Friday, April 27, 2018

The Folly of Counting Bodies: Using Regression to Transgress the State of Serial Murder Classification Systems

In their study of the self-reported psychopathy and criminal thinking of serial homicide
offenders, Culhane, Walker & Hildebrand (2017) summarize the unique difficulty facing
researchers regarding the definition of a serial killer in that definitive statements cannot be made
about these offenders without first defining the target population. In Too Few Victims, the
authors use a faulty mechanism of measurement – the criteria of body count – hoping to end the
debate of what constitutes a ‘true’ serial homicide series. In seeking an empirical rationale for
eliminating two-victim offenders from consideration as ‘true’ serial murderers, the author’s
narrow the scope of which offenders should be included data from the since defunct Serial Killer
Information Center (SKIC)1. Although the SKIC data collection team suspected from the outset
that regional coverage of news articles could contribute to how serial murderers are perceived
(Wiest, 2016) and could influence the “large differences between offenders with two-victims and
those who killed greater numbers” but the authors code certain variables in a somewhat ad hoc2
manner to prove that these murderers are “noticeably dissimilar from established serial homicide
offenders” and that the “conventional wisdom regarding the common traits of serial killers” is
confirmed. Further confusing matters is the reference by Kiger (1990) to the designation of the
number of killings that must occur before the incidents are counted as serial murder as arbitrary.
Wright, Pratt and DeLisi (2009) agree and characterize the protection of this kind of definitional
territory as theoretically irrelevant and methodologically defenseless. Still, the authors aim to
reclaim ground lost to supporters of an expansive view of how best to define the term serial
murder, referring to themselves as traditionalists, or critics of this approach. This commentary
addresses the current research and its capacity to return serial homicide research to an age of idolatry where the deeds of the serial killer are glorified and their worth measured by the subjective concept of deadliness.

A preoccupation with ordering offenders into groups – from would-be and typical serial
killers to those that attained prolific status – formed the basis of the study design of Too Few
Victims, even in light of implorations to stop ranking multiple murderers using these types of
superlatives (Marketwatch, 2017). Although Gibson (2010) postulates that “definitional diversity
is not a mere matter of semantics”, the authors choose to focus on such nuances when
categorizing serial killers by tabulating completed acts. This approach is problematic given what
we know about the frequent failings of offenders when attempting to achieve their imagined
goals (Clarke, 2017) and because multiple homicide offending is often itself a sensationalistic
episode imbedded within a larger criminal career (Delisi & Scherer, 2006). The juxtaposition of
the words ‘too few’, ‘only’ or ‘just’ with the concept of victimization in the current research not
only minimizes the actions of multiple homicide offenders and the lives of their victims while
giving agency to those able to take advantage of favorable circumstances to become successful
but also disregards the role circumstance and luck play in the careers of the serial murderer and
their eventual collapse (Yaksic, 2017). It is not known if the same factors that allow more lethal
predators to become “deadlier, infamous and ruthless” keep what are referred to here as “lowlevel
offenders” below the author’s recommended threshold because they are absent from
consideration in the author’s analysis.

By trying to predict aspects of the offender’s modus operandi and motive based on the
number of victims amassed, the current research frames the analysis in a backward fashion while
searching for statistically meaningful differences between serial killers who killed upwards of
two-victims. Separate regressions for each individual characteristic3 would have proven the
author’s point more convincingly, especially if the coefficients on the dummy variables representing two-victim or three-victims were consistently statistically significant. A second methodological weakness is the lack of consideration for victim typology since controlling for it could have helped to explain many of the variations observed between two-victim and other serial killers. The nature of the victim-offender relationship, characterized by Fox and Levin (1998) as one of the most striking dissimilarities between serial murder and criminal homicide, is neglected in the current research because the data on victims that was supplied to the authors was in the rudimentary stages of being compiled. The state of the victim data also accounts for the inability of the authors to authentically code rape and torture in the manner they claim since it would not be possible to differentiate the precise number of victims raped or tortured over a series, whether it is most or all. Perhaps if data from the United Kingdom had been included in the current research more information on victims would have been available due to the outward nature of how the media handles them (Wiest, 2016).

Differentiating offenders based on their level of lethality discounts not only the
oftentimes situational variables that impact the outcomes of these events but, as Beasley (2004)
notes, variations in post-offense behavior, transportation used, items taken or kept and media
awareness. For example, Howell Donaldson may have either been apprehended sooner if police
gave more credence to the criminal profile supplied to them (Sampson, 2017) or not at all if he
had not surrendered his firearm. If surveillance footage had not contributed to Brandon Scott
Lavergne’s capture after his second homicide, he may have gone on to kill other women aside from his two-victims in 1999 and 2012 (McLaughlin, 2012). Yaksic (2015) points out that the notions of serial murderers performing the same crime scene behaviors throughout their series, consistently taking souvenirs or leaving signatures, escalating in their violence as they continue killing, improving their methods and changing their strategies over their careers are tested by Bateman and Salfati (2007) and Schlesinger, Kassen, Mesa, and Pinizzotto (2010). Disparities among these factors contribute to the dissimilarity between murders committed by the same individual while demonstrating the fallacy behind the concept of a conventional serial killer.

