1:
Sampling Flow Chart
Sampling Frame
N = 864 SARTs

Contacted

Deemed Ineligible

n = 268 SARTs

n = 81

Retained
n = 187

Unable to Successfully
Contact

Interviewed

Refused to Participate

n = 172

n=4

n = 11
After this initial formal training, interviewers reviewed the entire interview. As a team,
the interviewers and project director discussed what each question was intended to capture, the
operational definition of various terms, and how they should answer questions regarding each
item. Then the trainees observed the trainer conduct a mock interview. The mock interview
modeled appropriate administration of the interview and provided examples of how to respond to
42
 

potential challenges (e.g., a participant that refuses to select an answer from the response set).
Then, the trainees conducted a mock interview with another interviewer-in-training, and then
conducted a mock interview with the trainer. After they interviewed the trainer, they received
constructive feedback and were either cleared to begin interviewing or assigned additional
practice. This trained also used these mock interviews to ensure interviewers were consistent in
their administration of the interview, as well as their response to several issues they were likely
to face during “real world” interviewing (e.g., the participant not wanting to choose from one of
the response options). This training approach has been successfully utilized in prior studies of
sexual assault victims and sexual assault service providers (Campbell, Adams, Wasco, Ahrens, &
Sefl, 2009).
Piloting. Pilot interviews were conducted with 12 SART leaders. The project director
reviewed all of these interviews to identify any measurement problems, such as questions that
were difficult to answer; missing filter questions or questions that led to a high degree of notapplicable responses; ambiguous or misunderstood terms; inadequate response sets, and
indications of participant frustration and/or fatigue. Piloting was also used to address issues with
the overall organization, flow, and timing of the interview and any issues with staff recruitment
and administration of the interview. The revised interview protocol is provided in Appendix A.
Interviewing procedures. The interview began with the consent process. The consent
process opened with the interviewer describing the purpose of the study, what participation in the
study entailed, and the potential risks and benefits of participation. The interviewer then
discussed the participants’ rights as a research participant, the study’s procedures for protecting
confidentiality and privacy, and contact information should the participant wish to contact the
faculty supervisor or the university IRB. At the end of the consent process, the interviewer asked
43
 

the participant if they had any questions and informed them that they indicate their voluntary
consent to participate in the study by continuing with the phone interview. Participants were also
be given the opportunity to receive a copy of the consent form if they wished.
During the structured interview, the interviewer read the introduction, transitions, and
questions from the interview protocol word-for-word. The interviewer was also responsible for
orienting the participant to choose an appropriate response from the response set as needed (e.g.,
“You said this is a goal of your team. Would you say it is somewhat important, very important,
or a primary goal?”), and answering the participants’ questions (i.e., clarifying the interpretation
of an item).
At the end of the interview, participants were offered the opportunity to receive a brief
summary of results. Interviews typically lasted a little over one hour (M = 68.08 minutes, SD =
16.74). With permission, all interviews were audio-recorded. During the interview, the
interviewer recorded participants’ answers on the hard copy of the interview. Data was entered
from these hard copies into SPSS by trained research assistants.
Weekly meetings were held to monitor interviewing quality and troubleshoot
interviewing and data entry. Questions and issues regarding interviewing and data entry were
also documented in the project logbook and discussed at team meetings. Quantitative data were
double-entered, meaning they were independently entered into two separate databases and
discrepancies between the two databases were checked and addressed by a graduate research
assistant.
Data storage and human subjects protection. All research assistants completed the
required university IRB training and received additional training on ethics and protecting privacy
and confidentiality within the context of our study. Confidentiality was protected at the
44
 

individual and community levels so that neither the participant nor the SART they belong to can
be identified. Hard copies of the interview data and recordings are kept in a locked file cabinet in
the locked university office. Electronic copies of the audio-recordings and all electronic
databases that include identifying data are stored on the project’s password protected sub-folder
of the university’s secure network server. Only project staff has access to the data. Upon
completion of the study, all links to identifiable information will be destroyed. Consistent with
university policy, hard copies of the data will be destroyed three years after closure of the
project. All procedures were approved by the Michigan State University IRB.
Measurement
In drafting the phone interview protocol and measures, the project director drew upon
personal experience working with multidisciplinary community stakeholders in the response to
sexual violence; SART practitioner manuals and guidelines; survey research and case studies of
SARTs that described their implementation; prior studies of domestic violence coordinating
councils (Allen, 2005; Allen et al., 2010; Nowell, 2006); and a national study that was conducted
in order to assess the heterogeneity of SANE programs (Campbell et al., 2006). Feedback from
the Associate Director of the National Sexual Violence Resource Center was solicited and
incorporated into interview revisions.
The interview captured the independent (SARTs’ membership breadth, formalization and
utilization of various collaborative activities) and dependent variables (perceived effectiveness at
improving legal outcomes and improving survivors’ help-seeking experiences). Several
contextual variables were also be captured and modeled as control variables in the analyses.
Specifics regarding the measurement of each of the variables in the study are provided below.
Descriptives for the measures of interest are provided in the results section.
45
 

Contextual variables. Contextual information was also be collected and modeled as
control variables in analyses. During the interview, the participant was asked to report on the
length of time the SART has been in operation (in years) their stakeholder group (e.g., whether
they are a medical/forensic examiner, police officer, etc.), the amount of time they have been a
member of the SART and their race/ethnicity, gender, age, and education level. Community-level
demographics (specifically, the number of counties/jurisdictions served, and whether the
community classified as rural (based on 2010 Census population density of less than 500 people
per square mile) were also be collected to be included in the analyses as control variables.
Membership breadth (independent variable). Participants were asked to indicate
(yes/no) whether each of 18 different stakeholder groups were currently represented as an active
member on their SART team. An active member is defined as “someone who attends SART
meetings or other functions reasonably frequently.” Stakeholder groups included rape crisis
center staff, prosecutor, crime lab personnel, corrections, clergy, higher education, and others.
Membership breadth was measured as the total number of stakeholder groups that are
represented on the SART (the sum of the 18 yes/no items).
Formalization (independent variable). Formalization was measured by 13 yes/no
questions. Participants were asked to report whether various formal structures and resources
(e.g., subcommittees, meeting agendas, a formal leader, and formal procedures of conflict
resolution) were in place in their SART (yes/no). Formalization was the sum of these 13 yes/no
items. This measure drew from the work of both Butterfoss (1998) and Allen and colleagues
(2010). Butterfoss’s original measure was adapted for use with DVCC’s by Allen and
colleagues. In turn, their measure of DVCC formalization was adapted for this study to capture
use of formal structures and resources in SARTs. In adapting the measure for this study, items
46
 

that were irrelevant in the context of SARTs were cut, several items were clarified to ensure they
would be clear to SART members, and several formal resources that are often used by SARTs
were added to the measure.
Use of collaborative activities (independent variables). Participants were asked about
SARTs’ use of four key collaborative activities: case review, multidisciplinary cross-trainings,
adoption and/or review of policies/protocols, and program evaluation. These specific activities
were chosen based on a review of SART manuals and the team’s experience with the community
response to sexual assault. For each activity, SARTs were asked whether their SART engaged in
that activity (yes/no), and if so whether the activity was utilized regularly or as needed. For the
variable program evaluation, SARTs were coded as 0 = does not engage in evaluation or 1 =
engages in program evaluation. For each of the other four activities (case review,
multidisciplinary cross-trainings and adoption and/or review of policies/protocols) a SART was
coded as 0 = does not use that activity, 1 = uses activity on an as needed basis, and 2 = used on a
regular basis. Thus, each activity was represented with its own variable.
Perceived effectiveness variables (dependent variables). Participants were asked to
rate the extent to which their SART’s efforts led to a variety of improvements in their
community on a five point Likert-type scale (1 = Not all; 2 = A little bit; 3 = Somewhat; 4 =
Quite a bit; 5 = To a great extent). The response set, and the structure of the scale (i.e., asking
participants to rate the extent to which the collaborative contributed to various improvements)
are modified from Allen and colleagues’ (2010) measure of perceived domestic violence
coordinating council effectiveness. The specific aspects of SARTs’ perceived effectiveness that
were captured (e.g., asking about improving victims’ willingness to participate in the criminal
justice process throughout the entirety of their case) were developed specifically for this study by
47
 

the project director and the PI based on their experience with the community response to sexual
assault.
Perceived effectiveness at improving victims’ experiences was assessed by a 10-item
measure that captured improvements in survivors’ post-assault interactions with legal, medical,
mental health, and advocacy services related to the accessibility and responsiveness of these
systems to rape victims. Sample items included “victims being more likely to receive referrals to
mental health services” “responders being more sensitive toward victims” and “responders
providing more comprehensive, higher quality services to victims.” It was anticipated that the
index was one-dimensional. An exploratory factor analysis (oblique) was conducted and visual
inspection of the scree plot as well as the Kaiser rule confirmed one factor. Thus, future analyses
were conducted based on the average of the ten items (α = .92; corrected item-total correlations
ranged from .610 to .775).
Perceived legal effectiveness was assessed by 26 items that captured a variety of
improvements related to the legal processing of sexual assault cases. Two items were dropped
due to high levels of missing data (the extent to which fewer prosecuted cases were dismissed by
a grand jury, judge, or prosecute, 16.3% missing; and the extent to which prosecutors’ improved
their education of juries, 11.0% missing). It was initially anticipated that the remaining 24 items
captured three domains: improvements in prosecution rates, improvements in criminal justice
system procedures and practice, and improvements in victims’ participation in the criminal
justice system. However, a confirmatory factor analysis was conducted to test this structure and
2

exhibited poor model fit (χ (296)= 837.5, CFI = .851, RMSEA = .103). This indicated that the
hypothesized factor structure did not adequately reflect the structure of the data.

48
 

An exploratory factor analysis (EFA) using oblique rotation was then used to determine
the structure of the items. Inspection of the scree plot suggested a two or three factor solution,
and the Kaiser rule indicated three factors; therefore the three factor solution was chosen. Items
were assigned to the factor that they loaded to the most strongly and subscale scores were created
by averaging across the items that were assigned to each factor (see Appendix B for a
presentation of which items were classified into which subscales). This resulted in three
subscales of perceived legal effectiveness. First, the “improvements in sexual assault victims’
participation in the criminal justice system” subscale was assessed by nine items. Sample items
included the extent to which the SARTs’ efforts have led to “an increase in the number of
assaults that are reported to the police” and “victims being more likely to continue participating
during the entire case.” (α = .95; corrected item total correlations ranged from .693 to .853).
Second, the improvements in police processing of sexual assault cases subscale consisted of six
items, including included the extent to which the SARTs’ efforts have led to “police being more
likely to refer cases to the prosecutor’s office” and “improvements in police utilization of
medical/forensic evidence” (α = .90; corrected item total correlation ranged from .624 to .792).
The third subscale, improvements related to the prosecution of sexual assault cases was assessed
by nine items. Example items include the extent to which the SARTs’ efforts have led to “an
increase in the number of prosecuted cases that result in conviction” and “improvements in
medical forensic personel expert witness testimony” (α = .95; corrected item total correlations
ranged from .713 to .850). 
Data Analyses
Cluster analysis. Cluster analysis was used to identify clusters or sub-groups of SARTs.
Cluster analysis is an analytic technique that is useful for identifying clusters or subgroups of
49
 

cases based on their profile across multiple variables (Luke, 2005; Rapkin & Luke, 1993).
Cluster analysis creates groupings of cases such that cases within a group are more similar to one
another on the variables of interest than they are to cases in a different group. In this case,
SARTs were grouped based on their scores on five variables: (1) formalization, (2) whether they
used program evaluation, and their institutionalization of (3) case review, (4) multidisciplinary
cross-training, and (5) policy and protocol development and review (see Figure 2).
Figure 2:
Cluster Analysis
Formalization
Institutionalization of Case Review
Institutionalization of Cross-Trainings
Institutionalization of Policy/Protocol Development and Review
Utilization of Program Evaluation 

SART Clusters

To prepare for clustering, data were inspected for out of range values and implausible
descriptive statistics in order to identify errors in data entry and coding. All five variables were
standardized prior to clustering in order to ensure that some variables would not be artificially
influential due to their scaling (if two variables with different scales are included in the analysis,
the variable with the larger scale would be more influential on the solution; Rapkin & Luke,
1993). High correlations among variables in a cluster analysis can also impact the results of the
analysis and therefore correlations among the five variables to be clustered were examined.
50
 

Correlations ranged from non-significant (r = .004, n.s.) to moderate (r = .395, p < .01) and
therefore no modifications were needed. Only three cases were missing data on any of the
clustering variables and therefore pairwise deletion was used (leaving the sample size intact).
In contrast to traditional hypothesis-testing statistical methods in which possible results
are constrained to rejecting or not rejecting the null hypothesis, cluster analysis is more
inductive. As cases are clustered, the statistical technique minimizes heterogeneity within
clusters and maximizes heterogeneity across clusters (Rapkin & Luke, 1993). However, cluster
analysis can group the cases into one to N groups (in this study, 1 to 172 groups) and the
researcher must decide which number of groups represents the best solution. As is typical, this
decision was informed by both statistical considerations (e.g., number of cases per cluster,
number of clusters, degree of heterogeneity/homogeneity in the cluster solution) and conceptual
considerations (e.g., are the clusters capturing groups that are conceptually meaningful and
interpretable?; Rapkin & Luke, 1993). In this study, a two-step procedure utilizing both
hierarchical and optimization cluster analysis methods was used.
First, a hierarchical (agglomerative) cluster analysis using Ward’s linkage method was
conducted to determine the appropriate number of clusters. In this method, the analysis begins
with N clusters (in this study 172 clusters), of one case each. The analysis creates a proximity
matrix that represents the extent to which each cluster (in this iteration, each consisting of one
case) is similar/dissimilar to the other clusters. Squared Euclidean distances (which are
appropriate for both categorical and continuous variables) were used to assess the degree of
similarity/dissimilarity among clusters (Rapkin & Luke, 1993). Then, the analysis combines (or
agglomerates) the two clusters that are the most similar to one another into one cluster consisting
of two cases. After that, a new proximity matrix of n-1 clusters (in this case 171) is created.
51
 

Next, the two clusters that are the most similar are combined. The process continues until all
cases have been combined into one cluster of N (172) cases.
The results provide a dendogram that visually displays each step of the agglomeration,
specifically the cases or clusters that were combined during each step, and how much
heterogeneity was added to the cluster that was combined during that step. The results also
include a list of heterogeneity coefficient for each agglomeration that indicates the amount of
heterogeneity at each agglomeration. An inverse scree plot of the heterogeneity coefficients and
visual inspection of the dendogram allow the analyst to look for unusually large jumps in
heterogeneity—such steep jumps indicate the last agglomeration increased heterogeneity a
disproportionate amount and therefore the cluster solution prior (i.e., involving one more cluster)
may have been preferable (Clatworthy, Buick, Hankins, Weinman & Horne, 2005). After the
hierarchical cluster analysis was interpreted to determine the appropriate number of clusters,
optimization cluster analysis was conducted.
In optimization cluster analysis, the analyst specifies the number of clusters a-priori
(Clatworthy et al., 2005; Rapkin & Luke, 1993). The statistical program generates a solution
consisting of that many clusters, and classifies cases into clusters in a way that maximizes the
homogeneity within clusters and minimizes the heterogeneity across clusters. When the
interpretation of hierarchical cluster analysis is not completely definitive and allows for different
numbers of clusters, optimization cluster analysis is conducted for both. Next, optimization
cluster analysis was conducted and centroids from the corresponding hierarchical cluster analysis
were used as cluster start values. The results of the optimization cluster analysis were used to
classify SARTs into clusters.

