Profiling Serial Killers: Literature Review

Abstract

This essay reviews the literature on serial killers, principally the more newsworthy ones in contemporary times. In the process, we recognize that gender, age, child-rearing, social adjustment, the need to over-compensate for other areas of weakness in one’s life, and introversion are strongly associated with serial killers. We then examine two qualitative research methods, case study, and content analysis, with an eye to informing more rigorous research in the future.

Introduction and Literature Review

Definition of a Serial Killer: A series of two or more murders, committed as separate events usually, but not a/ways, by one offender acting alone. The crimes may occur over a time ranging from hours to years. Quite often the motive is psychological, and the offender’s behavior and the physical evidence observed at the scene will reflect sadistic and sexual overtones.

The Definition of a Serial Killer (Brooks et al., 1988)

Beginning with Jack the Ripper, the most notorious and mysterious of them all, serial killers have terrorized entire cities and hogged headlines. The fact that they generally operated alone, never exposed themselves to witnesses, and evaded police manhunts for long periods of time (principally because there was nothing to link them to their victims no matter what evidence or clues they left behind) increased the mystery that shrouded them and had criminologists casting about for any method that might lead to their apprehension. When truly desperate, detectives have not been above consulting psychics.

In the six score years since Jack the Ripper ran riot in the anonymity and low life of London’s Whitechapel, there have been repeated attempts to apply personality theory to patterns of criminal behavior. The underlying assumption has been that aberrant behavior is associated with certain personality traits. Thus, the British Criminal Investigation Department circa 1888 asked a Dr. Thomas Bond to describe the likely perpetrator based on the modus operandi of the killings of prostitutes, going to the extent of showing him the cuts that seemed to reveal a surgeons’ skill with human anatomy (Kocsis, 2004). This effort came to naught as the Ripper was never found.

Modern psychology gained enormous credibility nearly 60 years later, however, when a long series of killings afflicted New York from the mid-1940s to 1956. Called in to consult, the psychiatrist James Brussel deduced an astoundingly detailed description of an unmarried immigrant from Europe in his 50s, meticulous to the point of compulsion, fond of the color blue, and fussy about wearing button-down suits. The prediction proved amazingly accurate and precisely fit the suspect when police finally caught up with him. Arrested for the bombings, George Metesky was found to be an unmarried man of European descent in his 50s who at the time of his arrest was wearing a blue suit with all the buttons neatly done up.

Against this striking success was the case of abject failure nearly half a century later when John Allen Muhammad went on a three-week killing spree in Maryland, Virginia, and Washington. Muhammad and young accomplice John Lee Malvo eluded police for so long, managing ten killings (and a handful of woundings) because he fit the profile of serial murderers in important aspects but diverged on three very visible characteristics (Leonard, 2002).

Lying prone in the trunk of a car and shooting through a hole just big enough to let him zero in on a random slew of victims with a sniper sight, Muhammad was reported to match the profile of notorious serial murderers mainly in point of attention-seeking, having had military training and an accomplice. Wanting to glorify himself was manifested in leaving provocative clues for the police. On the other hand, Muhammad was all of 41 whereas most serial killers were in their 20s and 30s. The criminologist and FBI profiler who helped build a psychological profile of New York’s “Son of Sam” declared that “It’s not traditional in any way, shape or form…It’s almost a unique thing” (Robert Ressler, as cited in Kelly, 2002). It was Ressler, however, who had contributed the insight that the Washington sniper must have had the benefit of military training and went around with an accomplice. A bit of research, Ressler claims, would have shown anyone that one-fourth of serial killers operated with a helper, partly because the killers themselves are “dominant psychopaths” and delight in the control that a willing accomplice hands to them.

But the police were misled by the unshakable opinion of psychologists and criminologists brought into the case about what they held to be firm elements of the serial killer profile. Prior to Muhammad, for example, the body of evidence from all previous cases in America had led profilers to expect that the perpetrator would be a White man; Muhammad was the first Negro ever apprehended for multiple killings of victims chosen at random. A second misleading element, according to the Brudnick Center on Violence and Conflict at Northeastern University (Levin, cited in Leonard, 2002) was that serial murderers usually delighted in killing as an end in itself. Muhammad made demands much as a kidnapper or terrorist would, a first in the grisly catalog of serial killers.

