Collaborative Partnerships and the DNA Testing of Unsubmitted SAKs

For many years, recommendations have been made for law enforcement agencies to form collaborative working relationships with local academic institutions. The benefit of these collaborations to law enforcement is acquisition of research capabilities beyond what most agencies possess internally; for academic institutions, collaboration can provide access to a highly sought-after commodity—law enforcement data.  

Currently, an exciting opportunity exists in the United States for this type of collaborative partnership, stemming from the Bureau of Justice Assistance’s National Sexual Assault Kit Initiative to forensically test and follow up (via investigation and prosecution) on evidence obtained from unsubmitted sexual assault kits (SAKs).1 Over the past several decades, hundreds of thousands of SAKs were collected and preserved across the United States, but many of them were never submitted for DNA testing. This initiative to test previously unsubmitted SAKs provides an opportunity to access previously untapped data, having the potential to significantly increase what is known about sexual assault and those who perpetrate it.  

An example of a productive law enforcement–academic collaborative working relationship currently exists in Cuyahoga County (Cleveland), Ohio. In 2013, the Cuyahoga County Prosecutor’s Office established the SAK Task Force (Task Force), which consists primarily of investigators, victim advocates, and prosecutors working as a team to follow up on the testing of nearly 5,000 previously unsubmitted SAKs in the county.2 Researchers from the Begun Center for Violence Prevention Research and Education at Case Western Reserve University were granted unprecedented access to be an embedded research partner—conducting participant observation at Task Force meetings and trainings, interviewing Task Force members, and obtaining access to the prosecutor’s office case management system, which includes original police reports, investigative reports, medical records, criminal histories, and laboratory reports. This level of access was granted with the understanding that successful collaborations require a high level of trust and communication among partners. 

Data Resulting from the Collaboration 

The research team created a database of the case files from the prosecutors’ office that encompassed a 20-year period (1992 to 2012, with the majority of sexual assaults in the files occurring from 1993 through 1999), which has, thus far, resulted in a sample of 433 completed case files that were coded for more than 500 unique variables. The variables included information on the offender, the victim, the SAK, the sexual assault, the investigation (then and now), and the prosecution (then and now).  

Access to such a large quantity of high-level law enforcement data permitted researchers to make some important findings regarding serial sex offenders. 

These data provide an opportunity to greatly expand on what is known about serial sex offenders. Linkages between multiple sexual assault reports made via DNA provide for more expansive and more objective data than previous research on the topic, which primarily relied on uncovering multiple convictions or offenders’ self-reporting of multiple offenses.  

Much of the past research on serial sex offenders has been limited to those who have sexually assaulted multiple strangers. However, the current DNA-derived data provide the ability to identify a broader group of serial sex offenders (i.e., those who sexually assault only strangers, those who sexually assault only non-strangers, or those who sexually assault both strangers and non-strangers). Findings from this collaborative initiative in Cuyahoga County show that offenders appear to frequently “cross over”, sexually assaulting both strangers and non-strangers (also known as sexual polymorphism), often drastically varying their offending patterns (or modus operandi or MO) across offenses.   

Key Findings 

“StrangerOnly” Offenders  

While the sample of cases collected by the research team disproportionally contains serial sex offenders, and sexual assaults committed by strangers, the researchers found that 57 percent of serial sex offenders with more than one unsubmitted SAK coded in their sample werestranger-only” offenders.3 “Stranger-only” offenders tended to exhibit the greatest amount of consistency in their MO across offenses. The majority consistently (in all the sexual assaults linked to them) moved or transported victims from the site of the first contact and frequently made the first contact with the victims outdoors or outside; half consistently used weapons to threaten victims; and half consistently used some form of bodily force during the sexual assault.  

However, these stranger-only offenders exhibited inconsistency when examining the age of their victims—they frequented sexually assaulted individuals who were considerably younger and considerably older than themselves. 

Additionally, 70 percent also varied their means of accessing victims across offenses. The most common means of accessing a victim, by this type offender was an immediate attack (“blitz”), followed by (in order of frequency): forcing the victim into a car; offering the victim a ride; and, finally, taking advantage of the victim’s offer of or request for assistance. 

“Stranger and Non-Stranger” Offenders  

Over a quarter (28 percent) of serial sex offenders in the sample (which, again, is disproportionally comprised of sexual assaults committed by strangers) sexually assaulted both strangers and non-strangers. These offenders significantly varied their offending patterns across offenses concerning the victims’ ages, places of the first contact, means of access, use of weapons, and use of force. For example, one offender sexually assaulted a 13-year-old girl (a stranger) at a party and, two months later, sexually assaulted his 3-year-old son. Another offender sexually assaulted two females in separate incidences and, in a third offense, sexually assaulted a sleeping 29-year-old man living in a group home. Of the three groups, this group varied their offending patterns the most. 