There will never be a constant and measureable progression from crime scene to crime scene in
what the offender does or reveals to a law enforcement organization (LEO) or what the LEOs
notice about that behavior – no matter the offender's body count – and even less in how much of
that quantifiable information is recorded in the SKIC data. Nevertheless, the authors firmly
believe that a stereotypical serial killer is characterized as an offender that partners with others
before killing for sexual enjoyment using torture and a singular method while taking totems.
While these five behaviors are used as ‘distinguishing attributes’ in the current research,
these variables may have been intentionally selected to best conform to offenders famously
examined by agents of the Federal Bureau of Investigation (FBI) Behavioral Science Unit. The
author’s desire to shoehorn current day serial murderers into archaic archetypes, now
mythologized in Netflix’s Mindhunter, is apparent when reviewing their measurement of motive
as “evidence of sexual assault, torture and totem-taking” and exclusion of female offenders
because they “rarely kill for sexual reasons”. These are thought to be the actions of sexual serial
killers that derive pleasure in their crimes but modern research has shown that the concept of
serial murder was wrongfully conflated with sexual violence due to the co-occurrence of both
crime trends (Reid, 2017). The authors both celebrate these old world views and stereotypes that
emerged from interviews conducted by FBI agents on 25 serial murderers (Ressler & Burgess,
1985) while denigrating the FBI’s recent work in adjusting the minimum victim threshold back
to its original form of two-victims (Federal Bureau of Investigation, 1994). The authors
characterize the FBI’s decision to alter this definition as a selfish endeavor and foregone
conclusion but also rely on the hallmarks of serial killing generalized from those early interviews
such as the erroneous belief – in light of new data stating otherwise – that serial murderers need a
“hands on” kill to be gratified. In defense of the FBI, they did not intend for the early research
done by its agents to be co-opted by academics, practitioners and researchers in the quick and
exacting manner in which each group has done. The FBI even gave representatives of these
groups the opportunity to participate in an open forum where the law enforcement version of the
serial killer was updated (Morton, Tillman, & Gaines, 2015).

Although the authors attempt to recast the mold of the ‘true’ serial murderer by truncating
the population down to such a narrow scope as to bring them more in line with outmoded
classification systems, the selection of partnership, rape, torture, totem-taking and enjoyment
variables may have been an artifact of working with a vastly incomplete database. Yaksic (2015)
noted that the SKIC’s database is far from complete as information across thousands of multiple
offenders is simply missing4. Imputations are relied on to fill those gaps rather than using what
Achen (2005) calls the classic skill of seeking out the missing information through independent
record review. The authors rigidly adhere to the use of empirical methods in Too Few Victims to
elucidate meaning from the contents of the database but that approach has resulted in a
romanticized view of serial killers and highlights, as Wright & DeLisi (2017) note, the gulf
between data points on a spreadsheet and the realities of modern day murderers. Extensive, systematic, and direct contact with serial murderers should ideally be coupled with empirical
data (Beasley, 2004) but another acceptable technique is explored by Reid (2018) where the
empirical is married with an exploration of the narrative behind the serial killer’s development.
Delving into the narrative behind the data points contained in the consolidated serial
killer database (Aamodt, Fox, Hickey, Hinch, Labuschagne, Levin, McClellan, Nelson, Newton,
Quinet, Steiger, White, & Yaksic, 2016) highlights the errors of lumping all two-victim
offenders together and excluding them from consideration. There is a potentially huge selfselection
problem in that two-victim killers by definition were not able to kill and hide from
police as long as multiple-victim killers. The characteristics of two-victim series provide some
possible explanations for this as these offenders are more likely to kill out of anger, less likely to
work with a partner and are slightly younger on average at first kill. Being stopped by police
after two-victims is different than desisting voluntarily after two-victims and never killing again.
In spite of the author’s statement that all two-victim offenders are “substantively different from
killers with higher victim counts”, other than circumstances sometimes beyond their control that
lead to their apprehension, there is fundamentally little to no difference between the actions of
the offenders cataloged in table 1. Although the authors insist that “the majority of two-victim
killers would likely never kill again, even given the opportunity”, there is nothing in the history
of these killers to suggest that they would have desisted from killing again if given the chance.
The authors view homicide as not featuring prominently in the lives of two-victim offenders as
they just happen to kill more than one person over time but they may be confusing the offenders
at the center of their research with the extremely small subgroup of recidivistic single-victim
homicide (RSVH) perpetrators (Bjorkly & Waage, 2005) that kill a second person after being
released from institutional confinement. Given that recidivism among homicide offenders is
lower than other forms of violent crime, it is imperative that two-victim offenders be viewed as
capable, motivated and unrelenting, as they have elevated their likelihood of victimizing a third
person merely by killing a second. A more substantive analysis would have been a comparison of
this group with the two-victim offenders and one-off killers as these groups may share more in
common than two-victim killers do with those that kill three or more victims.