Thirdly, the rage attributed to him had till then been typical of mass murderers more than serial killers. Soon after apprehending him at a rest stop in late October 2002, federal investigators determined the sorry circumstances of his life and the wellsprings of an abiding frustration: a failure at marriage, mediocrity during a stint in the military, and a business bankruptcy that threw him out on the streets. Prior to the shooting spree, Muhammad had muddled along as a petty thief. Clearly, he was a loser looking to make his mark, to lash back at society. On further consideration, there really was nothing to stop Muhammad from acting the mass murderer, shooting multiple victims in one day except for the fact that he had only unidirectional access from the car trunk and the low vantage point did not allow him to see more potential victims who scampered for cover after the first shot or two.

The picture that emerged from the early investigation – multiple failures, frustration, and rage that evidently boiled over into antisocial action – is really more typical of mass murderers or so Williams College psychology professor Saul Kassin attested from extensive research into the criminal mind (Leonard, 2002). In contrast, serial killers traditionally chose their victims with care and usually stopped between murderous episodes to conceal themselves in the anonymity of a community. This accounts for the fact that several cases of serial killing, solved or not, spanned years rather than the thankfully abbreviated period before Muhammad was apprehended.

Nonetheless, the sly concealment tactics Muhammad used can be considered typical of serial murderers who play a cat-and-mouse game with the police. Several that we know of even had the effrontery to try and make a mark with the media or the police. The Zodiac killer, whose reign of terror in San Francisco, spanned the 1960s and 1970s. sent over 20 coded missives to the police and was never caught. The “Unabomber”, Ted Kaczynski, transmitted a long manifesto to two newspapers with a demand for publication. While killing or wounding a total of 13 in New York, “Son of Sam” David Berkowitz positively gloried in the spotlight by constantly making columnist Jimmy Breslin his mouthpiece and stalking horse, as it were. That 8 in 10 serial killers are eventually caught can be attributed to their obsession with baiting the police and the press or their sheer carelessness with leaving evidence at crime scenes.

Profiling did a disservice in the case of the Washington sniper by steadfastly putting police “on the scent” of a white perpetrator in a white vehicle despite a number of witness statements that put Muhammad’s blue Chevrolet Caprice near the crime scenes. The probabilities based on all previous cases in the United States that had resulted in an arrest – Professor Hickey of Cal State in Fresno (Leonard, 2002) collected data on 399 serial murderers dating back to 1800 – did overwhelmingly favor a mainstream racial profile. In hindsight, it becomes clear that blacks are actually over-represented among serial killers. Comprising just 13 percent of the American population, they have been involved in up to 22% of serial killings from 1800 to 1995.

Confronted with this “inconvenient truth”, criminologist Scott Thornley of Mansfield University in Pennsylvania speculated that the liberal bias of American media and the fear of being labeled racist has fostered a collective inclination to sweep black criminals under the rug, as it were. “When you’re trying to appeal to a largely white audience, you don’t make Hannibal Lecter black,” said Thornley (as cited in Leonard, 2002)

William Oliver, criminal justice professor at Indiana University and author of “The Violent Social World of Black Men”, eschewed the “serial murder” label and reflected that the Muhammad case ought to be viewed through the lens of domestic terrorism. His data shows that one-third of young black men are in jail, on parole, or on probation. With many more existing on the social margins or eking out an existence outside the mainstream of economic opportunity, he argues, the real surprise is that “we haven’t had more public outbursts like this,” he said (Oliver, 1998).

With respect to the other misleading “clue” of car color, however, one wonders whether the still-inexact science of psychology did not do itself a disservice by promoting unjustified belief in the psychology of colors. Or perhaps the key difference between Metesky and Muhammad was that the former could indulge his personality traits and personality preferences, down to the color of clothes he habitually wore; in contrast, Muhammad was financially down and out.