52
 

Structural equation modeling. Structural Equation Modeling (SEM) was conducted to
examine structural characteristics as predictors of SARTs’ effectiveness. SEM is a flexible
approach which allows the analyst to test a theorized model of relationships among variables.
Categorical and continuous variables can be included in the model and unidirectional (predictive)
and correlational relationships among variables can be assessed (Tabachnik & Fidell, 2007). A
key advantage of SEM over multiple regression is that SEM models can include multiple
dependent variables (and allow the analyst to assess the relationship between them) within the
same model (Tabachnik & Fidell, 2007). The original version of the model is shown in Figure 3.
Figure 3:
Model of SART Structure and Perceived Effectiveness with Control Variables
SART Structure
Membership Breadth
Cluster Membership

Perceived Effectiveness
Improvements in Victims’ Help-Seeking
Experiences

Contextual Control
Variables
SART Length of Operation

Legal: Improvements in Victims’
Participation in the Criminal Justice
System

Length of Participant SART
Membership

Legal: Improvements in Police Processing
of Sexual Assault Cases

Serves a Rural County

Legal: Improvements related to the
Prosecution of Sexual Assault Cases

Number of Counties Served
Participant Stakeholder Group
Participant Demographics
 

 

53
 

 

The two multinomial variables (SART cluster membership and the participant’s

stakeholder group) were dummy coded. The interview participant’s stakeholder group was coded
into three groups: rape crisis center staff (55%), medical/forensic examiners (24%), and other
(20%), with rape crisis center staff, the largest group coded as the reference group.
To prepare for modeling, bivariate associations between the variables were examined,
and data were screened for outliers, non-normality, multicollinearity among the independent
variables, and non-linear relationships between the independent and dependent variables (as
recommended by Tabachnik & Fidell, 2007). This led to several modifications. First, none of the
participant demographics (age, gender, race, and education level) were correlated with any of the
four measures of perceived effectiveness and therefore these were not included as control
variables in the path model. Second, the length of time the participant had been involved in the
SART was strongly correlated with the length of time the SART had been in operation (r = .70, p
< .01); this correlation would cause multicollinearity if both variables were modeled as
predictors of the perceived effectiveness variables. The literature suggests the importance of the
length of time that a collaborative has been in operation as fundamental to its ability to succeed.
Therefore, the SARTs’ length of continuous operation was retained, and the participant’s length
of membership in the SART was dropped. Third, the distribution of the “number of counties
served” variable was non-normal. Therefore, the variable was dichotomized into whether the
SART served more than 1 county (yes/no). Fourth, the length of time the SART had been in
operation exhibited a non-linear relationship with each of the dependent variables. Inspection of
the bivariate scatter plots revealed that a quadratic line was a good fit for the data. Therefore, the
original variable (length of time in operation, centered) and the variable squared (length of time
in operation centered, squared) were modeled as predictors of the dependent variables. Finally,
54
 

four outliers were identified. Analyses were conducted with and without these cases to assess
their influence on the results. The revised model is pictured in Figure 4. In the initial path model,
all nine independent variables were modeled as predictors of each of the four dependent
variables and correlations were allowed between all independent variables and between the error
terms of the dependent variables. Full information maximum likelihood estimation (FIML) was
used to address the missing data in these analyses.
Figure 4:
Revised Full Path Model

SART Structure
Membership Breadth
Cluster Membership (Dummy)

Perceived Effectiveness
Improvements in Victims’ Help-Seeking
Experiences

Contextual Control
Variables

Legal: Improvements in Victims’
Participation in the Criminal Justice
System

SART Length of Operation
(Linear)

Legal: Improvements in Police Processing
of Sexual Assault Cases

SART Length of Operation
(Quadratic)

Legal: Improvements related to the
Prosecution of Sexual Assault Cases

Serves a Rural County
Serves more than 1 County
Participant Stakeholder Group
(Dummy)

A key indicator of the quality of a path model is model fit. Model fit refers to the extent
to which the variance covariance matrix that is generated based on the model specified by the
55
 

researcher is similar to the observed variance covariance matrix (Tabachnik & Fidell, 2007).
Poor model fit indicates that the theorized model is not well-suited to explaining the data. In this
study, overall model fit was assessed by significance of the Chi-square test of misfit (a
statistically significant Chi-Squared indicates a significant amount of misfit between the
observed and modeled matrices) and by the CFI and RMSEA indices of model fit (see Hu &
Bentler, 2003 for discussion and benchmarks). To conserve statistical power and increase
parsimony, non-significant parameters were sequentially trimmed from the model (see Hosmer &
Lemeshow, 2000). Chi-square difference tests were conducted to ensure the model trimming

process did not introduce a significant amount of misfit.
RESULTS
Descriptives: Participant and SART Characteristics
The first aim of the study was to document SARTs’ structural characteristics in practice,
specifically their membership breadth, formalization, and use of collaborative activities. In
addition to presenting this information, this section provides additional contextual information on
the SART key informants who participated in the interviews, the SARTs themselves and the
communities they serve, and SARTs perceived effectiveness.
Participants. SART key informants who participated in interviews were typically female
(95% female, 5% male), White (89% White, 5% African-American, 2% Hispanic/Latino(a), 2%
Native American or Alaskan Native, 1% Multiracial, and 0% Asian-American), middle-aged (M
= 42.46 years old, SD = 10.89), and college educated (84% had completed a bachelor’s degree or
higher). Over half were rape crisis center advocates/staff (55%), while approximately one-fourth
were medical/forensic examiners/SANEs (24%); 5% were advocates employed in the victim’s
witness unit of the prosecutor’s office, 5% were prosecutors, 4% were police, and 7% identified
56
 

as another stakeholder group (e.g., prevention educator, child advocacy center director). On
average, participants had been involved in their SART for five years (M = 5.08, SD = 4.40).
SART contextual characteristics. On average, SARTs had been in operation
continuously for eight years (M = 8.08 years, SD = 5.63). Approximately one-fifth served more
than one county (19%), while two-thirds (66%) served a rural community (with a 2010 census
population density of less than 500 people per square mile). As to geographic region, 31% of the
SARTs were located in the South, 29% in the Midwest, 25% in the West, and 15% in the
Northeast. The majority of SARTs’ active membership included police (98%), rape crisis center
advocates/staff (95%), Sexual Assault Nurse Examiners (90%), and prosecutors (85%). Threequarters of SARTs (75%) had active members from all four of these groups.
Structural characteristics of SARTs. In regards to membership breadth, out of 18
possible stakeholder groups, SARTs averaged 8.50 different stakeholder groups as active
members of their team (SD = 2.65; minimum = 2, maximum = 15). SARTs were moderately
formalized, utilizing an average of 6.26 out of 13 possible formal structures from the modified
formalization index (SD = 2.28, minimum = one, maximum = 12). SARTs varied in their use
and institutionalization of collaborative activities. The majority of teams engaged in case review
(74%), multidisciplinary cross-trainings (76%), and policy/protocol development and review
(82%). However, these collaborative activities were not typically institutionalized as a regular
SART process: 42% engaged in case review regularly, 38% had multidisciplinary cross-trainings
regularly, and 31% engaged in policy/protocol development or review on a regular basis. Only
15% engaged in program evaluation to assess their SART’s functioning.
SARTs’ perceived effectiveness. SART key informants reported relatively high levels of
perceived effectiveness. Average effectiveness ratings were highest for perceived effectiveness
57
 

at improvements in victims’ help-seeking experiences (M = 3.72, SD = .81), followed by
improvements related to police processing of sexual assault cases (M = 3.50, SD = .94),
improvements related to increasing victims’ participation in the criminal justice system (M =
3.40, SD = .86), and improvements related to the prosecution of sexual assault cases (M = 3.28,
SD = 1.02).
Clustering SARTs on their use of Formal Structures and Collaborative Activities
The second primary aim of the study was to employ cluster analysis to derive groups of
SARTs with similar profiles based on their use of formal structures and collaborative activities.
More specifically, SARTs were clustered on their formalization; their institutionalization of case
review, policy/protocol development, and multidisciplinary trainings; and whether or not they
engaged in program evaluation. Bivariate correlations among these five clustering variables are
presented in Table 2.
Table 2:
Bivariate Correlations among the Clustering Variables
1

2

3

4

Formalization
Institutionalization of Case Review

.115

Institutionalization of Cross-Trainings

.245**

.106

Institutionalization of Policy/Protocol Development .105
and Review

.079

.395**

Utilization of Program Evaluation

.004

.238**

.203**

** p < .01
58
 

.015

A hierarchical analysis was conducted to determine the appropriate number of clusters for
optimization cluster analysis. In order to choose the number of clusters, the inverse scree plot
and dendogram were inspected; these results were taken into consideration, along with the desire
to maintain a reasonable number of clusters. Based on these considerations, a three or four
cluster solution seemed best. Therefore, optimization cluster analysis was conducted for both
three and four clusters. Results of the three and four cluster solutions were compared to evaluate
their conceptual and statistical merits. Both the three and four cluster solutions had reasonable
cluster sizes, and clear differences between clusters. The three cluster solution was chosen based
on interpretability because it provided a more meaningful, interpretable picture of different
patterns of SARTs’ utilization of formal structures and collaborative activities. See Appendix C
for comparison of the three and four cluster solutions.
SARTs were then classified into one of the three clusters according to their profile on the
five clustering variables, representing SARTs’ use of formal structures and specific collaborative
activities. See Table 3 for descriptive information on SARTs’ formalization and use of various
collaborative activities by cluster. Table 3 also presents ANOVA/Chi-Squared results for each of
the five clustered variables; the ANVOAs/Chi-Squared tests assessed whether there were
statistically significant differences between clusters on each of the variables. Four of the five
clustering variables (with the exception of case review) varied by cluster, suggesting the clusters
did capture meaningful variability in SARTs’ use of formal structures and collaborative
activities. Next, the features of the three clusters will be described; for each cluster, a case
example will be discussed to illustrate the salient features of that cluster.
 

59
 

Table 3:
Cluster Analysis of SARTs’ Use of Formal Structures and Collaborative Activities Results
“Low
Adopters”

“High Adopters “High Adopters ANOVAs/Chi4
Except
Plus
Squared Tests
Evaluation”
Evaluation”

Cluster size

n=65

n=80

n=27

---

Formalization (0-13)

Low

High

High

F(2,169) =
21.08**

a3

b

b

M = 4.98

M = 7.44

SD = 1.74
Institutionalization of
1
case review (0-2)

M = 6.90

SD = 1.98

SD = 2.83

Average

Average

Average

a

a

a

F(2,169) =
2.10, n.s.

b

F(2,168) =
68.72**

c

F(2,167) =
44.61**

M = 1.00

M = 1.28

M = 1.19

SD = .79

SD = .81

SD = .83

Institutionalization of
multidisciplinary cross1
trainings (0-2)

Low

High

High

a

b

M = .48

M = 1.54

M = 1.58

SD = .61

SD = .53

SD = .64

Institutionalization of
policy/protocol
development and review
1
(0-2)

Low

High

Medium

Program evaluation
2
(0-1)

a

b

M = .64

M = 1.53

M = 1.15

SD = .54

SD = .53

SD = .67

Low

Low

High

a

a

M=0

SD = 0
1

M=0

SD = 0

b

M=1

χ2(2) =
170.00**

SD = 0
2

0 = activity not utilized; 1 = used as needed; 2 = used regularly; 0 = does not use program
3

evaluation; 1 = uses program evaluation; Within a row, means with the same letter superscript
indicate no statistically significant differences between groups based on Tukey’s HSD post-hoc
tests. Within a row, means with different superscripts indicate statistically significant differences
4
between groups based on Tukey’s HSD post-hoc tests; Cluster analysis selects for low variation
within clusters and high variation across clusters. Therefore, results of ANOVAs/Chi-Squared
should be interpreted in the context of describing the results of the cluster analysis only.
** p < .01

60
 

Cluster one. Cluster one comprised 38% of the sample and was named “Low Adopters”
due to generally lower use of formal structures and institutionalization of collaborative activities
than the other two clusters.
The sample “Low Adopter” SART serves two adjacent urban counties in the
Western/Rocky Mountain region. The group was formed in 2005 to expand help
strengthen the SANE program and create teamwork between various groups that respond
to sexual assault. At the time of the interview, five different stakeholder groups
participated in the team, including rape crisis center staff, SANEs, police, and advocates
from the victim’s witness unit in the prosecutor’s office (sample M = 8.50). They do not
have a regular meeting schedule; instead, they hold collaborative meetings on an as
needed basis. Out of a possible 13 formal structures and resources, this SART utilizes
only four: a formal mission statement, meeting agendas, meeting sign-in sheets, and
meeting minutes. The team does not have a formally elected leader, but a representative
from the SANE program calls their collaborative meetings. The team uses policies and
protocols, which are reviewed on an as needed basis and they also engage in
multidisciplinary case review on an as needed basis. They do not engage in
multidisciplinary cross-trainings and they do not engage in evaluation of their
coordinated efforts.
Cluster two. Cluster two, the largest cluster, included 47% of the sample of SARTs and
was termed “High Adopters except Program Evaluation” due to generally high use of formal
structures and institutionalization of collaborative activities, but no utilization of program
evaluation.

61
 

The sample “High Adopters except Program Evaluation” SART serves a rural
Midwestern county. Their SART formed in 1999, but stopped meeting for quite a few
years. They reformed, and at the time of the interview had been in operation continuously
for three years. The SART involves eight different stakeholder groups, including the
primary stakeholder groups (SANEs, police, prosecutors, rape crisis center staff) as well
as the local child advocacy center for child abuse and neglect victims, the local domestic
violence program, advocates from the victims’ witness unit in the prosecutor’s office, and
representatives from other social service agencies (sample M = 8.50). They utilize seven
out of 13 possible formal structures/resources to organize their team: the team has a
formal leader position, a mission statement, structured meetings (with agendas, sign-in
sheets, and meeting minutes), subcommittees, and a formal mechanism for holding SART
members accountable to the groups. On a quarterly basis, they engage in
multidisciplinary case review and review their policies/protocols, and twice a year they
conduct cross-trainings. This SART does not engage in program evaluation of their
coordinated efforts.
Cluster three. Finally, cluster three (16% of the sample) was called “High Adopters plus
Program Evaluation.” SARTs in this group had generally high use of formal structures and
institutionalization of collaborative activities and all engaged in program evaluation.
A sample “High Adopters plus Program Evaluation” SART serves a rural Southwestern
county. The group has been in operation continuously since 2003, and has 15 different
active stakeholder groups. Their members range from primary sexual assault responders
(police, rape crisis center staff, SANEs, and police) to clergy, judges, local schools, and
staff from corrections. They utilize nine formal structures and resources to organize their
62
 

team (out of 13 possible). They have a formal leader, a mission statement, bylaws, and
structured meetings (with meeting agendas, sign-in sheets, and meeting minutes). In
addition, they also organize their SART through formal procedures for decision-making,
conflict resolution, and holding SART members accountable to the broader group. They
engage in case review quarterly, multidisciplinary trainings twice a year, and review their
policies/protocols on an annual basis. They also utilize program evaluation: after each
sexual assault case, survivors and sexual assault responders complete surveys about
services that were provided, and their accessibility.
Predicting SARTs’ Perceived Effectiveness
The third aim of the study was to examine the relationships between SARTs’ structural
characteristics (membership breadth and SART cluster based on their use of formal structures
and activities) and SARTs’ perceived effectiveness at achieving improvements in legal outcomes
and victims’ help-seeking experiences. An SEM path model was tested to assess these
relationships, after controlling for contextual variables (SART length of continuous operation,
characteristics of the community served by the SART, and the sexual assault stakeholder group
the interviewee belonged to). To test the effect of SART cluster membership on perceived
effectiveness, SARTs in the “Low Adopters” cluster and SARTs in the “High Adopters except
Program Evaluation” cluster were compared to the reference group: SARTs in the “High
Adopters with Program Evaluation” cluster. The characteristics of this cluster are the most
similar to what the practitioner literature characterizes as best practice for SARTs. Therefore, the
dummy coding was designed allowed for comparisons against this cluster.
First, the full, saturated model (with the nine independent variables predicting all four
dependent variables, correlations modeled allowed between the independent variables and
63
 

between the error terms of the dependent variables; see Figure 4 on pp. 53) was tested. Then, the
model was trimmed such that non-significant paths were sequentially cut from the model to
9

increase model parsimony and conserve statistical power. Paths, that when trimmed altered the
pattern of results, were retained. Tables 4, 5, and 6 give the results of the final, trimmed model.
The model exhibited excellent fit and model trimming did not introduce a significant amount of
2

misfit based on the likelihood ratio test (χ (21) = 28.13, n.s.; CFI = .995, RMSEA = .045).
Analyses were replicated with and without the four cases that were outliers on the variables of
interest. The pattern of results did not change and therefore these cases were retained in the
sample.
The model accounted for 24% of the variance in perceived improvements in victims’
help-seeking experiences, 30% of the variance in perceived improvements in victims’
participation in the criminal justice system, 27% of the variance in perceived improvements in
police processing of sexual assault cases, and 28% of the variance in perceive improvements
related to the prosecution of sexual assault cases. Modeled correlations among the independent
variables and correlations among the error terms of the dependent variables are presented in
Tables 4 and 5 respectively. Standardized regression coefficients for predictive relationships
between the independent variables and the four dependent variables are provided in Table 6.
 