Support for the observation that serial murderers can be found lying dormant among
those responsible for either a singular homicide or two murders separated in time has grown
among researchers over the past decade. Yaksic (2015) conducted a query using the modified
Delphi technique, finding that the use of two-victim terminology is acceptable to the majority of
experts queried5. Perhaps it is not a leap to assume these are serial killers in the strictest
interpretation of the term as preeminent researcher Jack Levin has stated that “when you find
two, you almost always find three" (Kirkey, 2010). Researchers that support the expansive
definition that includes these offenders focus on the offender’s state of mind or criminal intent,
reminding investigators that two, three, four and five victim offenders can share similar
characteristics not reflected by their body count (Mouzos & West, 2007). Wright, Pratt and
DeLisi (2009) prefer a more straightforward definition of multiple homicide that reflects the
notion that a multiple homicide offender is simply someone who has killed more than one
person. Homant and Kennedy (2014) acknowledge that a lack of opportunity can hamper an
offender’s ability to act on their impulses stating that someone who has killed once may well be
a serial killer, psychologically speaking. Schlesinger, Ramirez, Tusa, Jarvis, & Erdberg (2017)
voice the concern that another murder might occur at some distant time if one or two murders are
committed. In practice, the Murder Accountability Project believes that approximately five thousand people kill someone and escape apprehension each year with a percentage of these men and women having killed more than once (Wilkinson, 2017). But gone unanswered in the current research is the question asked by Wright, Pratt and DeLisi (2009): how does escalating to a third or fourth victim change the landscape of an offender’s criminal history?

Fox and Levin (1998) argued that establishing a minimum body count of four victims
helps to distinguish a repetition of killing as serial from other forms of homicide but admit
twenty years later that selection of this victim tally “may have been an arbitrarily chosen
threshold” (Levin & Fox, 2017). The current research aims to support this unfounded viewpoint
by placing the four-victim group at the forefront while rendering offenders in that group
“genuine” serial murderers. By honing in on outmoded aspects of serial murder, current day
offenders are held to those standards. Instead of crediting the preemptive action taken by law
enforcement to effect change in this arena or recognizing that society has forced a reduction in
the serial murderers’ victim threshold by taking note of their activities, the authors envelop this
progress by simply stating that two-victim offenders are “significantly different from all other
serial homicide offenders”. As former federal investigator Bill Daley stated regarding the decline
in serial homicide, "We're able to get people right after they commit one murder as opposed to
waiting until they've committed several more murders" (Corbin, 2017). No credence is given by
the authors to the possibility that the old ways in which serial killers functioned are no longer
possible and that offenders operating in the modern age face obstacles not in place during the
phenomenon’s heyday (Yaksic, DeSpirito, Reid, 2018). The “fundamental differences” between
offender classes mentioned in Too Few Victims could be a byproduct of adaptive behavior
undergone by these killers after realizing that they can no longer take advantage of the ignorance
of potential victims and LEOs.

The contributions of researchers aiming to influence the work done by LEOs in
operationalizing the more traditional, three victim minimum may derail early work to form task
forces designed to intervene in an offender’s series before additional victims can be claimed. The
author’s recommendations handicap others from viewing certain homicide series, so-called
mundane killings such as those recently occurring in Tampa (Dawson, Altman, Guzzo,
Contorno, & Sampson, 2017) or Louisiana (Toohey, 2017), as serial murders because they
characterize that phenomenon as an “extreme form of violence”. Because of such discordance,
LEOs often disregard the application of serial killer terminology to unknown offenders (Toohey,
Skene & Discher, 2017; Sanchez & Levenson, 2017) and acknowledge that what these offenders
are called does not affect operations, in part due to a past reliance on archaic stereotypes and
outmoded thinking by researchers (Corbin, 2017). Because LEOs cannot wait for offenders to
conform to archetypes before triaging such cases to investigators’ for their immediate attention,
they prefer instead to pursue homicide investigations with suspected linkages both individually
and under the assumption they are linked.

Chief Dugan of the Tampa Police Department was recently vocal in his admonishment of
profilers, amateur detectives and researchers armed with knowledge about serial killers aiming to
assist in the Seminole Heights serial killer investigation, stating that speculation can harm the
efforts of law enforcement (Rosario, 2017). Because of a history of uninvited involvement from
outside parties, LEOs care not for relative risk percentages and have adopted a pragmatic “know
it when I see it” attitude when investigating homicides with suspected linkages in an effort to
avoid mishaps caused by conflicts. Muskegon County Sheriff Dean Roesler points to the "similar
patterns and characteristics” in the two homicides attributed to Jeffrey Willis (Gaertner, 2016).
Albuquerque Police Department homicide Sgt. Elizabeth Thomson said that “similarities in the
deaths struck her as soon as she arrived at the site where the second body was found” (Kaplan,
2017). As Chief Steve Caraway stated, "His [Laurent] intention was to commit a murder. This
was something that was planned". Deputy prosecutor Wayne Tashima noted that because Saito
“committed a murder he has the inherent ability to commit another murder" (Kawano, 2017).
Tulsa Police Department Homicide Sgt. Dave Walker referred to Brennon Lovett as an
“interesting study in killing” due to the high number of shooting attempts attributed to him and
his “mindset”, calling him a “serial killer who’d only been ‘successful’ once” (Goforth, 2016). In
summary, the offender’s style of victimization signals a particular psychological proclivity to
reoffend to those on the frontlines combating serial murderers, which LEOs assume will hold
constant until proven otherwise.