A more important dissonance had to do with motives. Muhammad shot at 13 people in a seemingly random fashion; he injured or killed 13 males and females, children whites, blacks, Indians, and from a distance, the way students who “go postal” use rifles in campus settings. On the other hand, serial killers like Dahmer, Williams, and Juan Corona were extreme sadists who liked to get up close and personal, exercise personal control over their victims, inflict pain, force their unwelcome sexual attentions on the victims, and glory in the screams (Levin, as cited in Leonard, 2002).

Overweening pride and self-glorification seem inbred in serial killers. This is no less true in the case of Ian Brady, the UK “Moors Murderer”, child pornographer, molester, and killer of children (Rotella, Gold and Andriana, 2001). Deciding to publicize his own case study and attempting a twice-removed analysis of Henry Lee Lucas, John Wayne Gacy, Ted Bundy, and other notorious serial killers, Brady wrote the self-serving “The Gates of Janus: Serial Killing and Its Analysis”. One cannot logically expect any kind of objective analysis from Brady, of course. But the inclusion of other serial killers in his “discourse” does yield some insight that only a peer or “expert” can give the oeuvre of others.

For himself, Brady defiantly avoids apologia. Instead, the convict-writer dives into a rambling litany of the sociological pressures and psychological vulnerabilities that impelled him and others to their crimes. The book places the blame for such crimes squarely on the shoulders of “self-satisfied, privileged societies” such as the UK and America. He contends that the ruling elites smugly ignore the oppressed and teeming under-classes, which naturally breed aberrant psychopaths like himself. Any redeeming or didactic value this book might have had is ruined by veering away from thoughtful profiles of the aforementioned colleagues in crime and opting to review the gruesome nature of the crimes they committed.

The foreword actually does more to shed light on the antecedent of Brady the murderer via a lengthy biographical prologue written by Colin Wilson (Brady, 2001). Brady was reportedly your basic self-effacing and bookish boy though he did turn antisocial adolescent. Wilson ascribes to Brady a chasm of amorality and a strong streak of moral relativism, absorbed from Dostoyevsky and in particular, the superman myth that Raskolnikov holds dear in Crime and Punishment. Hence, we see that the apparently senseless and surprising eruptions of serial violence come from psyches unfettered by religion and other influences on ethics. The warped personality feels no obligation to abide by “petty and arbitrary” rules. When Brady insists that the lives and integrity of others count for nothing besides his wanton desires, whim, caprice, and savage lust, one concludes with a shudder that only political power separates him from the worst genocidal maniacs mankind has produced: Hitler, Pol Pot, Stalin, Idi Amin, and bin Laden.

For instance, Taylor (1993) reinforces the point that amorality, moral relativism, and warped ethics are common to all multiple murderers. They do share many personality traits. That they are not more quickly apprehended or easily forestalled from killing rampages is due to two factors: they hide behind personable facades and there is no predictable pattern to their violent behavior.

Taylor maintains that, far from looking like sinister ogres, these modern-day monsters are “boy next door” types. They lead perfectly respectable lives and are in fact more gracious, charming, and charismatic than the run of law-abiding citizens.

Characterizing their crimes, Taylor brings together the skeins of control, compensation, and one-upmanship. Serial killers delight in eluding law enforcers and seek to control the last moments of their victims. In so doing, the killers gain satisfaction and achieve transitory control that eludes them in most other aspects of their lives. Killing and evading capture satisfy their pronounced need to achieve. Lacking the talent and drive to back up their ambition, they are mediocre or even abject failures in school and career. Rather than try to bounce back and redouble their efforts, budding multiple murderers blame society at large for exploiting, victimizing, rejecting, and in all other ways “shutting them out”. In their warped morality, they are able to justify lashing back at whichever segment of society deserves blame.