                                                            
9

A power analysis was conducted and determined that a sample of 172 would provide
reasonable statistical power to detect an effect of SART cluster membership on perceived
effectiveness.
64

 

Table 4:
Correlations among Predictors of SARTs’ Perceived Effectiveness from the Final Path Model
1
Membership
Breadth

0

Low Adopters
1
Cluster

0

SART Length of
Operation (Linear)

0

SART Length of
Operation
(Quadratic)

0

Serves a Rural
County

0

Serves more than
1 County

0

Participant
Stakeholder
Group: Medical/
Forensic
2
Examiner

0

Participant
Stakeholder
2
Group: Other

0

3

4

5

6

7

8

---

High Adopters
Except Evaluation
1
Cluster

2

3

---

3

-.727

---

3

.157

-.265

---

3

.063

-.021

.400

---

3

-.129

.242

-.177

-.108

3

3

3

3

0

3

0

3

0

3

3

0

0

3

3

0

0

3

3

0

0

3

--3

0

---

3

-.179

.162

---

3

.055

-.088

-.286

0

0

0

1

Italicized = significant at p < .05; Reference group = “High Adopters Plus Program
2

Evaluation” Cluster; Reference group = Participant Stakeholder Group: Rape Crisis Center
3

Staff; Parameter was non-significant and then constrained to zero in the model.
 

65
 

Table 5:
Correlations among Residuals of the Perceived Effectiveness (Dependent) Variables from the
Final Path Model
1

2

3

1. Residual: Perceived
Effectiveness at Improving
Victim’s Help-Seeking
Experiences
2. Residual: Perceived
Effectiveness at Improving
Victims’ Participation in the
Criminal Justice System

.713**

3. Residual: Perceived
Effectiveness at
Improvements in Police
Processing of Sexual Assault
Cases

.676**

.677**

4. Residual: Perceived
Effectiveness at
Improvements related to the
Prosecution of Sexual
Assault Cases

.566**

.632**

** p < .01

66
 

.722**

Table 6:
Predictors of SARTs’ Perceived Effectiveness from the Final Path Model
Perceived
Effectiveness at
Improving Victims’
Help-Seeking
Experiences
β

p value

Perceived
Effectiveness at
Improving Victims’
Participation in the
Criminal Justice
System
β

p value

Perceived
Perceived Effectiveness
Effectiveness at
at Improvements
Improving Police
Related to Prosecution
Processing of Sexual
of Sexual Assault
Assault Cases
Cases
β

p value

β

p value

Membership Breadth

.105

n.s.

.181

p < .01

.243

p < .01

.293

p < .01

Cluster: High Adopters
1
Except Program Evaluation

-.104

n.s.

-.285

p < .01

-.031

n.s.

-.128

n.s.

-.340

p < .01

-.497

p < .01

-.210

p < .05

-.219

p < .05

SART Length of Operation
(Linear)

.273

p < .01

.264

p < .01

.299

p < .01

.303

p < .01

SART Length of Operation
(Quadratic)

-.235

p < .01

-.222

p < .01

-.214

p < .01

-.196

p < .01

Serves a Rural County

.004

n.s.

.056

n.s.

.155

p < .05

.124

n.s.

Serves more than 1 County

.021

n.s.

.146

p < .01

.139

p < .05

.129

n.s.

Participant Stakeholder
Group: Medical/Forensic
2
Examiner

.133

n.s.

.082

n.s.

.218

p < .01

.236

p < .01

-.082

n.s.

-.012

n.s.

.072

n.s.

.078

n.s.

Cluster: Low Adopters

1

1

Participant Stakeholder
2
Group: Other

2

Reference group = High Adopters plus Program Evaluation Cluster; Reference group = Participant Stakeholder Group: Rape Crisis
Center Staff.
67
 

 

Improving victims’ help-seeking experiences. This dependent variable, “improving
victims’ help-seeking experiences,” captured perceptions of SARTs’ effectiveness with respect
to creating improvements related to victims’ experiences of seeking help from the legal, medical,
mental health, and advocacy systems post-assault. After controlling for the contextual variables
of interest, membership breadth was not related to SARTs’ perceived effectiveness at improving
victims’ help-seeking experiences. SARTs in the “Low Adopters” cluster were perceived as less
effective at contributing to improvements in victims help-seeking experiences than SARTs in the
“High All” cluster (β = -.349, p < .01); however, there was not a statistically significant
difference between SARTs in the “High All” and the “High Except Evaluation” clusters on this
dependent variable. Both the linear and quadric terms for SARTs’ length of continuous operation
were related to perceived effectiveness at improving victims’ help-seeking experiences (β = .273,
p < .01; β = -.235, p < .01). This indicates a curvilinear relationship, such that generally SARTs
that had been in operation for a longer period of time were perceived as more effective, with the
effect declining for the oldest SARTs. Calculations conducted using the unstandardized
coefficients in the predictive regression equation indicate that the effect of the length of time the
SART has been in operation on victims’ help-seeking experiences becomes zero when SARTs
reach approximately 17.74 years of operation. The two community context variables (whether
the SART served multiple counties and whether the SART served a rural county) were not
related to perceived effectiveness at improvements in victims’ help-seeking experiences. Finally,
the stakeholder group the SART key informant belonged to was not related to their perceptions
of the SART’s effectiveness at improving victims’ help-seeking experiences.
Improving victims’ participation in the criminal justice system. This dependent
variable captured SARTs’ perceived effectiveness at improvements related to increasing sexual
68
 

 

assault victims’ participation in the criminal justice system. After controlling for the other
variables in the model, membership breadth was significantly positively associated with SARTs’
perceived effectiveness at contributing to improvements related to victims’ participation in the
criminal justice system (β = .181, p < .01). Results indicated that SARTs that had more different
types of stakeholder groups actively involved in the team were perceived as more effective in
this domain. Both SARTs in the “Low Adopters” cluster and SARTs in the “High Adopters
except Evaluation” cluster were perceived as significantly less effective at contributing to
improvements in victims’ criminal justice system participation than SARTs in the “High
Adopters plus Evaluation” cluster (β = -.497, p < .01; β = -.285, p < .01). Both the linear and
quadric terms for SARTs’ length of continuous operation were related to perceived effectiveness
at improvements related to victims’ participation in the criminal justice system (β = .264, p < .01;
β = -.222, p < .01). Generally SARTs that have been in operation for a longer period of time are
perceived as more effective on this domain, with the effect declining for the oldest SARTs.
Calculations conducted using the unstandardized coefficients in the predictive regression
equation indicate that the effect of the length of time the SART has been in operation on victims’
participation becomes zero when SARTs reach approximately 17.74 years of operation. As to
community context, whether or not the SART served a rural community was not related to
perceived effectiveness; however, SARTs that served more than one county were perceived as
more effective at improvements to victims’ participation in the criminal justice system (β = .146,
p < .01). Finally, the stakeholder group the SART key informant belonged to was not related to
their perceptions of the SART’s effectiveness at improving victims’ participation in the criminal
justice system.

69
 

 

Improvements related to police processing of sexual assault cases and improvements
related to the prosecution of sexual assault cases. A very similar pattern of results existed for
the last two dependent variables: perceived effectiveness related to improvements related police
processing and prosecution of sexual assault cases. Therefore, the two are presented together.
One captured SARTs’ perceived effectiveness at making improvements in police processing of
sexual assault cases through the criminal justice system, while the other captured SARTs’
perceived effectiveness and making improvements related to the criminal prosecution of sexual
assault cases.
After controlling for the other variables in the model, membership breadth was associated
with SARTs’ perceived effectiveness at contributing to improvements in both police processing
and the prosecution of sexual assault cases, with SARTs that had a greater number of different
active stakeholder groups perceived as more effective (β = .243, p < .01; β = .293, p < .01).
SARTs in the “Low Adopters” cluster were perceived as less effective at contributing to
improvements in both the police processing and prosecution of sexual assault cases than SARTs
in the “High Adopters plus Evaluation” cluster (β = -.210, p < .05; β = -.219, p < .05); however,
there was not a statistically significant difference between SARTs in the “High Adopters Plus
Evaluation” and the “High Adopters Except Evaluation” clusters (β = -.031, n. s.; β = -.128, n.
s.). Both the linear and quadric terms for SARTs’ length of continuous operation were related to
perceived effectiveness at improvements in police processing and prosecution of sexual assault
cases (police: β = .299, p < .01; β = -.214, p < .01; prosecution: β = .303, p < .01; β = -.196, p <
.01). SARTs in operation for a longer period of time tended to be perceived as more effective,
with the effect declining for the oldest SARTs. Calculations conducted using the unstandardized
coefficients in the predictive regression equations indicate that the effect of the length of time the
70
 

 

SART has been in operation on police processing of cases becomes zero when SARTs reach
approximately 19.99 years of operation and the effect on improvements related to prosecution
becomes zero when the SART has been in operation for 21.24 years. Both aspects of community
context was related to perceived effectiveness in improvements related to police work, such that
SARTs that served a rural community, and SARTs that served more than one county were
perceived as more effective (β = .135, p < .05.; β = .139, p < .05). However, neither community
context variable was related to SARTs’ perceived effectiveness at improvements related to the
prosecution of sexual assault cases. The key informant’s stakeholder group was also associated
with their perceptions of the SARTs’ effectiveness at improvements related to both police and
prosecution, with medical/forensic examiners rating the SARTs’ effectiveness more highly than
rape crisis center staff (β = .218, p < .01; β = .236, p < .01). There were no significant differences
between rape crisis center staff and key informants classified in the “other” category (i.e., neither
a medical/forensic examiner nor rape crisis center staff; β = .072, n.s.; β = .078, n.s.).
DISCUSSION
SARTs are community-level interventions that were designed to create collaboration
among key sexual assault responders and thereby improve the prosecution of sexual assault cases
and victims’ experiences of seeking-help post-assault. To date, there are hundreds of SARTs in
the U.S.; however, there is little empirical evidence of how these interventions are structured,
and how well they are able to achieve their goals in practice. Therefore, this current study had
three primary aims: (1) to understand SART’s structural characteristics within a national random
sample; specifically, their membership breadth, formalization, and use of collaborative activities;
(2) to use cluster analysis to empirically derive groups of SARTs with similar profiles based on
their formalization and use of various collaborative activities; and (3) to examine the
71
 

 

relationships between SARTs’ structural characteristics and SARTs’ perceived effectiveness at
improvements in victims’ help-seeking experiences and legal outcomes.
Aim One: SARTs’ Structural Characteristics
The first aim of this study was to examine the structural characteristics of SARTs’ in
practice. Consistent with prior descriptive research based on convenience sampling methods
(e.g., Zajac, 2006, 2009), the current study found that SARTs’ are not uniformly implemented
across the U.S. Rather, SARTs vary in their membership breadth, formalization, and use and
institutionalization of collaborative activities. This finding is consistent with DOI theory
(Rogers, 2005) which suggests that as an intervention is adopted in different contexts, adopters
and implementers engage in “reinvention” of the intervention; this process results in different
forms of the interventions being put into practice in different communities—much like our
findings that different forms of SARTs are being put into practice in different communities. A
key focus of the current study was to capture different ways of implementing SARTs.
Aim Two: Clustering SARTs on their Formalization and Collaborative Activities
To capture different methods of implementing SARTs, the second aim of the study was to
use cluster analysis to classify SARTs into distinct subgroups, representing different profiles in
their use of formal structures and collaborative activities to organize their work. Three groups of
SARTs were identified. SARTs in the “Low Adopters” cluster utilized fewer formal structures,
were less likely to institutionalize multidisciplinary trainings and policy/protocol review into
their group’s collaboration, and did not engage in program evaluation. Both of the other clusters
tended to utilize more formal structures and collaborative activities to organize their work. The
SARTs in the “High Adopters except Evaluation” cluster tended to use more formal structures,
and engaged in more institutionalization of multidisciplinary trainings and policy/protocol
72
 

 

review into their group’s collaboration; however, none of the SARTs in this cluster engaged in
program evaluation. The “High Adopters plus Evaluation” cluster also tended to use more formal
structures and engaged in more institutionalization of multidisciplinary trainings and
policy/protocol review; in addition, they also engaged in formal program evaluation to assess
their collaborative efforts. As predicted, the study was able to identify discernible and
interpretable subgroups of SARTs that varied in their implementation of formal structures and
collaborative activities to structure their work. In the next aim, these subgroups were examined
in relation to SARTs’ perceived effectiveness.
Aim Three: Predicting SARTs’ Perceived Effectiveness
The third aim of the study was to examine predictors of SARTs’ perceived effectiveness.
Specifically, SARTs’ cluster membership, breadth of membership, and contextual factors
(characteristics of the community served and the interview participant) were examined as
predictors of perceived effectiveness at improving legal outcomes and victims’ help-seeking
experiences. First, the influence of cluster membership on SARTs’ perceived effectiveness will
be discussed, followed by membership, and finally the contextual factors that were studied.
Cluster membership and perceived effectiveness. Consistent with what would be
expected from DOI theory and the literature on dissemination of innovations, the High Adopters
with Program Evaluation group was perceived as more effective than the Low Adopters group on
each of the four domains of effectiveness that were measured in this study. Prior research and
theory on the Diffusion of Innovations suggests retaining core elements of an intervention is key
to effectiveness (Dearing, 2009; Durlak & Dupre, 2003; Emshoff, 2008; Hazel & Onaga, 2003;
Lee et al., 2008; Rogers, 2005). Indeed, this study found that the SARTs in the “High Adopters
plus Program Evaluation” group—the group that most closely mirrors recommendations for how
73
 

 

SARTs should operate—tended to perceived themselves as more effective than the Low Adopter
SARTs.
In addition, the study found that SARTs in the “High Adopters plus Program Evaluation”
group tended to perceive themselves as more effective at contributing to improvements in
victims’ participation in the criminal justice system than SARTs in the “High Adopters except
Evaluation” cluster. This finding was somewhat unexpected; rather, it was expected that the
“High Adopters plus Evaluation” cluster would be perceived as more effective than the “High
Adopters except Evaluation” cluster across all four domains of effectiveness due to the addition
of program evaluation. As such, it was somewhat surprising that the inclusion of program
evaluation was associated with only one of the four possible outcomes (and not all four).
Program evaluation theory suggests that both the process of engaging in an evaluation, as well as
the substantive findings of an evaluation can lead to programmatic changes; in turn, these
changes can create improvements in the effectiveness of a program (Cousins & Whitmore, 1998;
Patton, 2008). However, the current study’s findings did not show a universal contribution of
program evaluation; rather, the “High Adopters plus Program Evaluation” cluster was perceived
as more effective than the “High Adopters except Program Evaluation cluster” on only one of the
four domains of effectiveness.
There are two plausible reasons for this pattern of results. First, the “High Adopters plus
Program Evaluation” cluster consisted of 27 SARTs. Thus, there may have been limited
statistical power to detect an effect of cluster membership (type II error) due to the size of that
cluster. Second, the key difference between the “High Adopters plus Program Evaluation”
SARTs and the “High Adopters except Program Evaluation” SARTs was whether or not they
conducted formal evaluation. The current study did not capture whether they used the process
74
 