According to the authors, proponents of a definition based on intentionality fail to explain
how this ‘underlying psychological construct is to be evaluated’ but that could be because
DeLisi, Tahja, Drury, Caropreso, Elbert, & Heinrichs (2017) introduced the concept of homicidal
ideation to the criminological literature only recently. Reisner, Mcgee, & Noffsinger (2003) used
the case of Mr. X, an individual self-described as a budding serial killer, to demonstrate that
there are gradations of deviancy across the spectrum of multiple homicide, from those that are
satisfied with their existence in the realm of mere homicidal ideation (Reisner, Mcgee, &
Noffsinger, 2003), others that operate at the behest of compliance with the outcomes of deep
seeded fantastical thinking (Murray, 2017), those engaging in purely fictive and fanciful larks for
anticipated attention it would garner (Fischer, Beckson, & Dietz, 2016) to all others that strive to
successfully carry forth their plans (Aamodt, et al., 2016). Perhaps the “path to intended
violence” model used by Allely and Faccini (2017) to understand a mass murderer can be
applied to serial homicide offenders to better comprehend where among the six stages on the
path things go wrong for their plans. Although there is journalistic interest in wannabe serial
killers (Specker, 2017), no study has yet drawn the distinction between them and would-be serial
murderers, what the current research deems “potential” serial killers. To that end, Yaksic (2018a,
working manuscript) has tracked the existence of offenders at or below the two-victim threshold
which includes the gamut of individuals at various stages of criminality6. The difficulty in
parsing out the fanciful from those that will inevitably fulfill their intentions arises due to a high
rate of false positives and the lack of an available mechanism to ethically test a prospective the
offender’s follow through to fulfill their supposed destinies. Because of these unknowns, it is
easy to cast them aside with two-victim offenders even though DeLisi et al. (2017) found
offenders with greater homicidal ideation to be more likely to have been arrested for murder and
evince more extensive and severe criminal careers.

Far reaching are the ramifications of excluding a wide swath of offenders from study
merely because little is known about the mechanisms that assist such offenders in the transition
to becoming a full-fledged serial homicide offender. The regressive viewpoint on what
constitutes the purity of a ‘true’ serial killer in Too Few Victims is based in the author’s fixation
on using sexual elements to separate offenders into requisite classes and can stymie real-world
efforts to identify and apprehend modern day serial murderers that kill more than once. In light
of this, most detectives recognize the earmarks of serial homicide activity after the occurrence of
the offender’s first or second homicide while practitioners and researchers work furiously to catch up. The Murder Accountability Project conducts a training exercise related to the singular murder of Jon Benet Ramsey to test the likelihood that her death is related to others in a series. In their review of investigative challenges and failures related to serial murder cases, LePard, Demers, Langan, & Rossmo (2015) used a threshold of one or more missing persons and did not list definitional discordance as slowing down or otherwise negatively affecting inquires. Activist Mikki Kendall believes in the possibility that the murder of Theresa Bunn is connected to the killing of a second women, as did Chicago Police detectives at the time (Vice News, 2017; Ahmed & Rozas, 2007). Years earlier, Chicago Police probed linkages between the murders of Angela Jones and Roberta McKinney (Wilson, 2000). Rather than relying on the variables identified in Too Few Victims, investigators exploring potential linkages between one homicide and another generally review tangible aspects of the perpetration of the crime such as the isolation of the victim, manner of killing, disposal of the body, post-mortem mutilation, concealment, timing and victim selection (Pakhomou, 2004). Similarly, Mott (1999) found that body disposal, victim vulnerability, location of body disposal and offender mobility were helpful in predicating the status of the case, whether it be solved or unsolved.

The authors claim access to the largest serial killer database in existence yet utilize the
fewest variables when compared to other research efforts that employ empirical techniques.
Kraemer, Lord, & Heilbrun (2004) utilized demographic information on the offender and the
victim, information on the offender’s style of approach, use of a vehicle, location of events,
symbolic artifacts or writings at the crime scene, positioning of the body, use of restraints, the
body’s state of dress, use of weaponry, and medical forensic information including cause of
death, location and type of trauma, and evidence of sexual assault. Lester and White (2014)
capture aspects of the offender’s life history and the circumstances of the murders across 83
different variables. Pakkanen, Zappala, Bosco, Berti, Santtila (2015) focus on the location of the
victim’s body, evidence of sexual trauma, where the murder occurred and the offender’s level of
forensic awareness. The analyses performed by Gurian (2015) included specific offender patterns
such as mobility, time frame, victim selection, and method. In their descriptive analysis of 92
offenders, Morton, Tillman, & Gaines (2015) focused on factors such as an offender’s victim
selection and approach, method of killing, prior criminal history, demographic details of the
offender and the victim, location where the victim was murdered, body disposal location and
sexual activity, grouping offenders based on victim counts of two to four, five to nine and ten or