If Dr. Elliot Leyton (as cited in Taylor, 1993), professor of anthropology and criminology at Newfoundland’s Memorial University, is anything to go by, it should actually be easier to identify and forestall serial killers because they are apt to spring from troubled backgrounds. Collated profiles skew towards those with disadvantaged (or at least a working-class) upbringing, as well as confinement in either juvenile homes or mental hospitals. Mass killers, he says, tend to come from more stable, lower-middle-class homes.

Most serial killers are alienated, sexually dysfunctional, or outright sadists who get libidinous pleasure from mutilating, abusing, and finally dispatching their victims (Ressler and Schactman, 1995). These perpetrators have deep-seated fantasies that they long to act out and no amount of therapy resolves these fantasies. Hence, Ressler argues, there is simply no rehabilitating serial killers.

In addition, serial killers have a history of drug or alcohol abuse and typically have criminal or psychiatric episodes in their family backgrounds, according to a 1988 study by the U.S. Federal Bureau of Investigation. The FBI report also says that serial killers often dwelt on unsatisfactory relationships with fathers and mixed feelings for their mothers. Raised with a distinct absence of nurturing, these budding criminals so lack socialization skills that they treat victims much like inanimate objects, theirs to toss aside or destroy at whim. After assessment sessions, counselors almost uniformly describe serial killers as perfectly antisocial: withdrawn, introspective, shy, isolated, and perennially fearful of humiliation.

Schmid (2005) objects to much that characterizes the case-study approach in popular true-crime non-fiction about serial killers. In particular, he feels that such writers overstate the case for sexual deviance and “otherness”, the serial killer as a rank outsider by virtue of some psychopathology he suffered in childhood. Authors do society and the truth a disservice, he is convinced when they portray the serial killer as a monstrous sexual predator who nevertheless “fall prey” to the wiles of misbehaving females.

On the investigative end, Egger (2002) presents case studies packed with facts about four of the more notorious serial killers in modern U.S. history: John Wayne Gacy, Ted Bundy, Kenneth Bianchi, and Henry Lee Lucas. From the layman’s point of view, the more educational points Egger concern the difficulty of investigating and apprehending suspects. Since serial killers pick victims at random, even those clues that the careless ones leave do not immediately aid investigators because there are social or kinship links to draw suspects from. “Lack of linkage” is another blind spot, as when numerous murders using guns or knives occur in a jurisdiction like Los Angeles’ Southside. Then police who are inured to numerous spates of gang violence may be blind to the work of a serial killer working in the shadows.

Method – Primary

The research question: does a scientifically-derived profile of convicted serial killers enable the prediction of suspect characteristics in new cases? I shall simulate predictive value by:

• Investigating the biographical, socio-demographic, personality factors, and ethics of a sample of convicted serial killers.
• Ascertaining which of these predominantly characterize the group and deriving a composite score.
• “Predicting” based on their composite score the serial killers in another population of prisoners or even in the general population.

The case study method shall revolve around a statistically substantive sample of 50 convicts for serial killings and a control group of those convicted for other crimes. Since most of them will be incarcerated out of state or overseas, I shall formulate a standardized data collection form that can be filled up from case files sent by the relevant authorities or by sending them the form and soliciting their cooperation in filling up the form.

The validity of this data collection and organization form shall be based on a comprehensive list of variables that are hypothesized to have some relationship with pathological and criminal behavior. These shall include:

1. Socio-demographic information: age, educational attainment,
2. Personality factors observed by caseworkers, e.g. extraversion/introversion, social adjustment, dominance/submissiveness, the tendency to hysteria, gross psychopathology.
3. Upbringing: whether in inner-city, suburban or remote, rural settings; from unitary two-parent or single-parent family; incidence of living in orphanages or juvenile homes, experience with domestic violence and religiosity.
4. Pastimes and aspirations.

In turn, reliability shall be enhanced by constructing the form as a series of nominal or categorical items.

Method – Secondary

The fallback method is content analysis. Originating from the fields of linguistics, media, and advertising research, content analysis is essentially an observation technique used to analyze written material into meaningful units using carefully constructed and applied rules. It is also referred to as textual criticism.