 

and findings of those evaluation efforts to create programmatic changes. In other words, some
SARTs may have “gone through the motions” of evaluation, but may not have used the
evaluation to alter their SARTs. In fact, the evaluation literature verifies that quite frequently,
evaluations are conducted but do not translate into meaningful programmatic changes (Amo &
Cousins, 2007; Birkeland, Murphy-Graham, & Weiss, 2005; Cousins & Whitmore, 1998; Patton,
2008). Thus, simply conducting program evaluation may not have been sufficient to translate in
widespread improvements in all four domains of effectiveness.
Membership and effectiveness. Another critical aspect of a SART’s implementation is
its membership: the different stakeholder groups that participate in the coordinated team. The
current study found that SARTs with broader active membership from more stakeholder groups
had higher perceived effectiveness on all three forms of legal effectiveness; however there was
not a statistically significant relationship between membership breadth and perceived
effectiveness at improvements in victims’ help-seeking experiences. Prior literature on DVCC’s
suggests that broader active membership tends to be associated with higher perceived
effectiveness (Allen, 2005; Allen, 2006; Allen et al., 2010). Why in this study was breadth of
active membership only associated with the three forms of legal effectiveness? In the current
study, the primary sexual assault response groups (rape crisis center staff, police, prosecutors,
and SANEs) were most likely to be actively represented in SARTs. The remaining (“extra”)
stakeholder groups were less common, and therefore tended to be involved in SARTs with
broader membership, and not in SARTs with more narrow membership. In retrospect, these
“extra” stakeholder groups tend to be skewed toward the legal system. Many of those groups are
explicitly part of the criminal justice system and rarely deal with victims (e.g., probation and
parole, the crime lab), while others are not explicitly criminal justice focused, but have a great
75
 

 

deal of interaction with both the criminal justice system and victims (e.g., child advocacy
centers, domestic violence programs). Thus, the extra members that tended to be part of SARTs
with broader active membership tended to be associated more strongly with the legal response
than the response to victims. This may explain why broader active membership was associated
with higher perceived legal effectiveness, but was not significantly associated with perceived
effectiveness at improving victims’ help-seeking experiences.
Contextual features and effectiveness. The current study measured four additional
contextual features that may relate to SARTs’ perceived effectiveness: the length of time the
SART had been in operation consistently, two characteristics of the communities served by the
SARTs, and the role that the SART key informant played in the response to sexual assault. The
findings will be discussed and connected to prior literature in this order.
When considering the functioning of any collaboration, it is important to consider not
only how the group is structured, but also how long the group has been working together.
Consistent with scholarship that suggests community collaborations take time to develop
relationships and then translate those relationships into improvements in outcomes, this study
found positive associations between the length of time SARTs had been in operation
continuously and all four domains of their effectiveness, such that older SARTs were perceived
as more effective (Butterfoss et al., 1993; Roussos & Fawcett, 2000)10. However, this study
found a curvilinear (rather than linear) relationship between the collaboration’s age and its
perceived effectiveness. Generally, older SARTs are perceived to be more effective, with the
                                                            
10

It should be noted that this study used a cross-sectional design and therefore, does not provide
definitive evidence that SARTs do increase in effectiveness as they age. These findings may also
have been related to cohort effects (see subsequent paragraphs).
76
 

 

effect diminishing for the oldest SARTs; indeed, the effect of continuous length of operation
would have fell out around 17.74-21.24 years of operation (depending on the dependent
11

variable).

Given the cross-sectional nature of this study, it is difficult to disentangle the source

of these findings—there may be a cohort effect, or it may be that effectiveness changes as
SARTs mature (or there may be a cohort by maturation effect). It is possible that the oldest
SARTs, which represent the very first cohort of SARTs (i.e., the “first adopters” of the SART
model) are somewhat less effective than the next cohort of SARTs (“early adopters” but not first
adopters) because first adopters have to do the initial ground-breaking work, while early adopters
have the benefit of avoiding the initial ground-breaking and can learn from the first adopters. The
youngest, or most recent cohort of SARTs (“late adopters”) were perceived as least effective--perhaps these are the communities that had particularly challenging circumstances that
preventing them from adopting the SART intervention, and in turn, these circumstances inhibit
12

their effectiveness.

It is also possible that there is a maturation effect, such that generally, as SARTs reach a
certain age, they lose momentum, and there is a slight decrease in their effectiveness. In their
review of the literature on community collaborations, Foster-Fishman and colleagues (2001)
noted the importance of a collaborative having a “continuous learning orientation” in which they
                                                            
11

Due to the relatively small number of SARTs that had been in operation for over 17 years
(5%), these findings should be interpreted with caution.
12

 This finding was based on the amount of time the SART has been in operation continuously.
Late adopters could include SARTs that became a SART for the first time quite recently, or tried
having a SART before, failed, and recently re-adopted the SART model. Either way, it is
possible that challenging circumstances contributed to the need to adopt or re-adopt the SART
model quite recently. 
77

 

 

find new ways to improve and adapt to changing conditions. This may be a difficult level of
commitment to sustain over decades, and may help to explain why they were no was longer a
positive effect of length of operation for the oldest SARTs. Or, it may be that as SARTs age, the
way in which SART members perceive the effectiveness of their group shifts, with the oldest
SARTs having a somewhat less optimistic view of their functioning, and thereby making the
effect of the age of the SART diminish for the oldest SARTs.
In addition to characteristics of the SART itself, this study examined features of the
community the SART serves and their relationships with perceived effectiveness. Diffusion of
Innovation Theory suggests it is important to consider the environment in which an intervention
is implemented and how the environment impacts the intervention’s effectiveness (Rogers,
2005). Specifically, SARTs that serve rural communities were perceived as more effective at
creating improvements related to police processing of sexual assault cases and SARTs that
served multiple communities were perceived as more effective at improvements related to
prosecution and police processing of sexual assault cases.
SARTs that served rural communities may have been perceived as more effective at
improvements due to differences in community climate between rural and urban communities.
For example, Orbst and colleagues (2003) found that residents of rural communities reported a
stronger psychological sense of community (i.e., stronger sense of belonging, identification, and
influence in relationship to their community) than residents of urban communities. This in turn
may cause a greater willingness to engage in and commitment to collaborative efforts to improve
the community response to sexual assault. In addition, anecdotal accounts from open-ended
questions in these interviews suggested rural communities tended to be more tightly connected,
and this helped build positive multidisciplinary relationships. More specifically, many SART
78
 

 

leaders noted that doing their work in a small, tight-knit, rural community made it easier to
develop positive relationships; there was a sense of “everyone knows everyone else” in these
communities, and it was easier to build positive relationships from that starting point, rather than
starting from scratch.
In addition, rural communities will tend to have fewer police departments, and the
departments that exist will tend to be smaller. Such conditions may make it easier for rape crisis
centers and SANEs (who are typically responsible for creating and leadings SARTs) to have
sufficient time resources to develop positive working relationships with all of the different police
departments in their communities. In addition, the smaller size of the departments may make it
easier to reach all of the police officers in a community who respond to sexual assault in efforts
to improve individual officers’ practices in processing cases. Thus, differences in the number and
size of police department in smaller, rural vs. larger, urban communities may also account for
differences in the perceived effectiveness of rural and urban communities.
As to the number of counties served, results reveal that SARTs that serve multiple
communities tended to be perceived as more effective at improvements related to police
processing of sexual assault cases. This has two possible explanations. First, it may be that
SARTs that are particularly successfully at bringing police on board in one county—bolstered by
that success—decide to expand their efforts and move toward creating improvements in multiple
county. Secondly, it is possible that the actual act of expansion to additional counties helps
SARTs to succeed at creating perceived improvements related to police processing of sexual
assault cases. By adding additional counties, a SART will engage more people and organizations
per stakeholder group, and this will be particularly true of police departments. Such expansion
may create more opportunities for cross-fertilization within stakeholder groups across counties:
79
 

 

members from the same stakeholder groups from different communities can share strategies and
information with one another. In addition, a key function that collaboratives can serve is to create
accountability to engaging in the desired response to sexual assault (e.g., Allen et al., 2010).
SARTs may be particularly effective at holding an individual member accountable when other
members of the collaboration also come from the same stakeholder group. For example, the
sheriff department from county A may have a stronger desire to be perceived as a good
collaborator and responder and therefore be more accountable to the team when the sheriff from
county B is also part of the group. Thus, serving multiple counties may change the collaborative
dynamics within a SART.
Limitations and Implications for Future Research
The study’s findings should be interpreted in light of three notable methodological
limitations. First, this is the only study to date to have developed a sampling frame of U.S. and
use random sampling methods. Although sampling frame development involved a careful,
multistep process, it is possible that some U.S. SARTs were not identified by our study. In
particular, SARTs that are less connected to the field (e.g., are not promoted in the literature or
online, are not connected to other professional organizations in the field) may have been
inadvertently omitted from the sampling frame and therefore under-sampled. Although we used
multiple complementary strategies to identify SARTs and directly contacted local SANE
programs that were likely to participate in a SART, it is still possible that some SARTs were not
identified. It is possible that such SARTs may have been less connected and therefore also less
formalized and less likely to follow recommendations for how SARTs should operate. Thus, it is
possible that under-sampling our SARTs could have led to over-estimation of the formality and

80
 

 

use of collaboration activities within U.S. SARTs. Despite this limitation, this is by far the most
comprehensive attempt to obtain a nationally representative sample of SARTs to date.
Second, the current study was a broad national study to understand the structural
characteristics of SARTs in practice, and the implications of differences in structure for SARTs’
(perceived) effectiveness. As such the current study was primarily focused on SARTs’
structure; while this is an important characteristic of SARTs, other aspects of SARTs may also
be related to variations in SARTs’ effectiveness. SARTs may vary in their climate and process,
such as their ability to create an inclusive collaborative climate, the nature and quality of
multidisciplinary relationships and communication, the extent to which various team members
have a shared understanding of the problem and how to respond to it, the commitment and
leadership of SART members, and the extent to which the team is engaged in authentic reflection
and action to create sustainable changes in policy, practice, and procedures in their community,
etc. (see Foster-Fishman et al., 2001 for these and other features of collaborations that have been
associated with collaborative effectiveness). Future research on SARTs can expand from the
current study on SARTs’ structure, to provide a deeper understanding of SARTs’ innerworkings. In particular in-depth, multi-method ethnographic (including observations, archival
records, and stakeholder interviews) case studies with a small number of SARTs could be used
to: capture key climate and process-related features of the focal SARTs; examine how salient
climate and process-related characteristics unfolded, and why; understand the implications of
climate and process for SARTs’ effectiveness; and understand the mechanisms by which
bringing together diverse groups to work together results in improvements in victims’ helpseeking experiences and legal outcomes.

81
 

 

The third methodological limitation stems from measuring SART leaders’ perceptions of
their SART’s effectiveness. While asking collaboration members’ to report on their perceptions
of the collaboration’s effectiveness is standard in large-scale studies (e.g., Allen, 2005, 2006;
Allen et al., 2010; Nowell, 2006), it is possible that SART leaders may have under or overestimated the impact of their team on victims’ experiences seeking help post-assault and the
processing of sexual assault cases through the criminal justice system. In particular, it seems
likely that social desirability could have caused SART leaders to want their SART to seem
effective during the phone interviews. While measures were taken to reduce this bias (e.g.,
asking leaders to give us their honest opinions, avoiding interviewer responses that suggested the
SART was good or bad, noting the confidential nature of the study), it is likely that this still had
an influence on leaders’ responses and therefore, may have led to inflated reports of SARTs’
effectiveness.
To address problem in future research, there are several alternate measurement
approaches that could be used to study SARTs’ effectiveness. For example, data could be
collected from survivors (for example, on the services they received, their perceptions of the
accessibility and responsiveness of the systems they sought help from, their participation in the
criminal justice system, and their well-being and recovery). Archival records could be coded to
identify changes in practices in the response to sexual assault and changes in legal outcomes
such as availability of forensic evidence, arrest rates, conviction rates, and sentence lengths.
Observations could be conducted to document how services are provided and how criminal
processes unfold. However, interviewing survivors, obtaining archival records, and conducting
observations are extremely resource-intensive modes of data collection, and therefore were not
suitable for the current, national scale study. However, future small-scale case studies of SARTs
82
 

 

could improve our understanding of SARTs’ effectiveness, and the factors that promote or
detract from SARTs’ effectiveness by using such methods to capture SARTs’ effectiveness
directly, rather than SART members’ perceptions of effectiveness
Implications for Policy and Practice
Structure. Despite these limitations, this study has several important implications for
SART policy and practice. Diffusion of Innovation Theory (Rogers, 2005) and research on the
dissemination of interventions, suggests that ideally, when adopting an intervention in a new
setting, the “critical components” of the intervention should be retained, and other, less critical,
features of the intervention should be adapted to improve the intervention’s effectiveness in the
local context (Dearing, 2009; Durlak & Dupre, 2003; Emshoff, 2008; Hazel & Onaga, 2003;
Lee, Atschul, & Mowbray, 2008; Rodríguez et al., 2010). In this vein, the current study found
that SARTs that tended to be higher in formalization and their institutionalization of
collaborative activities (the “High Adopters plus Program Evaluation” cluster) were perceived as
more effective than SARTs that were lower in their formalization and use of collaborative
activities (“Low Adopters” cluster). This suggests that a “critical component” of SARTs’ is
having enough structure to organize the SARTs’ teamwork. Collaboration is a shifting dynamic
process: members join and leave the team, goals shift, ideas for what constitutes best practice
evolve, resources change, and so on. In this changing environment, through formalization and
institutionalization of collaborative activities, the SART should have a “backbone” that supports
them. The team members know what their process is for working together and they have
structures in place to keep themselves focused on making progress on their collaborative goals.
In turn, SARTs that are less formalized and only engage in collaborative activities occasionally
may find themselves unsure of how it is they work together, and may have to spend substantial
83
 

 

time each time they meet as a group figuring out what they are trying to achieve and how they
will navigate working together.
One of the implications, then, is that SARTs should organize their teamwork through
formal structures/resource and institutionalizing collaborative activities. Another implication is
that organizations and funders that support SARTs’ efforts should help them to do so. Typically,
SART members are sexual assault responders, and are unlikely to have had substantial training
and/or experience in building multidisciplinary relationships and creating a collaborative
infrastructure. Thus, manuals, trainings, and technical assistance for SARTs should go beyond
naming different ways of structuring a SART and recommending that SARTs should engage in
those practices; rather, they should identify the necessary resources, knowledge, and skills that
will be required to implement various structures, identify likely barriers to implementation, and
help SART members identify specific strategies for successful implementation.
For example, a workbook on policy/protocol development could begin by describing the
necessary elements to successfully drafting and agreeing upon a policy or protocol (e.g., buy-in
from organizational leaders; knowledge of best practices in responding to sexual assault;
knowledge of different stakeholder groups’ roles and limitations in that community; SART
members’ time to draft and review drafts of the policy). Then, the workbook could discuss
potential barriers to implementation of the protocol (e.g., conflict between organization’s policies
and the drafted SART protocol; no one knows what to do when there is disagreement, people
who are supposed to be following the protocol do not know it exists, etc.). Then the workbook
could, provide specific strategies SART can chose from to help their team successfully engage in
policy/protocol development and adoption. For example, the SARTs could form a small
multidisciplinary working group to engage in a multistage development process (identification of
84
 

 

the goals and desired use of the protocol, identification of a process for how to handle buy-in and
disagreement, discussions with organizations about their needs/concerns regarding a protocol,
reviewing best practices and sample protocols, drafting language, drafting plans for sharing the
protocol with the people who are supposed to follow it) and then have the working group bring
draft ideas and language to the broader group for discussion and revision. Thus one option to
promote SARTs’ effectiveness is to provide them with resources that can specifically help them
establish structure to organize their teamwork.
In addition, SARTs can support each other in establishing structure by sharing
information between SARTs. If one SART is struggling to successfully institutionalize case
review, it is likely that many other SARTs have had similar struggles, and have ideas for how to
overcome those problems. Thus, a venue that allows for communication and mentoring between
different SARTs could be a valuable contribution to the field. In particular, a venue that is easy
to access, and can be anonymous, may make SARTs more likely to reach out to one another for
support. For example, there could be a SART webpage with a members discussion section where
SART members can bring up different issues they have faced and share advice with one another.
An expert in collaboration could review the discussion and identify additional resources and tips
the teams may find helpful.
Evaluation. A final implication of the current study is that SARTs are under-evaluated.
Only 27 out of 172 SARTs engaged in formal program evaluation, and it is unclear to what
extent the evaluations they conducted met their information needs and were used to guide
practice. SARTs could benefit from systematically evaluating their efforts, and then using their
evaluation efforts to guide how they work together to improve the response to sexual assault.
While the findings regarding the role of program evaluation were somewhat mixed (differences
85
 

 

were only found between “High Adopters with Evaluation” and “High Adopters except
Evaluation” on only one of four domains of perceived effectiveness), it seems likely that if
SARTs were able to conduct evaluations that met their information needs and then use those
evaluations to guide their efforts, they would benefit (Patton, 2008). However, they need support
to do so. SART members are sexual assault responders, and are unlikely to have significant
expertise in designing, conducting, and using evaluation to guide practice. Thus, efforts to assist
SARTs in engaging in useful evaluations may be particularly beneficial. For example, sample
evaluation instruments could be provided for SARTs to use, along with corresponding plans for
how to conduct data collection, analyze the data, and use the findings to guide practice. SARTs
could also receive an initial training and the ongoing technical assistance from an organization
that supports sexual assault responders (such as a state coalition) to help them design, implement,
and use their own evaluation. Such an approach has been used successfully with SANE programs
to develop their evaluation capacity (Campbell, Townsend, Bybee, Shaw, & Markowitz, 2013)
and could be easily adapted to the context of SARTs. Additionally, financial resources could be
allocated by funders specifically to SARTs’ evaluation efforts. This would provide SARTs with
incentives to engage in evaluation and offset the time and other resources that would go into
engaging in evaluation.
Conclusion
For many decades, researchers and practitioners have been concerned with the many
problems with the community response to sexual assault. SARTs appear to be a promising
intervention model to bring together diverse stakeholders to work together to create
improvements in the response to sexual assault victims and cases (Greeson & Campbell, 2013).
Results of the current study suggest that how a SART is structured—its use of formal structures
86
 

 

and institutionalization of collaborative activities to organize its teamwork—has implications for
its (perceived) effectiveness. However, future work needs to (1) examine additional factors (such
as SART climate and process) that may also promote SARTs’ effectiveness and (2) employ
different measurement approaches to understand SARTs’ effectiveness at improving legal
outcomes and victims’ help-seeking experiences.