To prevent a myopic view of the crime based on a single unusual factor, Morton,
Tillman, & Gaines (2015) recommend reviewing the case materials taking a “totality of
circumstances”. As such, Kraemer, Lord, & Heilbrun (2004) focused on intent rather than
motive while Gurian (2015) did not include motive in the final analysis due to the problems of
vague boundaries between categories, subjectivity in determining primary motive and lack of
empirical support. Beasley (2004) highlights how problematic identifying a precise motive can
be given the combinations of influences on an offender and the variability of human behavior.
Williams, Thomas & Arntfield (2017) demonstrate that there are still motives such as leisure that
have yet to be considered among those typically identified. Working with a smaller number of
offenders would have allowed the authors to conduct an in depth analysis using more variables
(Messori, 2016). The variation in variables and criteria used to design the study cohorts in the
current research and other empirical endeavors will prevent the direct comparison the authors
hoped their recommendation to regress to a three victim minimum threshold would have

In reconstructing the data used in the current research, it was found that it encompasses
2,395 offenders including 1,602 solo offenders, 287 operating in teams, 199 suspected, accused
or self-proclaimed, 81 spree killers and 226 members of an organization7. The high number of
individuals participating in organizations raises the question of homogeneity among offenders
within the chosen grouping schemes of two, three – seven and eight or more victims. Can it be
expected that hitmen and gang members that kill between three and seven victims share the same
typological makeup as other ‘typical’ serial murderers in the data? The absence of investigation
into the ‘cooling off period’ is also questionable due to the finding upon reconstruction that little
variation exists between the average amounts of time offenders remain at large8. The authors
assume that a four-victim offender’s series is exactly twice as long as that of a two-victim
offender and fail to acknowledge that these homicides could occur over a similar period of time.
Perhaps it would be more appropriate to organize offenders based on their status as either a
habitual or dedicated perpetrator, a dichotomy of historical depth introduced by Pakhomou
(2004) based on a scale of the offender’s criminal versatility, or conduct an analysis that explains
how such supposedly different groups can be at large for commensurate amounts of time.
The authors have four concerns related to lowering the victim threshold, stating that
doing so dilutes the population of serial killers, artificially increases the prevalence of serial
murder, changes the mathematical nature of the word ‘series’ and prevents direct comparison of results across studies. First, the argument that one diverse cohort can be watered down by the
addition of another diverse cohort lacks theoretical basis because Caucasian and African
American serial murderers are examined under the same umbrella of serial murder even though
there are several differences between them (Lester & White, 2014). Second, instead of increasing
the prevalence of serial murder, the inclusion of two-victim offenders has had the opposite
effect; researchers now know that serial homicide is declining, particularly among the group of
offenders the authors deem prolific. Third, because serial killers were originally called sequence
killers and a sequence can begin with one and either continue or end with a second, the semantic
argument is irrelevant. Finally, direct comparison of results across studies is complicated by
research teams that use different methods when manipulating the data and studies that classify
offenders by type of offense, not count.

Despite use of mechanical methods, identified as such by Achen (2005), the authors are
surprised to find that their theoretical insights differ from what they previously thought was
common among these offenders. These ‘new’ postulations are more in line with current
mainstream thought regarding the portrayal of serial murder, already widely accepted after
having been established in the expansive period of time between Douglas and Ressler and
today’s researchers. A consequence of abiding by the conclusions drawn by the authors in Too
Few Victims is a loss of investment in further research dedicated to an important group of killers
that allegedly “cannot be combined with any other offender type”. Such research should
investigate if the empirical distinction between two-victim serial killers and other multiplevictim
serial killers may simply be a statistical anomaly caused by the lack of data or lower data
quality for two-victim series9. As it stands, the regression results show that serial killers who kill for enjoyment and work with partners tend to kill more victims. However, the authors do not prove that serial killers who kill more victims necessarily have statistically different characteristics once all other exogenous variables are controlled for. k-means and Expectation-Maximization clustering is another unexplored avenue that could serve to reveal hidden aspects of the data by grouping serial killers into typological clusters in a principled manner based on various observed characteristics. If the average number of victims within the groups of serial killers produced by the algorithm would end up being different then the data would support the point made by the authors. If not, then it would provide a contradictory argument. In closing, the authors regard impeded research efforts and hampered communication as resultant from definitional discordance but recent research demonstrates that practitioners themselves can often contribute to institutional rivalries that slow efforts to understand these offenders (Yaksic, 2018b). The undue emphasis placed on body count, a niche aspect of serial homicide research, by the authors in Too Few Victims has the capacity to further alienate disparate groups of researchers and practitioners from each other and from those law enforcement officers they are aiming to help. By redirecting the focus back to issues that exist beyond definitional problems, future research will transcend mere regressions and deliver us all from the motivations of those seeking to transgress the field of serial homicide research.

1 The Serial Killer Information Center has not been updated in more than a year

2 e.g. “usual patterns” of rape or torture vs. “at least one” souvenir-taking vs. 80-percent threshold to establish who was a “partnered killer” and “singular-method killer” idolatry where the deeds of the serial killer are glorified and their worth measured by the subjective concept of deadliness.