“Non-StrangerOnly” Offenders  

Over 15 percent of serial sex offenders in the sample assaulted only non-strangers. Their level of consistency in offending pattern was high, similar to that of stranger-only offenders. However, unlike what was observed in stranger-only offenders, this consistency extended to their victims’ ages, which was often similar to their own. In most cases, the offender’s relationships with the victims were found to be either casual or a recent acquaintance. Most non-stranger assaults occurred in residences and did not involve weapons or bodily force.  

Key Takeaways 

Based on the findings presented here, some long-standing beliefs regarding serial sexual offenders may now be in question, including the perception that serial sex offenders consistently show preference for a certain “type” of victim. Investigators often attempt to identify unknown assailants by linking crimes with similar offending patterns, especially patterns that pertain to victims of similar race, age, sex, and relationship to the offender. The research coming from Cuyahoga County shows that serial sex offenders follow an offending pattern … until they don’t. Therefore, relying on traditional offending pattern (e.g., race, age, sex, relationship) analysis has the potential to result in overlooking sexual assaults that are linked but do not fit the pattern.  

Additionally, the sheer number of serial sex offenders identified as a result of this initiative has been surprising to all involved.4 This outcome has important implications for current investigative policy and practice, impacting how law enforcement views victims and offenders as well as the crime itself. This finding identifies the need to update training for first responders on conducting victim interviews and collecting forensic evidence in what can likely be a series of sexual offenses. It also identifies the need to provide follow-up investigators with updated training and resources to thoroughly investigate suspects for the possibility of serial perpetration. The recommended training should emphasize viewing sexual assault offenses as being course-of-conduct—based crimes rather than being incident-based crimes. This will shift the primary focus of these investigations to an offender’s ongoing (serial) behavior, mirroring an approach that has been successful in domestic violence investigations. 

The collaborative law enforcement–academic working relationship in Cuyahoga County is producing research findings that are helping shape new evidence-based strategies for the investigation and prosecution of sexual assault. All across the United States, the push to test previously unsubmitted SAKs and the reinvestigation of those cases are providing the same opportunity for others to gather these critically important data. In addition, establishing a collaborative working relationship with local academic institutions provides law enforcement executives the opportunity to publicly demonstrate their commitment to better serving sexual assault victims while holding perpetrators accountable and making their communities safer.  

Dan R. Clark is a contracted consultant with IACP for the OVC Integrity, Action, and Justice: Strengthening Law Enforcement Response to Domestic and Sexual Violence National Demonstration Initiative. In 2017, he worked as a research assistant with the Begun Center for Violence Prevention Research and Education at Case Western Reserve University after spending the previous two years as a special investigator with the Sexual Assault Kit Task Force in the Cuyahoga County Prosecutor’s Office. Mr. Clark served 26 years in the City of Lakewood, Ohio, Division of Police, retiring as chief.  

Rachel Lovell, PhD, is a sociologist and methodologist who studies gender-based violence and victimization, particularly sexual assault, human sex trafficking and sex work, and intimate partner violence. She is the lead researcher on the Cuyahoga County Sexual Assault Kit (SAK) Research Project, which is an action research study that examines untested SAKs in Cuyahoga County, Ohio. She is a senior research associate at the Begun Center for Violence Prevention Research and Education at the Jack, Joseph and Morton Mandel School of Applied Social Sciences at Case Western Reserve University. Dr. Lovell received her PhD in sociology from the Ohio State University in 2007.


1 Bureau of Justice Assistance, “National Sexual Assault Kit Initiative,” March 14, 2018.  
2 Cuyahoga County Office of the Prosecutor, “Department of Justice Awards $2 Million to Cuyahoga County Sexual Assault Kit Task Force,” news release, September 10, 2015.
3 “Serial sex offenders” are defined as individuals associated “with more than one linked SAK (‘kit[s]-to-kit’) … or one SAK and at least one arrest” for a registerable sexually-based offense—an offense that, if convicted, would have required them to register as a sex offender in Ohio “(‘kit[s] plus criminal history’).” Rachel Lovell et al., “Offending Patterns for Serial Sex Offenders Identified via the DNA Testing of Previously Unsubmitted Sexual Assault Kits,” Journal of Criminal Justice 52 (September 2017): 68–78. 
4 The disproportionate number of serial sex offenders is due to their prioritization by the Task Force for investigation and prosecution, and the high number of stranger-only cases is due to the sample consisting of cases that were not successfully prosecuted.