87
 

APPENDICES

88
 

 

APPENDIX A

Study 1: SART Leader Interview
 

89
 

 

Participant ID ____________________

Interviewer ID ____________________

Date

Start Time(s) ____________________

____________________

Interview Administration: 1 = Continuous; 2 = Stopped; Continued Later

Thank you for agreeing to participate in our study. I recognize that people from
different communities and different disciplines use different terms. Throughout the
interview I will use “SART” to refer to a multidisciplinary team, and victim to refer to
people who have been sexually assaulted. For some questions, I will ask you to choose
an answer that best represents your team, like on a survey, while for other questions, I
will ask you to describe your SART to me. There are no right or wrong answers- we are
just looking for your honest opinions. Remember- everything we speak about today will
be kept strictly confidential.

Date Entered_____________________________

By_________________________

Date Double Entered______________________

By_________________________

Date Transcribed_________________________

By_________________________

Document Quality Checking Here:

90
 

 

SECTION ONE: BACKGROUND INFORMATION ON THE SART 
I would like to start by getting some background information on the SART.
 
1. How long has the SART been in operation? 
____________________
____________________ [years]

SPSS

2. Has the SART been in operation continuously or have there been times when it
has stopped and re-started?
1 = CONTINUOUS
2 = STOPPED AND STARTED
888 = Not Applicable [due to SART being continuous]
a. If stopped and started, how long has the SART been in operation this time?
____________________
____________________ [years]

3. How long have you been a member of the SART?
____________________
____________________ [years]

SPSS

BACKGROUND 
91

 

4. What is your current title or position within the SART?
____________________________________________________________

SPSS

5. How long have you held this position?
____________________
____________________ [years]

SPSS

6. What stakeholder group do you belong to?
1 = Medical/forensic examiner
2 = Rape crisis center advocate or counselor
3= Prosecutor
4 = Police
5 = Crime lab personnel
6 = Advocate within victim’s witness unit in prosecutor’s office
7 = Judge
8 = Other [Specify__________________________________________]SPSS

Now, I have some quick yes or no questions about your SART.
7. Is there a specific group of sexual assault victims that your SART exclusively
serves (for example, a university population or a military base)?
1 = YES [Specify________________________________________]
0 = NO
BACKGROUND 
92

SPSS

 

8. What age groups of sexual assault victims does your SART serve?
____________________ [get cut-off years]

9. Is your SART a for-profit organization?

1=YES

0=NO

[If yes, go to Q11.]
[If no, go to Q10.]

10. Is your SART a 501c3 non-profit organization?

1 = YES

0 = NO

11. Does your SART currently have any formal sources of funding other than
fundraising (such as federal, state, or local grants)?

1 = YES 0 = NO

12. Does your SART have a leader, coordinator, or administrator?
1 = YES 0 = NO

13. Does your SART have a paid staff person?
a. If yes, is this a full-time position?
1 = YES
0 = NO
888 = Not Applicable [due to no paid staff]

BACKGROUND 
93

1 = YES 0 = NO

 

14. Does your community have a Child Advocacy Center? By community I mean the
community that your SART primarily serves.

1 = YES 0 = NO

15. Does your community have Sexual Assault Nurse Examiner [SANE] or Sexual
Assault Forensic Examiner [SAFE] trained nurses?

1 = YES 0 = NO
If no, go to Q18.

16. Does your community have a SANE or SAFE program? [defined as having
more than two nurses and a coordinator, and being recognized as a program by
their institution and the broader community]
1 = YES
0 = NO If no, go to Q18

888 = N/A due to no SANE nurses

17. Which came first, the SANE program or the SART?
1 = SANE FIRST
2 = SART FIRST
3 = SAME TIME

888 = N/A due to no SANE nurses or no SANE program

BACKGROUND 
94

 

SECTION TWO: MEMBERSHIP 
18. Now I would like to ask you questions about the members of your SART team.
By member I mean someone who attends SART meetings or other functions
reasonably frequently. I will read off a list of different stakeholder groups- just let
me know with a simple yes or no whether someone from that group is a member
of your SART.

a. Sexual Assault Nurse Examiner [SANE] or Sexual Assault Forensic Examiner
1 = YES

0 = NO

b. Other medical personnel

1 = YES

0 = NO

c. Rape crisis center staff

1 = YES

0 = NO

d. Prosecutor

1 = YES

0 = NO

e. Police

1 = YES

0 = NO

f. Crime lab personnel

1 = YES

0 = NO

[SAFE]

g. Advocate within the victim’s witness unit in the prosecutor’s office
1 = YES

0 = NO

h. Judicial

1 = YES

0 = NO

i.

1 = YES

0 = NO

j. Corrections (includes probation and parole)

1 = YES

0 = NO

k. Clergy or the faith community

1 = YES

0 = NO

l.

1 = YES

0 = NO

Sex offender treatment

School (grades k-12)
MEMBERSHIP 
95

 

m. Higher education (meaning, a college or university)
1 = YES

0 = NO

n. Child Advocacy Center

1 = YES

0 = NO

o. Domestic violence agency

1 = YES

0 = NO

p. Other social services (for example, drug abuse, welfare)
1 = YES

0 = NO

q. Sexual assault victims/survivors who represent the perspective of victims and
not an organization on the SART

1 = YES

0 = NO

r. Are there other groups that are represented in your SART that I didn’t
1 = YES

mention?

0 = NO

i. If yes, Specify[______________________________________]

SPSS

19. How many different organizations or agencies belong to your SART?

___________________________________

CONTINUE ON NEXT PAGE

MEMBERSHIP 
96

SPSS

 

20. Next I am going to ask you about the extent to which your SART is composed of
members who are front line staff, middle management, and executives within
their organization. Of the individuals involved in the SART approximately what
proportion are:

a. General staff or front-line workers that do not hold primary leadership or
decision-making roles in their organization? Would you say…
1 = None
2 = Some, but less than half
3 = Most members
4 = The vast majority or all members

b. Middle Management (such as captains or sergeants, department heads, or
program coordinators?)
1 = None
2 = Some, but less than half
3 = Most members
4 = The vast majority or all members

MEMBERSHIP 
97

 

c. The leader or executive of their organizations/agencies (for example, the chief
of police, elected prosecutor, rape crisis center director)?
1 = None
2 = Some, but less than half
3 = Most members
4 = The vast majority or all members

CONTINUE ON NEXT PAGE

MEMBERSHIP 
98

 

SECTION THREE: GOALS 
21. Next, I would like to ask you some questions about your SART’s goals. What was
the main reason for forming the SART?
___________________________________________________________
___________________________________________________________
___________________________________________________________
________________________________________________________ TRANSCRIBED

22. Next, I will read aloud a list of possible goals. I am going to ask you to rate how
important each of these goals is to your team currently. Please choose from: “not
a primary goal,” “somewhat important,” “very important,” or “a primary goal,” for
your team.
Goal

Would you say this is…
1 = Not a primary goal

a. To increase prosecution rates?

2 = Somewhat important
3 = Very important
4 = A primary goal
b. To increase reporting of sexual assaults to the police?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

GOALS 
99

 

Goal

Would you say this is…

c. To reduce barriers to seeking medical/forensic
services post-assault?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

d. To reduce barriers to seeking mental health services
post-assault?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

e. To reduce barriers to seeking advocacy services
post-assault?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

f. To improve the quality and comprehensiveness of

1 = Not a primary goal

medical/forensic services that are offered to

2 = Somewhat important

survivors who seek help from the medical system?

3 = Very important
4 = A primary goal

g. To improve the quality and comprehensiveness of

1 = Not a primary goal

mental health services that are offered to survivors

2 = Somewhat important

who seek help from the mental health system?

3 = Very important
4 = A primary goal

h. To improve the quality and comprehensiveness of

1 = Not a primary goal

advocacy services that are offered to survivors who

2 = Somewhat important

seek help from advocates?

3 = Very important
4 = A primary goal
GOALS 
100

 

Goal

Would you say this is…

i. To reduce negative treatment of victims by primary

1 = Not a primary goal

responders? By primary responders, I mean victim
advocates, medical/forensic examiners, police, and

3 = Very important

prosecutors.

4 = A primary goal

To educate primary responders about responding to

1 = Not a primary goal

sexual assault?

j.

2 = Somewhat important

2 = Somewhat important
3 = Very important
4 = A primary goal

k. To improve primary responders’ attitudes toward
sexual assault victims?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

l. To improve primary responders’ skills in responding to
sexual assault?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

m. To hold responders accountable to responding to
sexual assault appropriately?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

n. To educate the general public about sexual

1 = Not a primary goal

assault/rape and the services that are available to

2 = Somewhat important

victims in your community?

3 = Very important
4 = A primary goal
GOALS 
101

 

Goal

Would you say this is…

o. To increase coordination among stakeholders? By

1 = Not a primary goal

stakeholders, I mean community partners who work

2 = Somewhat important

on sexual assault issues.

3 = Very important
4 = A primary goal

p. To increase communication among stakeholders?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

q. To increase stakeholders’ understanding of one
another’s roles and limitations?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

r. To improve the quality of relationships among
stakeholders?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

s. To improve local organizations’ (such as the rape

1 = Not a primary goal

crisis center, hospital, police department, prosecutor’s

2 = Somewhat important

office) policies and procedures regarding sexual

3 = Very important

assault?

4 = A primary goal

t. To provide a victim-centered response to sexual
assault?

1 = Not a primary goal
2 = Somewhat important
3 = Very important
4 = A primary goal

GOALS 
102

 
 

23. What are other goals of your SART that I didn’t ask you about?
___________________________________________________________
___________________________________________________________ SPSS
 

24. If you were to rank the primary goals of your SART, is your SART’s time and
energy most focused on improving legal outcomes (like increasing prosecution
rates), improving victims’ experiences, or prevention education?
#1___________________________________________________________

SPSS

#2___________________________________________________________

SPSS

#3___________________________________________________________

SPSS

They can rank some or all of them equally. If equally, put them on the same line.
Go to 25.

DURING DATA ENTRY CODE BASED ON ANSWER TO 24
Legal outcomes vs. victim experiences
1 = Legal is more important; 2 = Victim is more important; 3 = Equally important

Legal outcomes vs. prevention/education
1 = Legal is more important; 2 = Prevention is more important; 3 = Equally important

Victim outcomes vs. prevention/education
1 = Victim is more important; 2 = Prevention is more important; 3 = Equally important

GOALS 
103

 

SECTION FOUR: STRUCTURE 
Now, I am going to ask you some questions about how you work together as a
team.
25. I will read a list of formal structures that some collaborations use. Please answer
yes or no which of these are in place in your SART.
a. Written mission statement

1 = YES

0 = NO

b. Meeting attendance and/or sign-in sheets

1 = YES

0 = NO

c. Written meeting agendas

1 = YES

0 = NO

d. Recorded and distributed meeting minutes

1 = YES

0 = NO

e. Newsletters for SART members

1 = YES

0 = NO

f. Subcommittees

1 = YES

0 = NO

g. SART organizational chart

1 = YES

0 = NO

h. Formal bylaws or operating rules/procedures specific to the operation of the
SART. This does NOT include policies/procedures for responding to sexual
assault victims.
1 = YES

0 = NO

i. Formal procedures for decision-making

1 = YES

0 = NO

j. Formal procedures for conflict resolution

1 = YES

0 = NO

k. An established mechanism for ensuring members are accountable to the rest
1 = YES

of the SART in completing tasks

 
STRUCTURE  
104

0 = NO

 

SECTION FIVE: ACTIVITIES 
26. Now, I will read a list of activities that some SARTs engage in. Please give me a
simple yes or no to indicate whether or not your SART uses this activity.

a. Case review (specifically reviewing individual cases)
1 = YES 0 = NO
i.

If yes, do these occur regularly, or as needed?
1 = As needed
2 = Regularly [Specify how often]_________________________SPSS
888 = Not applicable [does not utilize this activity]

b. Meetings which include business other than or in addition to case review
1 = YES 0 = NO
i.

If yes, do these occur regularly, or as needed?
1 = As needed
2 = Regularly [Specify how often]_________________________SPSS
888 = Not applicable [does not utilize this activity]

ACTIVITIES 
105

 

c. Cross-disciplinary trainings (meaning, one stakeholder group trains another;
does not include within discipline trainings)
1 = YES 0 = NO
i.

If yes, do these occur regularly, or as needed?
1 = As needed
2 = Regularly [Specify how often]_________________________SPSS
3 = Some regularly, some as needed
[For each, specify how often]_________________________SPSS
______________________________________________

SPSS

888 = Not applicable [does not utilize this activity]

d. Trainings conducted by non-SART members or presentations from guest
speakers that all SART members are invited to attend
1 = YES 0 = NO
i.

If yes, do these occur regularly, or as needed?
1 = As needed
2 = Regularly [Specify how often]_________________________SPSS
3 = Some regularly, some as needed
[For each, specify how often]_________________________SPSS
______________________________________________
888 = Not applicable [does not utilize this activity]
ACTIVITIES 
106

SPSS

 

e. Development of resources (such as checklists for responders, or
informational pamphlets for victims)
1 = YES 0 = NO
i.

If yes, does this occur regularly, or as needed?
1 = As needed
2 = Regularly [Specify how often]_________________________SPSS
3 = Some regularly, some as needed
[For each, specify how often]_________________________SPSS
______________________________________________

SPSS

888 = Not applicable [does not utilize this activity]

f. Discussion of the response to sexual assault in your community (including
problems, potential improvements, or other issues)
1 = YES 0 = NO
i.

If yes, does this occur regularly, or as needed?

1 = As needed
2 = Regularly [Specify how often]_________________________SPSS
3 = Some regularly, some as needed
[For each, specify how often]_________________________SPSS
______________________________________________
888 = Not applicable [does not utilize this activity]
ACTIVITIES 
107

SPSS

 

g. Policy or protocol adoption, development, or revision
1 = YES 0 = NO
i.