3 e.g., Offender Characteristic 1 = f( Number of Victims, All Other Offender Characteristics, MO, Most Likely Motive) where “Number of Victims” would be dummy variables {0,1} representing Two-Victim, Three-Victim, Four-Victim killers, and onwards

4 Out of the 2,395 offenders meeting the author’s criteria, information in the totem variable, for example, was absent 64 percent of the time, or across 1,548 records.

5 Thirteen researchers, practitioners and law enforcement officers participated in the modified Delphi panel

6 1) described by police, prosecutors or practitioners as a fledgling or budding serial killer maintaining trademarks or hallmarks, characteristics, tendencies or traits; 2) offenders that desire, intend or plan to kill more than one person; 3) the self-proclaimed statements, utterances and admissions from offenders made to convince others of their capacity to kill serially; 4) attempting additional homicides occurring before or after successfully completed acts; 5) obsession with the aspirations to attain the serial killer designation; 6) membership in fandom dedicated to offering admiration towards actual serial killers and celebrating their exploits; 7) the presence of a serial mindset juxtaposed with numerous failures to amass enough victims; 8) incessant comparison to previous serial killers and a clearly delineated plan to emulate them

7 approximately 967 US based male offenders killed two people with a first kill since 1970. 730 acted as solo offenders, 97 operated in teams, 62 were suspected, accused or self-proclaimed, 20 were spree killers and 58 were members of an organization (drug, enterprise, cult or gang). Approximately 1,212 US based male offenders killed three, four, five, six or seven people with a first kill since 1970. 760 acted as solo offenders, 146 operated in teams, 120 were suspected, accused or self-proclaimed, 57 were spree killers and 129 were members of an organization. Approximately 216 US based male offenders killed eight or more people with a first kill since 1970. 112 acted as solo offenders, 44 operated in teams, 17 were suspected, accused or self-proclaimed, 4 were spree killers and 39 were members of an organization.

8 1,860 days or five years for two-victim killers; 1,500 days or four years for three-seven victim killers; 1,650 days or five years for killers of eight or more victims.

9 as 17.7% of all two-victim killers were coded under the “Other Motive” category


Kat Albrecht, Hester Brink, Heng Choon (Oliver) Chan, Steve Daniels, Simon Demers, Steve
Egger, Steve Giannangelo, Elizabeth Gurian, Sean Maddan, Wade Myers, Hannah Scott


Aamodt, M., Fox, J. A., Hickey, E., Hinch, R., Labuschagne, G., Levin, J., McClellan, J.,
Nelson, B., Newton, M., Quinet, K., Steiger, C., White, J., & Yaksic, E. (2016). Consolidated
Serial Homicide Offender Database. Harvard Dataverse. Retrieved from

Achen, C. H. (2005). Let’s Put Garbage-Can Regressions and Garbage-Can Probits Where They
Belong. Conflict Management and Peace Science. 22:327–339.

Ahmed, A. & Rozas, A. (2007). 1st murder victim identified: City cops canvass worried
neighbors. Chicago Tribune. Retrieved from

Allely, C. S., & Faccini, L. (2017). “Path to intended violence” model to understand mass
violence in the case of Elliot Rodger. Aggression and Violent Behavior. 37:201-209.

Bateman, A. L., & Salfati, G. (2007). An examination of behavioral consistency using individual
behaviors or groups of behaviors in serial homicide. Behavioral Sciences and the Law, 25(4),

Beasley, J. O. (2004). Serial murder in America: Case studies of seven offenders. Behavioral
Sciences & the Law, 22, 395–414.

Bjorkly, S. & Waage, L. (2005). Killing Again: A Review of Research on Recidivistic Single-
Victim Homicide. International Journal of Forensic Mental Health. 4(1):99-106.

Clarke, T. (2017). Is glorified murder to blame for new breed of killer? The West Australian.
Retrieved from

Corbin, C. (2017). Profile of a serial killer: Police hunt killer behind 4 Tampa murders. Fox
News. Retrieved from

Culhane, S. E., Walker, S., & Hildebrand, M. M. (2017). Serial Homicide Perpetrators’ Self-
Reported Psychopathy and Criminal Thinking. Journal of Society for Police and Criminal
Psychology. 1-13.

Davis, P. (2017). Anne Arundel judge: 'manipulative serial killer' deserved a tougher sentence.
Capital Gazette. Retrieved from

Dawson, A., Altman, H., Guzzo, P., Contorno, S., & Sampson, Z. T. (2017). Friends stunned by
arrest of suspected serial killer in Seminole Heights. Tampa Bay Times. Retrieved from

Delisi, M., & Scherer, A. M. (2006). Multiple homicide offenders: Offense characteristics, social
correlates, and criminal careers. Criminal Justice and Behavior, 33(3), 367–391.

DeLisi, M., Tahja, K., Drury, A. J., Caropreso, D., Elbert, M., & Heinrichs, T. (2017). The
Criminology of Homicidal Ideation: Associations with Criminal Careers and Psychopathology
among Federal Correctional Clients. American Journal of Criminal Justice. 42(3):554-573.