In the context of understanding the mind and motivations of the serial killer, applying content analysis means an objective, systematic description of the content of communication elicited from a sample of perpetrators and convicts. To start with, the study objectives can be phrased as follows:

• Identify the modes of thought, inclinations, desires, aspirations, and self-concept of the modern-day serial killer.
• Determine what degree of amorality and ethics, if any, pervade his thinking about the criminal acts he is accused of.

The material to be subjected to content analysis shall be of two types. First, I shall make use of recorded interrogations and transcripts of trial testimony. The former shall be solely interrogations of the perpetrator himself while the scope of trial testimony will be expanded to include the statements of both the perpetrator and all those giving testimony, whether for the defense or prosecution, in respect of the background and personal circumstances of the convict.

The recording form shall consist of categorical ratings for the four groups of variables described under the primary method for the case study. As the primary method, this shall use a matched control group of men incarcerated for other, non-violent crimes.

The alternative, of course, is to conduct an extensive literature review covering:

• Magazine articles
• Journal articles
• Newspaper articles
• Books were written about specific serial killers
• Psychiatric files
• Internet web sites

Such an approach could investigate patterns of association and reveal whether similarities exist.

This alternative employs the “life history methodology”; an analysis of small numbers of cases. It is useful for reconstructing the life history of an individual. The investigation employs the lens of the story of the person as expressed in his own words.

Accordingly, the key research questions become descriptive and exploratory, and the mode of reasoning for this study would be principally inductive: the investigator works from personal documents. In reconstructing the life history of serial killers, the analysis will involve the “reproductive” forms of reasoning.

Just as for the case study and first content analysis options, the structure of data analysis covers a multiplicity of themes:

• Child-rearing environment
• Any experience with trauma
• Psychological factors can be accounted for.
• Adjustment and coping skills exhibited during childhood, including a history of anti-social behavior
• Behavior during adolescence

Results and Data Analysis

Other than the standard descriptive statistics, multiple discriminant analyses have a central role and can be applied to both research methods. MDA is a procedure for predicting group membership on the basis of the four sets of independent variables. Unlike the requirement of metric dependent variables in multiple regression analysis, the dependent variable can be either nominal or categorical in MDA. Concretely, this means a simple “yes”/“no” on being a serial killer. On the other hand, the independent variables can include metric measures such as age, number of siblings, years of education attained.

The goals for MDA are as follows:

• Determine if there are statistically significant differences between the average discriminant score profiles of serial killers and the general prison population.
• Establish a model for classifying other prisoner populations (not included in the original assessment) into serial killers or not.
• Determine how much of the difference in the average “serial killer” score profile of the test and control groups is accounted for by each of the independent variables.

MDA calculates a discriminant or Z score for each of the test group serial killers and prisoners convicted of other crimes. At the end of the process, this score is the basis for “predicting” whether a convict belongs among the serial killers or not. As well, the analysis will extend to the size of the discriminant weights associated with the independent variables. Given what we have seen in the review of literature, one can reasonably expect that the larger weights (i.e., those evincing larger differences between a serial killer and ordinary prisoners or possessed of greater discriminatory power) will be such variables as age, race, introversion, having lived in orphanages or juvenile homes, a prideful bent, social maladjustment, absent or sparse romantic relationships, a weak id, control, compensation, one-upmanship, weak or absent parents, drug and alcohol abuse.

Accordingly, we test the hypothesis of being able to successfully profile serial killers with this discriminant model of selected variables:

Z= b₁X₁ + b₂X₂ + b₃X₃ + b₄X₄ + b₅X₅

where

Z = discriminant score

b_1-n = discriminant coefficients or weights

X₁ = Individual’s composite introversion score

X₂ = Composite strength-of-ethics score

X₃ = Quality of family ties

X₄ = Adolescent crisis score

X₅ = Control score

The discriminant analysis results were:

Z=.18X₁ +.04X₂ +.12X₃ +.10X₄ -.37X₅

This suggests that the lack of control in most aspects of one’s life is the most important variable followed by the degree to which the criminal is withdrawn and therefore socially maladjusted while the least important is the strength of id, a factor on which serial killers differ little from the general population of inmates.