If yes, does this occur regularly, or as needed?
1 = As needed
2 = Regularly [Specify how often]_________________________SPSS
3 = Some regularly, some as needed
[For each, specify how often]_________________________SPSS
______________________________________________
888

SPSS

Not applicable [does not utilize this activity]

h. Development, adoption, or revision of memoranda of understanding between
different stakeholder groups
1 = YES 0 = NO
i.

If yes, does this occur regularly, or as needed?
1 = As needed
2 = Regularly [Specify how often]_______________________SPSS
3 = Some regularly, some as needed
[For each, specify how often]_________________________SPSS
______________________________________________
ACTIVITIES 
108

SPSS

 

888

i.

Not applicable [does not utilize this activity]

Program evaluation (such as research or quality assurance) of the SART
team as a whole
1 = YES 0 = NO
i.

If yes, does this occur regularly, or as needed?
1 = As needed
2 = Regularly [Specify how often]_______________________SPSS
3 = Some regularly, some as needed
[For each, specify how often]_________________________SPSS
______________________________________________
888

ii.

SPSS

Not applicable [does not utilize this activity]

[If yes]: Please describe your evaluation activities
[Probe for type of evaluation and topic of evaluation (for example,
prosecution rates vs. victim outcomes)]
________________________________________
________________________________________
_______________________________________

CONTINUE ON NEXT PAGE

ACTIVITIES 
109

SPSS/TRANSCRIBED

 

SECTION SIX: PERCEIVED EFFECTIVENESS 
We are a little over half of the way through the interview now. Thanks very much
for answering these questions. Next, I am going to ask you some more specific
questions about how you believe your SART has impacted your community. By
community I mean the community that your SART primarily serves.

27. To what extent has the SART’s efforts led to:
a. An increase in the
number of assaults
that are reported to
police? Please
choose from the
following
options…

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

b. Victims giving more
complete accounts
of the incident to law
enforcement?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
110

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
c. Victims being more
engaged during the
investigation of their
case?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

d. Victims being more
likely to continue
participating
throughout the
entirety of their
case?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

e. Victims being more
forthcoming with
other evidence that
may support their
accounts?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
111

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
f. Victims being more
willing to prosecute?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

g. Victims being more

engaged with
prosecutors during
court preparations?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

h. Victims being more
at ease with legal
personnel
throughout the
process?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
112

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
i.

Improvements in
support for victims
as they participate in
the criminal justice
process?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

j.

Improvements in the
quality of forensic
evidence?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

k. Improvements in
police knowledge of
medical/forensic
evidence?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT
PAGE

EFFECTIVENESS 
113

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
l.

Improvements in
prosecutors’
knowledge of
medical/forensic
evidence?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

m. Improvements in
police utilization of
medical/forensic
evidence?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

n. Improvements in
prosecutors’
utilization of
medical/forensic
evidence?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
114

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
o. Rape kits being
more likely to be
submitted to the
crime lab?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

p. Improvements in law
enforcements’
investigations of
sexual assault
cases?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

q. Improvements in
police building
rapport with victims?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
115

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
r. Improvements in
medical/forensic
personnel expert
witness testimony?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

s. Improvements in
prosecutors arguing
cases?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

t. Improvements in
prosecutors’
education of juries
about sexual
assault?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
116

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
u. Improvements in
prosecutors building
rapport with victims?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

v. Improvements in
prosecutors
preparing victims for
testimony?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

w. Police being more
likely to refer cases
to the prosecutor’s
office?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT
PAGE

EFFECTIVENESS 
117

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
x. Cases that are
referred to the
prosecutor’s office
being more likely to
be charged [same
as warranted or
authorized] by the
prosecutor?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

y. Cases that were
charged being less
likely to be
dismissed by a
grand jury, judge, or
prosecutor?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

z. An increase in the
number of
prosecuted cases
that result in
conviction or a guilty
plea?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
118

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
aa. A decrease in the
prevalence of sexual
assault in your
community?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

bb. The creation,
expansion, or restructuring of
programs or
services for sexual
assault victims in
your community?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

cc. Victims being more
likely to receive
referrals to
medical/forensic
services?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
119

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
dd. Victims being more
likely to receive
referrals to mental
health services?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

ee. Victims being more
likely to receive
referrals to
advocacy services?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

ff. Victims being more
likely to seek out
needed services
post-assault?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
120

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
gg. Responders being
more
knowledgeable
about sexual assault
and responding to
sexual assault
effectively? By
responders, I mean
victim advocates,
medical/forensic
examiners, police,
and prosecutors.

1= Not at all

hh. Improvements in
responders’ skills in
responding to
sexual assault?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

ii. Improvements in
responders’
attitudes toward
sexual assault
victims?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
121

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:

jj.

Responders
providing more
comprehensive,
higher quality
services to victims?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

kk. Responders being
more sensitive
toward victims?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

ll. Victims having more
positive experiences
with the legal
system?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

EFFECTIVENESS 
122

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
mm. Victims having
more positive
experiences with the
medical system?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

nn. Victims having more
positive experiences
with the mental
health system?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

oo. Victims having
more positive
experiences with
advocacy?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
123

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
pp. Improvements in the
general public’s
knowledge of sexual
assault and services
for sexual assault
victims in your
community?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

qq. Improvements in
local organizations’
policies and
procedures for
responding to
sexual assault?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

rr. An increase in
communication
among SART
stakeholders?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
124

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

To what extent has the SARTs’ efforts led to:
ss. SART stakeholders
having a better
understanding of
one another’s roles
and limitations?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great
extent

tt. SART responders
being more
accountable to
providing an
appropriate
response to sexual
assault?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great
extent

uu. Improvements in
SART stakeholders’
knowledge of
services available
to victims in your
community?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great
extent

1 = there was little to no room
for improvement?
2= this was not a goal of your
SART?
3= or your team was unable to
achieve this goal?
4=other[Specify______________
_________________]
1 = there was little to no room
for improvement?
2= this was not a goal of your
SART?
3= or your team was unable to
achieve this goal?
4=other[Specify______________
_________________]
1 = there was little to no room
for improvement?
2= this was not a goal of your
SART?
3= or your team was unable to
achieve this goal?
4=other[Specify______________
_________________]

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
125

 

To what extent has the SARTs’ efforts led to:
vv. SART stakeholders
working together
more efficiently?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

ww. Improvements in
the quality of
relationships among
SART stakeholders?

1= Not at all

If not at all, is that
because... ---->
2 = A little bit
3 = Somewhat
4 = Quite a bit
5 = To a great extent

CONTINUES ON NEXT PAGE

EFFECTIVENESS 
126

1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]
1 = there was little to no
room for improvement?
2= this was not a goal of
your SART?
3= or your team was unable
to achieve this goal?
4=other[Specify___________
____________________]

 

28. What factors have helped your SART achieve your goals?
Probe for factors related to:


Individuals



Organizations



Relationships (among individuals and among organizations)



Broader community



SARTs operations and activities
_______________________________________________
_______________________________________________
_______________________________________________
_______________________________________________
_______________________________________________
_______________________________________________ TRANSCRIBED

29. What factors have made it harder to achieve your goals?
Probe for factors related to:


Individuals



Organizations



Relationships (among individuals and among organizations)



Broader community



SARTs operations and activities
_______________________________________________
_______________________________________________
EFFECTIVENESS 
127

 

_______________________________________________
_______________________________________________
_______________________________________________
_______________________________________________ TRANSCRIBED
 

30. What advice or lessons learned do you have to share based on your experiences
with your SART? [OK if they don’t want to add anything. This is just a chance for
them to share more.]
________________________________________________________________________
_______________________________________________________ TRANSCRIBED 

EFFECTIVENESS 
128

 

SECTION SEVEN: OTHER CHARACTERISTICS OF COMMUNITY RESPONSE 
Next I want to ask you real quick, just yes or no, whether a few things exist in
your community.
31. Does your community have any multidisciplinary joint service agencies that are
specific to sexual assault? (for example, an organization that houses both police
and medical/forensic examiners)
1 = YES
0 = NO

a. If yes, Can you describe this for me?
[Probe for which stakeholder groups work together and how they work
together.]
_______________________________________________ TRANSCRIBED

32. In your community, do medical/forensic examiners and police typically conduct
interviews jointly with victims?
1 = YES
0 = NO

33. Does the local prosecutor’s office have a specialized unit that handles sexual
assault crimes?
1 = YES
0 = NO
a. If yes, what type of unit is it? [for example, a sex crimes unit]
_________________________________________________________
COMMUNITY RESPONSE 
129

SPSS

 

34. Are there local task forces, councils, or coalitions (other than the team we have
discussed during this interview) that some of your members are a part of?
1 = YES [Specify_________________________________]
0 = NO

SPSS

a. If yes, approximately how many of your SARTs’ members are also involved
in one of these teams? Would you say…
1 = None
2 = Some, but less than half
3 = Most members
4 = The vast majority or all members

888 = Not Applicable; no to 34.

35. How often are victims who report the assault to police given information about
services from the local rape crisis center? Would you say that…
1 = The option is offered to the survivor every time, for every survivor
when appropriate/applicable
2= The option is offered to the survivor most of the time
3= The option is offered to the survivor occasionally
4= The option is offered to the survivor rarely or never

888 = Not applicable, due to no rape crisis center

COMMUNITY RESPONSE 
130

 

36. How often are victims who report the assault to police given information about
receiving a medical forensic exam?
1 = The option is offered to the survivor every time, for every survivor
when appropriate/applicable
2= The option is offered to the survivor most of the time
3= The option is offered to the survivor occasionally
4= The option is offered to the survivor rarely or never

37. How often are victims given the opportunity to have a victim advocate
accompany them during the medical/forensic exam?
1 = The option is offered to the survivor every time, for every survivor
when appropriate/applicable
2= The option is offered to the survivor most of the time
3= The option is offered to the survivor occasionally
4= The option is offered to the survivor rarely or never

38. How often are victims given the opportunity to have a victim advocate
accompany them during the initial report to the responding police officer?
1 = The option is offered to the survivor every time, for every survivor
when appropriate/applicable
2= The option is offered to the survivor most of the time
3= The option is offered to the survivor occasionally
4= The option is offered to the survivor rarely or never

COMMUNITY RESPONSE 
131

 

39. How often are victims given the opportunity to have a victim advocate
accompany them during victim interviews with the detective?
1 = The option is offered to the survivor every time, for every survivor
when appropriate/applicable
2= The option is offered to the survivor most of the time
3= The option is offered to the survivor occasionally
4= The option is offered to the survivor rarely or never

40. How often are victims given the opportunity to have a victim advocate
accompany them during court hearings?
1 = The option is offered to the survivor every time, for every survivor
when appropriate/applicable
2= The option is offered to the survivor most of the time
3= The option is offered to the survivor occasionally
4= The option is offered to the survivor rarely or never

41. How often does law enforcement consult with medical/forensic examiners
regarding forensic exam findings? Would you say that…
1= Law enforcement consults with medical forensic examiners regarding
every case, when applicable
2= Law enforcement consults with medical forensic examiners regarding
most cases
3= Law enforcement consults with medical forensic examiners regarding
occasional cases
4= Law enforcement consults with medical forensic examiners regarding
cases rarely or never
COMMUNITY RESPONSE 
132

 

42. How often do prosecutors consult with medical/forensic examiners regarding
forensic evidence findings?
1= Prosecutors consult with medical/forensic examiners regarding every
case, when applicable
2= Prosecutors consult with medical/forensic examiners regarding most
cases
3= Prosecutors consult with medical/forensic examiners regarding
occasional cases
4= Prosecutors consult with medical/forensic examiners rarely or never

43. How often do medical forensic examiners conduct an examination of suspects in
sexual assault cases when a suspect has been apprehended?
1= A medical forensic examiner conducts a suspect exam for every
suspect that have been apprehended, when applicable
2= A medical forensic examiner conducts a suspect exam for most
suspects that have been apprehended
3= A medical forensic examiner conducts a suspect exam occasionally of
suspects that have been apprehended
4=A medical forensic examiner conducts a suspect exam rarely or never
for suspects that have been apprehended

CONTINUES ON NEXT PAGE

COMMUNITY RESPONSE 
133

 

44. When sexual assault cases go to trial, how often do medical/forensic examiners
testify? Would you say that…
1 = A medical/forensic examiner testifies every time a sexual assault case
goes to trial, when applicable
2= A medical/forensic examiner nurse testifies in most sexual assault
cases that go to trial
3= A medical/forensic examiner testifies in occasional sexual assault
cases that go to trial
4= A medical/forensic examiner testifies rarely or never in sexual assault
cases that go to trial

CONTINUES ON NEXT PAGE

COMMUNITY RESPONSE 
134

 

SECTION EIGHT: DEMOGRAPHICS 
Now, I will end the interview with some demographic questions about you and
your community.

45. What communities does your SART primarily serve? Probe for specific counties,
cities, and/or jurisdictions.
[Do not include people who happen to come in from other communities]
________________________________

community, state

SPSS

________________________________

community, state

SPSS

________________________________

community, state

SPSS

________________________________

community, state

SPSS

46. How many counties does your SART serve?
________________________________________

SPSS

47. Would you characterize the community/communities that your SART serves as…
a. Urban?

1 = YES

0 = NO

b. Suburban?

1 = YES

0 = NO

c. Rural?

1 = YES

0 = NO

48. What is your gender?
1 = Male
2 = Female
3 = Transgender
135
 

 

49. Age?
________________________________________ (in years)

SPSS

50. Race?
1 = Native American or Alaskan Native
2 = Asian American
3 = Black or African American
4 = Native Hawaiian or Pacific Islander
5 = White
6 = Hispanic, Latino(a), or of Spanish origin
7 = Biracial or Multiracial
[Specify____________________________________________]SPSS
8 = Other [Specify_____________________________________________]SPSS

51. Education Level?
1 = Did not complete high school
2 = High school graduate/GED
3 = Some college or trade school, no degree
4 = Associates degree
5 = Trade school degree
6 = Bachelor’s degree
7 = Advanced degree
Turn off tape recorder.
136
 

 

SECTION NINE: CONCLUSION 
Turn off tape recorder.
Thank you very much for participating in our study. Before we get off the phone, I
do want to let you know that some of the SARTs that we are interviewing will be
selected for more in-depth study in the future. If your team is selected again, we will be
in touch.

Would you like to be included in a list of people who will receive a brief report of findings
from this study?

Do you have any questions for me?

End Time(s): ___________________________

SPSS

Calculate length of interview: _____________________(minutes)

SPSS

137
 

 

CODE LATER
UCR and CENSUS DATA
1.

Numbers of sexual assaults in the past year (UCR)
_________________________________

Date Entered___________________ By___________________________

2.