Douglas, J. E., Burgess, A. W., Burgess, A. G., & Ressler, R. K. (1992). Crime classification
manua. New York, NY: Simon & Schuster.

Elfrink, T. (2017). Fort Lauderdale Police Catch Serial Killer Who Wrote Message In Blood.
(2017). Miami New Times. Retrieved from

Federal Bureau of Investigation. (1994). Child Abduction and Serial Killer Unit, Morgan P.
Hardiman Task Force on Missing and Exploited Children [brochure].

Fischer, C. A., Beckson, M., & Dietz, P. (2016). Factitious Disorder in a Patient Claiming to be a
Sexually Sadistic Serial Killer. Journal of Forensic Sciences. 62(3):822-826.

Fox, J. & Levin, J. (1998). Multiple Homicide: Patterns of Serial and Mass Murder. Crime and
Justice. 23:407-455.

Gaertner, E. (2016). Why Jeffrey Willis would be considered a serial killer if convicted. Retrieved from

Gibson, D. C. (2010). Serial killing for profit: Multiple murder for money: Greenwood
Publishing Group.

Goforth, D. (2016). How did this elusive shooter avoid Tulsa police for three years? The
Frontier. Retrieved from

Gurian, E. A. (2015). Reframing Serial Murder Within Empirical Research: Offending and
Adjudication Patterns of Male, Female, and Partnered Serial Killers. International Journal of
Offender Therapy and Comparative Criminology. 61(5):544-560.

Henry, C. (2014). Woman says she escaped from Steven Zelich in 2001. Retrieved

Hepp, R. (2008). Trucker charged in Hunterdon killing hated women, those who knew him say. Retrieved from

Homant, R. J., & Kennedy, D. B. (2014). Understanding serial sexual murder: A biopsychosocial
approach. In W. Petherick (Ed.), Profiling and serial crime (3rd ed., pp. 341–372). Boston:
Anderson Publishing, Ltd.

Kaplan, E. (2017). Is ‘serial offender’ killing homeless men? Albuquerque Journal. Retrieved

Kawano, L. (2017). Documents: Hospital escapee had sexual relations with staff. Hawaii News
Now. Retrieved from

Kiger, K. (1990). The darker figure of crime: The serial murder enigma. In S. Egger (Ed.), Serial
murder: An elusive phenomenon (pp. 35–52). New York, NY: Praeger.

Kirkey, S. (2010). Serial killers become chameleons to get away with murder: Experts. Retrieved from

Kraemer, G. W., Lord, W. D., & Heilbrun, K. (2004). Comparing Single and Serial Homicide
Offenses. Behavioral Sciences & the Law. 22(3):325-343.

LePard, D., Demers, S., Langan, C., & Rossmo, D. K. (2015). Challenges in serial murder
investigations involving missing persons. Police Practice and Research. 16(4), 328–340.

Lester, D., & White, J. (2014). A Study of African American Serial Killers. Journal of Ethnicity
in Criminal Justice. 12(4):308-316.

Levin, S. (2013). Scottie Thompson, on Parole for Murder, Kills Dakota Jones in "Tragic,
Random" Slaying: Cops. Riverfront Times. Retrieved from

Levin, J., & Fox, J. (2017) Multiple Homicide: Understanding Serial and Mass Murder. In The
Handbook of Homicide. F. Brookman, E. R. Maguire, & M. Maguire (Eds.) Wiley.

Marketwatch. (2017). Biggest? Worst? Mass-shooting expert implores media to drop the superlatives.
Retrieved from

McAvoy, A., & Kelleher, J. S. (2017). Escaped Psychiatric Patient, Once Described as a Classic
Serial Killer, Could Be in California. Time. Retrieved from

McClure, S. (2002). ‘I just got away with murder’. The Tennessean. Retrieved from

McLaughlin, M. (2012). Brandon Scott Lavergne Pleads Guilty To Murdering Mickey
Shunick And Lisa Ann Pate. Huffington Post. Retrieved from

Messori, L. (2016). Frequencies between serial killer typology and theorized etiological factors.
(Unpublished doctoral dissertation).

Moriarty, T. (2017). Prosecutors: N.J. man killed 3 women, tried to kill 4th. Retrieved

Morton, R. J., Tillman, J. M., & Gaines, S. J. (2015). Serial murder: Pathways for investigations.
National center for the analysis of violent crime. Washington, D.C.: Federal Bureau of
Investigation. Retrieved from

Mott, N. (1999). Serial Murder: Patterns in Unsolved Cases. Homicide Studies. 3(3)241-255.

Mouzos, J. & West, D. (2007). An examination of serial murder in Australia. Trends and Issues
in Crime and Criminal Justice, 346. Canberra: Australian Institute of Criminology.

Murray, J. L. (2017). The Role of Sexual, Sadistic, and Misogynistic Fantasy in Mass and Serial
Killing. Deviant Behavior. 38(7).

Pakhomou, S. M. (2004). Serial killers: offender’s relationship to the victim and selected
demographics. International journal of police science and management. 6(4): 219–233.