Proceeding to the other important goal of discriminant analysis – being able to predict membership in the serial killer sub-class – one derives a confusion matrix.

This shows that the estimated discriminant function correctly classified 80.3 percent of the serial killers as such and incorrectly categorized 19.7 percent of ordinary prisoners as serial killers. Taken at face value, this model already boasts 4-in-5 odds of picking out a serial killer from a line-up of suspects with criminal records.

Proceeding further, however, there are statistical tests to indicate whether being able to predict a serial killer from available profiling tools with this model yields better results than might be obtained by pure chance. Among the many approaches, a simple option is the proportional chance criterion, applicable when group sizes are unequal and our investigatory goal is to correctly predict membership in the serial killer sub-class:

C_PRO = p² + (1 – p)²

where

p = proportion of criminals in group 1 (serial killers)

1 – p = proportion of criminals in group 2 (ordinary convicts)

In this case, the serial killer group subsumes 117 convicts or 28.3 percent of the total (117/413). Calculation of the proportional chance criterion proceeds:

C_PRO =0.283² + (1 -.283)² =0.08 +0.514 =0.594 or 59.4%

To conclude, the model correctly predicts 94 of the serial killers and 235 of the ordinary convicts for a total of 329 out of 413 or 79.7 percent correctly classified. This exceeds the proportional chance criterion of 59.4 percent by a wide margin and affirms that this discriminant model with its correctly chosen component variables did a better job than we would expect due to chance.

Methodological Benefits and Drawbacks

The case study method is the mainstay of criminal profiling, that criminology endeavor aimed at predicting the characteristics and identity of serial killers. Unhappily, the spotlight is thrown by media on such “experts” whenever a new spate of killing pops up overstates the case for the reliability of case studies (Gorelick, 2002). Television anchormen and would-be authors put too much store by statements of the profilers that they already know precisely what a killer looks like, what he must have been thinking when pulling off that last murderous episode, even what food and drink he consumed that very day.

It is easy enough to digest the validity of interviewing apprehended serial killers in the hope of identifying characteristics typical of the breed. From there, it is a giant step, however, to predict what the next serial killer must be like in every respect. The uninformed public that prefers certainty and the veneer of expertise do not appreciate the fact the probability statistics leave room for error. The informed theory of case study profiling raises the chance of identifying common characteristics but, given the diversity of the “universe” of humans and the randomness of criminal acts, no professional engaged in criminal profiling can claim to use the tool for investigative purposes, to zero in on suspects with anything like 100% certainty. At best, profiling can help narrow down the search for probable offenders.

References

Brady, I. (2001). The gates of Janus: Serial killing and its analysis. Los Angeles: Feral House.

Egger, S. A. (2002). The killers among us: An examination of serial murder and its investigation. New York: Prentice Hall.

Gorelick, S. M. (2002). Putting a face on evil. The Christian Science Monitor, p. 11.

Leonard, M. (2002). Sniper suspect defies profile. Boston Globe, A.1.

Kelly, S. (2002). This time, profiles didn’t fit. Denver Post, A. 1.

Kocsis, R. N. (2004). Profiling the criminal mind: does it actually work? The Lancet: Medicine, Crime, and Punishment, 364, s14-5.

Oliver, W. (1998) The violent social world of black men. Hoboken, NJ: /J. Wiley/Jossey-Bass.

Rotella, M., Gold, S. F. & Andriana, L. (2001). The gates of Janus: Serial killing and its analysis. Publishers Weekly, 248(47), 58.

Schmid, D. (2005). Natural-born celebrities. Serial killers in American culture. Chicago: The University of Chicago Press.

Taylor, Linda E (1993, January). Multiple murderers: What makes a monster? Canada & the World, 58(5), 16.