Total population of communities served (CENSUS)
________________________________________

Date Entered___________________ By___________________________

 

138
 

 

APPENDIX B

Perceived Legal Effectiveness Items and their Corresponding Subscales
Table 7:
Perceived Legal Effectiveness Items and their Corresponding Subscales
To what extent has the SARTs’ efforts led to…

Subscale

Victims being more at ease with legal personnel

Improvements in victims’ participation in
the criminal justice system

 

Victims being more likely to continue
participating during the entire case

Improvements in victims’ participation in
the criminal justice system

Victims being more engaged in the investigation

Improvements in victims’ participation in
the criminal justice system

Victims being more forthcoming with other
evidence to help their cases 

Improvements in victims’ participation in
the criminal justice system

Victims being more engaged with prosecutors
during court preparations 

Improvements in victims’ participation in
the criminal justice system

Victims being more willing to prosecute

Improvements in victims’ participation in
the criminal justice system

 

Victims giving more complete accounts to law
enforcement 

Improvements in victims’ participation in
the criminal justice system

Increase in the number of assaults reported to the
police

Improvements in victims’ participation in
the criminal justice system

Improvements in support for victims in the
criminal justice process

Improvements in victims’ participation in
the criminal justice system

Improvements in law enforcement’s
investigations of sexual assault cases

Improvements related to police
processing of sexual assault cases

Improvements in police knowledge of medical
forensic evidence

Improvements related to police
processing of sexual assault cases

Improvements in police building rapport with
victims

Improvements related to police
processing of sexual assault cases

139
 

 

Table 7 (cont’d)
Improvements in police utilization of medical
forensic evidence

Improvements related to police
processing of sexual assault cases

Police being more likely to refer cases to the
prosecutor’s office

Improvements related to police
processing of sexual assault cases

Rape kits being more likely to be submitted to the Improvements related to police
crime lab
processing of sexual assault cases
Improvements in prosecutors’ knowledge of
medical/forensic evidence

Improvements related to the prosecution
of sexual assault cases

Improvements in prosecutors’ utilization of
medical forensic evidence

Improvements related to the prosecution
of sexual assault cases

Improvements in prosecutors’ arguing cases

Improvements related to the prosecution
of sexual assault cases

Increase in number of prosecuted cases that result
in conviction

Improvements related to the prosecution
of sexual assault cases

Cases that are referred to the prosecutor’s office
being more likely to be charged

Improvements related to the prosecution
of sexual assault cases

Improvements in medical personnel expert
witness testimony

Improvements related to the prosecution
of sexual assault cases

Improvements in prosecutors preparing victims
for testimony

Improvements related to the prosecution
of sexual assault cases

Improvements in prosecutors’ building rapport
with victims

Improvements related to the prosecution
of sexual assault cases

Improvements in the quality of medical/forensic
evidence

Improvements related to the prosecution
of sexual assault cases

140
 

 

APPENDIX C

Comparison of Three and Four Cluster Solution Results
Based on hierarchical cluster analysis results, both a three or four cluster solution were
reasonable choices. Therefore k-means cluster analysis was conducted for both a three and four
cluster solution. The results of these analyses are presented below. ANOVA and Chi-squared
tests were conducted to test whether the clusters differed on the variables of interest, and posthoc comparisons (Tukey’s HSD) were conducted to which specific pairs of clusters were
statistically significantly different from one another. The three and four cluster solutions are
presented in the tables below, and then after, a discussion of the rationale for choosing the four
cluster solution is given.
 

141
 

 

Table 8:
Results of Three Cluster Solution (Chosen)
“Low
Adopters”

“High
Adopters
Except
Evaluation”

“High
Adopters
Plus
Evaluation”

Results of Post-hoc
comparisons for
significant differences
between clusters

Cluster size

n=65

n=80

n=27

n/a

Formalization (0-13)

Low

High

High

2 and 3 are not
statistically different

a3

b

b

M = 4.98

M = 6.90

M = 7.44

SD = 1.74

SD = 1.98

SD = 2.83

Institutionalization
of case review (01
2)

Average

Average

Average

a

a

a

No statistically significant
differences

b

2 and 3 are not
statistically different

M = 1.00

M = 1.28

M = 1.19

SD = .79

SD = .81

SD = .83

Institutionalization
of multidisciplinary
cross-trainings (01
2)

Low

High

High

a

b

M = .48

M = 1.54

M = 1.58

SD = .61

SD = .53

SD = .64

Institutionalization
of policy/protocol
development and
1
review (0-2)

Low

High

Medium

Program evaluation
2
(0-1)

a

b

2 and 3 are higher than 1
2 and 3 are higher than 1
c

M = .64

M = 1.53

SD = .54

SD = .53

SD = .67

Low

Low

2 is higher than 3

M = 1.15

High

a

a

M=0

M=0

SD = 0
1

2 and 3 are higher than 1

M=1

SD = 0

1 and 2 are not
statistically different.

b

3 is higher than 1 and 2

SD = 0
2

0 = activity not utilized; 1 = used as needed; 2 = used regularly; 0 = does not use program
3

evaluation; 1 = uses program evaluation; Within a row, means with the same letter superscript
indicate no statistically significant differences between groups based on Tukey’s HSD post-hoc
tests. Within a row, means with different superscripts indicate statistically significant differences
4
between groups based on Tukey’s HSD post-hoc tests; Cluster analysis selects for low variation
within clusters and high variation across clusters. Therefore, results of ANOVAs/Chi-Squared
should be interpreted in the context of describing the results of the cluster analysis only.
 

142
 

 

Table 9:
Results of Four Cluster Solution (Not Chosen)
Low
formalization,
cross-training,
policy, and
evaluation

Low case
review and
evaluation;
High crosstraining

Medium
crosstraining;
High case
review

Medium
case
review;
High
program
evaluation

Results of Post-hoc comparisons for
significant differences between
clusters

Cluster size

n=52

n=39

n=55

n=26

n/a

Formalization

Low

High

High

High

1 is lower than 2, 3, and 4

(0-13)

M=4.66

M=6.25

M=7.30

M=7.27

No significant differences between 2, 3,
4

Institutionalization of case
review

Average

1

(0-2)

Institutionalization of
multidisciplinary crosstrainings

a

M=.85

b

a

b

b

b

c

Low

High

Average

M = .49

M=1.93

a

M=1.15

No significant differences between 1 and
4
3 is higher than 1, 2, and 4
1 and 4 are higher than 2

a

b

c

Low

High

Average

M = .35

M = 1.74

M = 1.26

1

(0-2)

143

MediumHigh
M=1.58

bc

1 is lower than 2, 3, and 4
2 is higher than 3, but 2 and 4 are not
significantly different
3 and 4 are not significantly different

 

Table 9 (cont’d)
Institutionalization of
policy/protocol development
1
and review (0-2)
2

a

b

b

b

Low

High

High

High

M = .63

M = 1.49

M =1.35

M=1.15

a

a

a

b

1
3

Low

Low

Low

High

M=0

Program evaluation (0-1)

M=0

M=0

M=1

1 is lower than 2, 3, and 4
No differences between 2, 3, 4

4 is higher than 1, 2, and 3
No differences between 1,2,3

2

0 = activity not utilized; 1 = used as needed; 2 = used regularly; 0 = does not use program evaluation; 1 = uses program evaluation;

Within a row, means with the same letter superscript indicate no statistically significant differences between groups based on
Tukey’s HSD post-hoc tests. Within a row, means with different superscripts indicate statistically significant differences between
4
groups based on Tukey’s HSD post-hoc tests; Cluster analysis selects for low variation within clusters and high variation across
clusters. Therefore, results of ANOVAs/Chi-Squared should be interpreted in the context of describing the results of the cluster
analysis only.
Both solutions had reasonable cluster sizes and were differentiated by the clustering variables. The four cluster solutions has
somewhat more balanced cluster sizes and all 5 clustering variables had differences across clusters, while the three cluster solution had
somewhat less balanced cluster sizes, and one of the 5 cluster variables (case review) did not differ across clusters. However, the
results of the four cluster solution were less conceptually meaningful. The clusters were difficult to interpret and name, due to the
many, varied patterns of ways in which clusters differed from one another on the variables of interest. It was also expected that these
would be much more difficult for SART stakeholders to make sense of. Therefore, the three cluster solution was chosen for further
analyses.
144

 

REFERENCES

145

 

REFERENCES
Ahrens, C. E., Cabral, G., & Abeling, S. (2009). Healing or hurtful: Sexual assault survivors’
interpretations of social reactions from support providers. Psychology of Women
Quarterly, 33, 81-94.
Allen, N. E. (2005). A multi-level analysis of community coordinating councils. American
Journal of Community Psychology, 35, 49-63.
Allen, N. E. (2006). An examination of the effectiveness of domestic violence coordinating
councils. Violence Against Women, 12, 46-67.
Allen, N. E., Javdani, S., Anderson, C. J., Rana, S., Newman, D., Todd, N., … Davis, S. (2010).
Coordinating the criminal justice response to intimate partner violence: The effectiveness
of councils in producing systems change. Washington, DC: National Institute of Justice.
Allen, N. E., Watt, K. E., & Hess, J. Z. (2008). A qualitative study of the activities and outcomes
of domestic violence coordinating councils. American Journal of Community Psychology,
41, 63-73.
American College of Emergency Physicians (1999). Evaluation and management of the sexually
assaulted or sexually abused patient. Dallas, TX: American College of Emergency
Physicians.
Amey, A. L., & Bishai, D. (2002). Measuring the quality of medical care for women who
experience sexual assault with data from the National Hospital Ambulatory Medical Care
Survey. Annals of Emergency Medicine, 39, 631-638.
Amo, C. & Cousins, J.B. (2007). Going through the process: An examination of the
operationalization of process use in empirical research on evaluation. New Directions in
Evaluation, 116, 5‐26.
Amstadter, A. B., McCauley, J. L., Ruggiero, K. J., Resnick, H. S., & Kilpatrick, D. G. (2008).
Service utilization and help seeking in a national sample of female rape victims.
Psychiatric Services, 59(12), 1450-1457.
Amstadter, A. B., McCauley, J. L., Ruggiero, K. J., Resnick, H. S., & Kilpatrick, D. G. (2010).
Self-rated health in relation to rape and mental health disorders in a national sample of
women. American Orthopsychiatric Association, 81, 202-210.
Basile. K. C., Chen, J., Black, M. C., & Saltzman, L. E. (2007). Prevalence and characteristics of
sexual violence victimization among U.S. Adults, 2001-2003. Violence & Victims, 22,
437-438.
Basile, K. C., & Smith, S. G. (2011). Sexual violence victimization of women: Prevalence,
characteristics, and the role of public health and prevention. American Journal of
Lifestyle Medicine, 5, 407-417.
146

 

Berkowitz, B. (2001). Studying the outcomes of community-based coalitions. American Journal
of Community Psychology, 29, 213-227.
Birkeland, S., Murphy-Graham, E., & Weiss, C. (2005). Good reasons for ignoring good
evaluation: The case of the drug abuse resistance education (D.A.R.E.) program.
Evaluation and Program Planning, 28, 247-256
Booth, B. M., Mengeling, M., Torner, J., & Sadler, A. G. (2011). Rape, sex partnership, and
substance use in women veterans. Journal of Traumatic Stress, 24, 287-294.
Brecklin, L. R., & Ullman, S. E. (2010). The roles of victim and offender substance use in sexual
assault outcomes. Journal of Interpersonal Violence, 25, 1503-1522.
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and
design. Cambridge, MA: Harvard University Press.
Bronstein, L. R. (2003). A model for interdisciplinary collaboration. Social Work, 48, 297-306.
Bryant-Davis, T., Chung, H., & Tillman, S. (2009). From the margins to the center: Ethnic
minority women and the mental health effects of sexual assault. Trauma, Violence, &
Abuse, 10, 330-357.
Burgess, A. W., Lewis-O’Connor, A., Nugent-Borakove, M. E., Fanflik, P. (2006). SANE/SART
services for sexual assault victims: Policy implications. Victims and Offenders, 1, 205212.
Butterfoss, F. D., Goodman, R. M., & Wandersman, A. (1993). Community coalitions for
prevention and health promotion. Health Education Research, 8, 315-330.
Campbell, R. (1998). The community response to rape: Experiences with the legal, medical, and
mental health systems. American Journal of Community Psychology, 26, 355-379.
Campbell, R. (2005). What really happened? A validation study of rape survivors' help-seeking
experiences with the legal and medical systems. Violence and Victims, 20, 55-68.
Campbell, R. (2008). The psychological impact of rape victims’ experiences with the legal,
medical, and mental health systems. American Psychologist, 68, 702-717.
Campbell, R., Adams, A. E., Wasco, S.M., Ahrens, C. E., & Sefl, T. (2009). Training
interviewers for research on sexual violence: A qualitative study of rape survivors’
recommendations for interview practice. Violence Against Women, 15, 595–617.
Campbell, R., & Ahrens, C. (1998). Innovative community services for rape victims: An
application of multiple case study methodology. American Journal of Community
Psychology, 26, 537-571.

147

 

Campbell, R., & Bybee, D. (1997). Emergency medical services for rape victims: Detecting the
cracks in service delivery. Women's Health: Research on Gender, Behavior, and Policy,
3(2), 75-101.
Campbell, R., Bybee, D., Ford, J. K., Patterson, D., & Ferrell, J. (2009). A systems change
analysis of SANE programs: Identifying the mediating mechanisms of criminal justice
system impact. Washington, DC: National Institute of Justice.
Campbell, R., Dworkin, E., & Cabral, G. (2009). An ecological model of the impact of sexual
assault on women’s mental health. Trauma, Violence & Abuse, 10, 225-246.
Campbell, R., Greeson, M. R., Bybee, D., & Kennedy, A. (2011). Adolescent sexual assault
victims’ experiences with SANE-SARTs and the criminal justice system. Washington, DC:
National Institute of Justice.
Campbell, R., & Martin, P. Y. (2001). Services for sexual assault survivors: The role of rape
crisis centers. In C. Renzetti, J. Edleson, & R. Bergen (Eds.), Sourcebook on violence
against women (pp. 227–241). Thousand Oaks, CA: Sage.
Campbell, R., & Patterson, D. (2011). Services for victims of sexual violence. In M. P. Koss, J.
W. White, & A. E. Kazdin (Eds.), Violence against women: Navigating solutions (volume
2) (pp. 95-114). Washington DC: American Psychological Association.
Campbell, R., Patterson, D., & Lichty, L. F. (2005). The effectiveness of sexual assault nurse
examiner (SANE) program: A review of psychological, medical, legal, and community
outcomes. Trauma, Violence, & Abuse: A Review Journal, 6, 313-329.
Campbell, R., & Raja, S. (1999). The secondary victimization of rape victims: Insights from
mental health professionals who treat survivors of violence. Violence and Victims 14,
261-275.
Campbell, R., & Raja, S. (2005). The sexual assault and secondary victimization of female
veterans: Help-seeking experiences with military and civilian social systems. Psychology
of Women Quarterly, 29, 97-106.
Campbell, R., Townsend, S. M., Bybee, D., Shaw, J. L., & Markowitz, J. (2013). Implementation
of a Sexual Assault Nurse Examiner (SANE) practitioner toolkit. Washington, DC: U.S.
Department of Justice.
Campbell, R., Townsend, S. M., Long, S. M., Kinnison, K. E., Pulley, E. M., Adames, S. B., &
Wasco, S. M. (2006). Responding to sexual assault victims’ medical and emotional
needs: A national study of the services provided by SANE programs. Research in
Nursing & Health, 29, 384-398.
Campbell, R., Wasco, S. M., Ahrens, C. E., Sefl, T., & Barnes, H. E. (2001). Preventing the
"second rape": Rape survivor's experiences with community service providers. Journal of
Interpersonal Violence, 16, 1239-1259.
148

 

Chen, Y., & Ullman, S. E. (2010). Women’s reporting of sexual and physical assaults to police
in the national violence against women survey. Violence Against Women, 16, 262-279.
Clark, S. J., Burt, M. R., Schulte, M. M., & Maguire, K. (1996). Coordinated community
response to domestic violence in six communities: Beyond the justice system. Retrieved
from: http://aspe.hhs.gov/hsp/cyp/domvilnz.htm.
Clatworthy, J., Buick, D., Hankins, M., Weinman, J., & Horne, R. (2005). The use and reporting
of cluster analysis in health psychology: A review. British Journal of Health Psychology,
10, 329-358.
Cloutier, S., Martin, S. L., & Poole, C. (2002). Sexual assault among North Carolina women:
prevalence and risk factors. Journal of Epidemiology & Community Health, 56, 265-271.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 115-159.
Cousins, J.B., & Whitmore, E. (1998). Framing participatory evaluation. New Directions for
Evaluation,80, 5‐23.
Dearing, J. W. (2009). Applying diffusion of innovation theory to intervention development.
Research on Social Work Practice, 19, 503-518.
Dearing, J. W., Maibach, E. W., & Buller, D. B. (2006). A convergent diffusion and social
marketing approach for disseminating proven approaches to physical activity promotion.
American Journal of Preventive Medicine, 31, S11-S23.
Department of Justice (2004). A national protocol for sexual assault medical forensic
examinations: Adults/adolescents. Washington, DC: Author.
Durlak, J. A., & Dupre, E. P. (2008). Implementation matters: A review of research on the
influence of implementation of program outcomes and the factors affecting
implementation. American Journal of Community Psychology, 41, 327-350.
El-Mouehly, M. (2004). Violence against women: A public health problem. The Journal of
Primary Prevention, 25, 289-303.
Emshoff, J. G. (2008). Researchers, practitioners, and funders: Using the framework to get us on
the same page. American Journal of Community Psychology, 41, 393-403.
Emshoff, J., Blakely, C., Gray, D., Jakes, S., Brounstein, P., Coulter, J., & Gardner, S. (2003).
An ESID case study at the federal level. American Journal of Community Psychology, 32,
345-357.
Filipas, H. H. & Ullman, S. E. (2001). Social reactions to sexual assault victims from various
support sources. Violence and Victims, 16, 673-692.
Florida Council Against Sexual Violence (2007). SART toolkit. Tallahasse, FL: Author.