Pakkanen, T., Zappala, A., Bosco, D., Berti, A., & Santtila, P. (2015). Can hard-to-solve one-off
homicides be distinguished from serial homicides? Differences in offence behaviours and victim
characteristics. Journal of Criminal Psychology. 5(3):216-232.

Reid, S. (2017). Compulsive Criminal Homicide: A New Nosology for Serial Murder,
Aggression and Violent Behavior. Aggression and Violent Behavior. 34:290-301.

Reid, S. (2018) Taking back narrative control: an in depth analysis of serial killers reflections of
their development. Retrieved from

Reisner, A. D., Mcgee, M., & Noffsinger, S. G. (2003). The Inpatient Evaluation and Treatment
of a Self-Professed Budding Serial Killer. 47(1):58-70.

Ressler, R. K., & Burgess, A. W. (1985). The men who murdered. FBI Law Enforcement
Bulletin, 54(8), 2–31.

Ressler, R. K., Burgess, A. W., & Douglas, J. E. (1988). Sexual homicide: Patterns and motives.
Lexington, MA: Lexington Books.

Rosario, S. (2017). Dugan to amateur detectives: Stop spreading Seminole Heights rumors,
WTSP. Retrieved from

Sampson, Z. (2017). Experts have some theories on who's carrying out Seminole Heights
killings. Tampa Bay Times. Retrieved from

Sanchez, R. & Levenson, E. (2017). Tampa police release new video in hunt for mystery killer of
3. CNN. Retrieved from

Schlesinger, L. B., Ramirez, S., Tusa, B., Jarvis, J. P., & Erdberg, P. (2017). Rapid-Sequence
Serial Sexual Homicides. J Am Acad Psychiatry Law. 45:72–80.

Schlesinger, L. B., Kassen, M., Mesa, V. B., & Pinizzotto, A. J. (2010). Ritual and signature in
serial sexual homicide. Journal of the American Academy of Psychiatry and the Law, 38(2),

Sparacello, M. (2011). Bridge jumper tied to Kenner motel killing by phone, license, debit card.
The Times-Picayune. Retrieved from

Specker, L. (2017). Serial killer or wannabe? Expert skeptical in Mobile case.
Retrieved from

The Republic. (2016). What to know about the Phoenix canal murders. Retrieved

Thompson, P. (2010). 'I will kill, I know I will... I am an animal': The chilling TV warning from
the man who murdered Chelsea King and Amber Dubois. MailOnline. Retrieved from

Toohey, G. (2017). Alleged East Feliciana serial killer pleads not guilty Tuesday to charges in
two of four shootings. The Advocate. Retrieved from

Toohey, G., Skene, L., & Discher, E. (2017). New reality after East Feliciana shootings:
Changing routines, unfamiliar sense of unease. The Advocate. Retrieved from

Vice News. (2017). Is there a serial killer roaming the streets of Chicago?
Retrieved from

Wiest, J. B. (2016). Casting Cultural Monsters: Representations of Serial Killers in U.S. and
U.K. News Media. Howard Journal of Communications. 4:327-346.

Wilford, G. (2017). Bondage murder trial suspect 'wanted to become serial killer after fleeing
UK'. The Independent. Retrieved from

Wilkinson, A. (2017). The Serial-Killer Detector. The New Yorker. Retrieved from

Williams, D. J., Thomas, J. N. & Arntfield, M. (2017): An Empirical Exploration of Leisure-
Related Themes and Potential Constraints across Descriptions of Serial Homicide Cases, Leisure
Sciences. Leisure Sciences. 1-16.

Wilson, T. (2000). 2 Stranglings, Rape May Be Linked, Cops Say: Police Probing Attacks On
Women On Far South Side. Chicago Tribune. Retrieved from

Wright, K. A., Pratt, T. C., & DeLisi, M. (2009) Multiple Homicide Offenders: Arbitrary Cut-
Off Points and Selection Bias. Homicide Studies. 13 (2): 193-199.

Wright, J. P, & DeLisi, M. (2017). What Criminologists Don’t Say, and Why. City Journal.
Manhattan Institute for Policy Research. Retrieved from

Yaksic, E. (2015). Addressing the challenges and limitations of utilizing data to study serial
homicide, Crime Psychology Review. 1:1, 108-134.

Yaksic, E. (2017). Novel Techniques to Detect the Multiple-event Murderer: An Exploration of a
Researcher’s Field Notes. In Violent Offenders: Theory, Research, Public Policy and Practice.
M. DeLisi & P. J. Conis (Eds.) MA: Jones & Bartlett Learning.

Yaksic, E. (2018a). Would-bes and Wannabes: An Analysis of the Unmet Desires of Potential
Serial Killers. Working manuscript.

Yaksic, E. (2018b). The Perpetual Influence of Dark Traits on Alienists: Dissecting the Use of
Transgressive Behavior and Antagonistic Schemes to Obtain Esteem by Inhibiting Social
Cohesion and Altering Functional Norms, Routledge International Handbook of Psychopathy
and Crime. In press.

Yaksic, E., DeSpirito, L., & Reid, S. (2018). Detecting an Observable Decline in Serial
Homicide: Have We Banished the Devil from the Details? Draft manuscript.

No comments:

Post a Comment