149

 

Foster-Fishman, P. G., Berkowitz, S. L., Lounsbury, D. W., Jacobson, S., & Allen, N. A. (2001).
Building collaborative capacity in community coalitions: A review and integrative
framework. American Journal of Community Psychology, 29, 241-261.
Frazier, P., & Haney, B. (1996). Sexual assault cases in the legal system: Police, prosecutor and
victim perspectives. Law and Human Behavior, 20, 607-628.
George, L.,Winfield, I., & Blazer, D. (1992). Sociocultural factors in sexual assault: Comparison
of two representative samples of women. Journal of Social Issues, 48, 105-125.
Golding, J. M. (1999). Sexual assault history and long-term physical health problems: Evidence
from clinical and population epidemiology. Current Directions in Psychological Science,
8, 191-194.
Greeson, M. R., & Campbell, R. (2013; online first). Sexual Assault Response Teams (SARTs):
An empirical review of the effectiveness and challenges to successful implementation.
Trauma, Violence, & Abuse.
Hanson, R. F., Self-Brown, S., Borntrager, C., Kilpatrick, D. G., Saunders, B. E., Resnick, H. S.,
Amstader, A. (2008). Relations among gender, violence exposure, and mental health: the
National Survey of Adolescents. American Journal of Orthopsychiatry, 78(3), 313-321.
Hazel, K. L., & Onaga, E. (2003). Experimental social innovation and dissemination: The
promise and its delivery. American Journal of Community Psychology, 32, 285-294.
Hosmer, D. W., Jr., & Lemeshow, S. (2000). Applied logistic regression (2nd ed.). New York:
Wiley.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling: A
Multidisciplinary Journal, 6, 1-55.
Illinois Criminal Justice Information Authority (2003). The efficacy of Illinois’ sexual assault
nurse examiner (SANE) pilot program. Chicago, IL: Author.
Jordan, C. E., Campbell, R., & Follingstad, D. (2010). Violence and women’s mental health: The
impact of physical, sexual, and psychological aggression. Annual Review of Clinical
Psychology, 6, 607-628.
Kentucky Association of Sexual Assault Programs (2002). Developing a sexual assault response
team: A resource guide for Kentucky communities. Frankfort, KY: Author.
Kilpatrick, D. G., & Acierno, R. (2003). Mental health needs of crime victims: Epidemiology
and outcomes. Journal of Traumatic Stress, 16, 119-132.
Kilpatrick, D. G., Resnick, H. S., Ruggiero, K. J., Conoscenti, L. M., & McCauley, J. (2007).
Drug-facilitated, incapacitated, and forcible rape. Washington, DC: National Institute of
Justice.
150

 

Kimerling, R., & Calhoun, K.S. (1994). Somatic symptoms, social support, and treatment
seeking among sexual assault victims. Journal of Consulting and Clinical Psychology,
62, 333–340.
Koss, M. P., Bailey, J. A., Yuan, N. P., Herrera, V. M., & Lichter, E. L. (2003). Depression and
PTSD in survivors of male violence: Research and training initiatives to facilitate
recovery. Psychology of Women Quarterly, 27, 130-142.
Koss, M.P., Gidycz, C.A., & Wisniewski, N. (1987). The scope of rape: Incidence and
prevalence of sexual aggression and victimization in a national sample of higher
education students. Journal of Consulting and Clinical Psychology, 55, 162-170.
Kreuter, M. W., Lezin, N. A., & Young, L. A. (2000). Evaluating community-based
collaborative mechanisms: Implications for practitioners. Health Promotion Practice, 1,
49-63.
Krug, E. G., Dahlberg, L. L., Mercy, J. A., Zwi, A. B., & Lozano, R. (2002). World health report
on violence. Geneva: World Health Organization.
Ledray, L. E. (2001). Evidence collection and care of the sexual assault survivor: the SANESART response. St. Paul, MN: Violence Against Women Online Resources.
Lee, S. J., Altschul, I., & Mowbray, C. T. (2008). Using planned adaptation to implement
evidenced-based programs with new populations. American Journal of Community
Psychology, 41, 290-303.
Lewis-O’Connor, A. (2009). The evolution of SANE/SART -- are there differences? Journal of
Forensic Nursing, 5, 220-227.
Lisak, D. & Miller, P. M. (2002). Repeat rape and multiple offenders among undetected rapists.
Violence and Victims, 17, 73-84.
Littel, K. (2001). Sexual assault nurse examiner programs: Improving the community response to
sexual assault victims. Office for Victims of Crime Bulletin, 4, 1-19.
Logan, T. K., Cole, J., & Capillo, A. (2007). Sexual assault nurse examiner program
characteristics, barriers, and lessons learned. Journal of Forensic Nursing, 3, 24-34.
Logan, T. K., Evans, L., Stevenson, E., & Jordan, C. E. (2005). Barriers to services for rural and
urban survivors of rape. Journal of Interpersonal Violence, 20, 591-616.
Lonsway, K. A., & Archambault, J. (2012). The “justice gap” for sexual assault cases: Future
directions for research and reform. Violence Against Women, 18, 145-169.
Luke, D. A. (2005). Getting the big picture in community science: Methods that capture context.
American Journal of Community Psychology, 35, 185-200.

151

 

Macy, R. J., Giattina, M., Sangster, T. H., Crosby, C., & Montijo, N. J. (2009). Domestic
violence and sexual assault services: Inside the black box. Aggression and Violent
Behavior, 14, 359-373.
Madigan, L., & Gamble, N. (1991). The second rape: Society's continued betrayal of the victim.
New York: MacMillan.
Malefyt, M. B., Littel, K. M., & Walker, A. H. (1998). Promising practices: Improving the
criminal justice system’s response to violence against women. Washington DC: STOP
Violence Against Women Grants Technical Assistance Project.
Martin, P. Y. (2005). Rape work: Victims, gender, and emotions in organization and community
context. New York: Routledge.
Martin, P. Y., DiNitto, D., Byington, D., & Maxwell, M.S. (1992). Organizational and
community transformation: The case of a rape crisis center. In D. Bargal & H. Schmid
(Eds.). Organizational change and development in human service organizations. (pp.123145.). New York: Haworth Press.
Martin, P. Y. & Powell, R. M. (1994). Accounting for the second assault: Legal organizations
framing of rape victims. Law and Social Inquiry, 18, 853-890.
Martin, S. & Macy, R. (2009). Sexual violence against women: Impact on high-risk health
behaviors and reproductive health. Harrisburg, PA: VAWnet, a project of the National
Resource Center on Domestic Violence/Pennsylvania Coalition Against Domestic
Violence. Retrieved from: http://www.vawnet.org
Martin, S. L., Macy, R. J., & Young, S. K. (2011). Heath and economic consequences of sexual
violence. In Violence against women: Mapping the terrain (volume 1) (pp. 173-195).
Washington DC: American Psychological Association.
Mattesich, P. W., Murray-Close, M., & Monsey, B. R. (2001). Collaboration: what makes it
work (2nd ed.). St. Paul, MN: Amherst H. Wilder Foundation.
Mayer, J. P., & Davidson, W.S. (2000). Dissemination of innovation as social change. In J.
Rappaport & E. Seidman (Eds.), Handbook of community psychology (pp. 421-438). New
York: Plenum.
Miller, R. L., & Shinn, M. (2005). Learning from communities: overcoming difficulties in
prevention and promotion efforts. American Journal of Community Psychology, 35, 169183.
Monroe, L. M., Kinney, L. M., Weist, M. D., Dafeamekpor, D. S., Dantzler, J. & Reynolds,
M.W. (2005). The experience of sexual assault: Findings from a statewide victim needs
assessment. Journal of Interpersonal Violence, 20, 767-777.
National Sexual Violence Resource Center (2011). Sexual assault response team development: A
guide for victim service providers. Enola, PA: Author.
152

 

National Victim Center (1992). Rape in American: A report to the nation. Arlington, VA:
National Victim Center.
Noble, A., Brannon-Patel, E., & Tysoe, I. (2001). An evaluation of the sexual assault response
team victim advocate program. Sacramento, CA: Governor’s Office of Criminal Justice
Planning Program Evaluation Division.
Nowell, B. (2006). The role of social capital in interorganizational alliances. East Lansing, MI:
Michigan State University.
Nugent-Borakove, M. E., Fanfilk, P., Troutman, D., Johnson, N., Burgess, A., & O‘Connor, A.
L. (2006). Testing the efficacy of SANE/SART programs: Do they make a difference in
sexual assault arrest and prosecution outcomes? Washington, DC: National Institute of
Justice.
Orbst, P., Smith, S. G., & Zinkiewicz, L. (2002). An exploration of sense of community, Part 3:
Dimensions and predictors of psychological sense of community in geographical
communities. Journal of Community Psychology, 30, 119-133.
Oregon Attorney General‘s Sexual Assault Task Force (2009). SART handbook version III.
Salem, OR: Author.
Palm, K. M., & Follette, V. M. (2008). Sexual victimization and physical health: An examination
of exploratory mechanisms. Journal of Child Sexual Abuse, 17(2), 117-132.
Patterson, D., Greeson, M. R., & Campbell, R. (2009). Protect thyself: Understanding rape
survivors’ decisions not to seek help from social systems. Journal of Health & Social
Work.
Patton, M.Q. (2008). Utilization‐focused evaluation (4th edition). Thousand Oaks, CA: Sage.
Pennsylvania Coalition Against Rape (2002). Sexual assault response team (SART) guidelines.
Enola, PA: Author.
Plichta, S. B., Vandecar-Burin, T., Odor, R. K., Reams, S., & Zhang, Y. (2006). The emergency
department and victims of violence: An assessment of preparedness to help. Journal of
Health and Human Services Administration, 29, 285-308.
Post, L. A., Biroscak, B. J., & Barboza, G. (2011). Prevalence of sexual violence. In J.W. White,
M. P. Koss, & A. E., Kazdin (Eds.), Violence againstwomen: Mapping the terrain
(volume 1) (pp. 101-124). Washington DC: American Psychological Association.
Rambow, B., Adkinson, C. Frost, T. et al. (1992). Female sexual assault: Medical and legal
implications. Annals of Emergency Medicine, 21, 727-731.
Raphael, J., & Logan, T. K. (2008). The use (and misuse) of data on rape: Restoring sexual
assault to the national agenda. White paper prepared for the CounterQuo Conference.

153

 

Retrieved from: http://www.counterquo.org/assets/files/reference/The-Use-and-Misuseof-Data-on-Rape.pdf
Rapkin, B. D., & Luke, D. A. (1993). Cluster analysis in community research: Epistemology and
practice. American Journal of Community Psychology, 21, 247-277.
Resnick, H., Acierno, R., Holmes, M., Dammeyer, M., & Kilpatrick, D. (2000). Emergency
evaluation and intervention with female victims of rape and other violence. Journal of
Clinical Psychology, 56, 1317-1333.
Resnick, H., Holmes, M., Kilpatrick, D., Clum, G., Acierno, R., Best, C. L., & Saunders, B. E.
(2000). Predictors of post-rape medical care in a national sample of women. American
Journal of Preventive Medicine, 19, 214-219.
Rodríguez, M. M. D., Baumann, A. A., & Schwartz, A. L. (2010). Cultural adaptation of an
evidence based intervention: From theory to practice in a Latino/a community context.
American Journal of Community Psychology, 47, 170-186.
Rogers, E. M. (2003). Diffusion of innovations (5th ed.). New York, NY: Free Press.
Roussos, S. T., & Fawcett, S. B. (2000). A review of collaborative partnerships as a strategy for
improving community health. Annual Review of Public Health, 21, 369-402. 
Rovi, S. & Shimoni, N. (2002). Prophylaxis provided to sexual assault victims seen at US
emergency departments. Journal of American Medical Women’s Association, 57, 204207.
Schafer, J., & Graham, J. W. (2002). Missing data: Our view of the state of the art.
Psychological Methods,7(2), 147-177.
Seidman, I., & Pokorak, J. J. (2011). Justice responses to sexual violence. In M. P. Koss, J. W.
White, & A. E. Kazdin (Eds.), Violence against women: Navigating solutions (volume 2)
(pp. 137-157). Washington DC: American Psychological Association.
Singleton, R. A., & Straits, B. C. (2005). Approaches to social research (4th ed.). New York,
NY: Oxford University Press.
Spohn, C., & Tellis, K. (2012). The criminal justice systems’ response to violence against
women. Violence Against Women, 18, 169-192.
Starzynski, L.L., Ullman, S.E., Filipas, H.H., & Townsend, S.M. (2005). Correlates of women’s
sexual assault disclosure to informal and formal support sources. Violence and Victims,
20, 417-432.
Tabachnick, B., G., & Fidell, L. S. (2007). Using multivariate statistics (2nd ed). Boston, MA:
Pearson.

154

 

Tjaden, P., & Thoennes, N. (2006). Extent, nature, and consequences of rape victimization:
Findings from the National Violence Against Women Survey. Washington, DC: National
Institute of Justice.
Ullman, S. E. (1996). Do social reactions to sexual assault victims vary by social support
provider? Violence and Victims, 11, 143-157.
Ullman, S. E. (2007). Comparing gang and individual rapes in a community sample of urban
women. Violence and Victims, 22, 43-51.
Ullman, S. E. (2010). Talking about sexual assault: Society’s response to survivors. Washington,
DC: American Psychological Association.
Ullman, S. E., Filipas, H. H., Townsend, S. M., & Starzynski, L. L. (2007). Psychosocial
correlates of PTSD symptom severity in sexual assault survivors. Journal of Traumatic
Stress, 20, 821-831.
Walsh, W. A., Banyard, V. L., Moynihan, M. M., Ward, S., & Cohn, E. S. (2010). Disclosure
and service use on a college campus after an unwanted sexual experience. Journal of
Trauma & Dissociation, 11(2), 134-151.
Wilson, D., & Klein, A. (2005). An evaluation of the Rhode Island sexual assault response team
(SART). Washington DC: National Institute of Justice.
Wolitzky-Taylor, K. B., Resnick, H. S., McCauley, J. L., Amstadter, A. B., Kilpatrick, D. G., &
Ruggiero, K. J. (2011). Is reporting of rape on the rise? A comparison of women with
reported versus unreported rape experiences in the national women’s study replication.
Journal of Interpersonal Violence, 26, 807-832.
World Health Organization (2003). Guidelines for medico-legal care for victims of sexual
violence. Geneva: Author.
Yuan, N. P., Koss, M. P., & Stone, M. (2006). The psychological consequences of sexual
trauma. Harrisburg, PA: VAWnet, a project of the National Resource Center on
Domestic Violence/Pennsylvania Coalition Against Domestic Violence. Retrieved from:
http://www.vawnet.org
Zajac, J. J. (2006). Report on the national needs assessment of Sexual Assault Response Teams.
Enola, PA: National Sexual Violence Resource Center.
Zajac, J. J. (2009). National sexual assault response team survey report. Enola, PA: National
Sexual Violence Resource Center.
Zakocs, R. C., & Edwards, E. M. (2006). What explains community coalition effectiveness?
American Journal of Preventive Medicine, 30, 351-361.
Zinzow, H. M., Resnick, H. S., Amstadter, A. B., McCauley, J. L., Ruggiero, K. J., & Kilpatrick,
D. G. (2010). Drug- or alcohol-facilitated, incapacitated, and forcible rape in relationship
155

 

to mental health among a national sample of women. Journal of Interpersonal Violence,
25, 2217-2236.
Zinzow, H. M., Resnick, H. S., McCauley, J. L., Amstadter, A. B., Ruggiero, K. J., & Kilpatrick,
D. G. (2010). The role of rape tactics in risk for posttraumatic stress disorder and major
depression: Results from a national sample of college women. Depression and Anxiety,
27, 708-7

156