16/06/2026
ABSTRACT
Though stalking is often portrayed as a solo offender crime, victims can also be targeted by multiple stalkers working together or separately. This project uses data from the 2006 Stalking Victimization Supplement to explore the negative consequences and protective measures experienced by victims of solo stalkers, multiple stalkers working together, and multiple, independent stalkers. Results suggest the presence of multiple stalkers on its own does not aggravate the victim consequences, but having multiple stalkers working together increases the average number of negative outcomes experienced. These findings highlight the need for research in this area, as the presence of multiple stalkers jointly targeting a victim can heighten the negative effects of stalking.
KEYWORDS:
Stalkingvictimizationmultiple-perpetrator stalkingproxy stalking
Previous article
View issue table of contents
Next article
Introduction
Even though stalking is often thought of as a crime involving one victim and one offender, multiple perpetrator stalking has been observed in the form of proxy stalking — situations in which a primary stalker uses other individuals to further target the victim (Logan & Walker, Citation2017; Melton, Citation2007b; Spitzberg & Cadiz, Citation2002). This tactic can take several forms, perpetrators can manipulate or trick the victims’ family and friends into gaining information or access to them, they can enlist their friends or family to follow or harass the victim, or they can use unwitting service providers, for example, delivery drivers, to further terrorize them (Logan & Walker, Citation2017; Mullen et al., Citation2000). Multiple perpetrator stalking can also include situations where victims are harassed by multiple stalkers operating independently of each other, an issue yet to be explored.
Much of what is known regarding “stalking by proxy” focuses on the victim-offender relationship, and primarily draws from the experiences of female victims of domestic violence (Brewster, Citation2003; Melton, Citation2007b; Mullen et al., Citation2000). When researchers discuss proxy stalking, it is identified as a tactic used by perpetrators to extend their control over victims, and examples are extensively described, but its effects on victim outcomes have not been examined. This is problematic as multiple perpetrator stalking affects approximately 30% of stalking victims (Logan, Citation2019); and presents a challenge for victim safety planning (Logan & Walker, Citation2018). Findings from research with victim services advocates suggest that it is a significant concern when developing a victim safety plan since third parties could help circumvent protection orders and heighten the negative effects of stalking (Logan & Walker, Citation2018). Finally, the existing literature does not distinguish between victims of multiple, independent stalkers, and those of multiple stalkers working together, which could result in different victim consequences.
The current study attempts to fill these gaps by exploring how victims are affected by the presence of multiple perpetrators, and the relationships among them. Regression analyses are employed to examine if the number of negative outcomes experienced, and the number of protective measures taken by victims differ among the stalking groups. First, the experiences of victims targeted by solo stalkers and multiple stalkers are examined, and second, the experiences of victims of multiple, independent stalkers and multiple stalkers working together are compared. Additionally, the victim-offender relationship of victims in the different stalker groups are examined to determine if victims in certain relationships are at heightened risk of multiple perpetrator stalking. Though exploratory, study findings further our understanding of the victim consequences of multiple perpetrator stalking, and have the potential to inform stalking education, victim safety planning, and risk assessment, as well as provide new directions for future research.
Background
Stalking
Stalking is a form of intimate partner violence that accounts for 2.3% of all victimization reported by males and 9.7% of all victimization reported by females in the 2010—2012 National Intimate Partner and Sexual Violence Survey (Smith et al., Citation2017). According to the Model Stalking Code, a person is guilty of stalking if they engage “in a course of conduct directed at a specific person and knows or should know that the course of conduct would cause a reasonable person to (a) fear for his or her safety or the safety of a third person; or (b) suffer other emotional distress … ” (National Center for Victims of Crime [NCVC], Citation2007, p. 24). The “course of conduct” required by stalking laws can be established by any two or more incidents in which an individual directly or indirectly uses a variety of tactics to harass or surveil a victim, either in person or through a third party (NCVC, Citation2007, p. 24). Stalking by multiple perpetrators is not accounted for in stalking legislation, however, it affects a large number of stalking victims (Logan, Citation2019), highlighting the need for more research in the area.
It is well established in the literature that a majority of stalking victims are female and most perpetrators are male (Smith et al., Citation2017; Tjaden, Citation2009), most victims know their stalker (Baum et al., Citation2009;Tjaden, Citation2009; Black et al., Citation2011; Smith et al., Citation2017;), and only a small percentage of stalking incidents are perpetrated by strangers, with estimates ranging from 9% (Catalano, Citation2012) to 16.6% (Smith et al., Citation2017). In the United States, 15.8% of women and 5.3% of men reported being stalked in their lifetime, and reported feeling fearful, or feared themselves, a friend or relative could be seriously harmed (Smith et al., Citation2017).
Men are at risk of stalking victimization primarily from current or former intimate partners (42.8%), acquaintances (37.2%), strangers (16.6%), or family members (11.0%), and are commonly stalked by both male (43.0%) and female (45.7%) perpetrators (Smith et al., Citation2017). Female victims report being primarily stalked by a current or former intimate partner (61.5%), an acquaintance (26.1%), stranger (14.7%), or family member (7.2%), but are disproportionately stalked by male perpetrators (84.8%) (Smith et al., Citation2017). Thompson et al. (Citation2013) found similar rates of stalking perpetration and stalking violence among males and females, and Churcher and Nesca (Citation2013) concluded that male stalkers did not pose a greater risk of harm than female stalkers.
Proxy & multiple perpetrator stalking
Instances of multiple perpetrator stalking have been identified in the literature as proxy stalking, cases where a stalker can enlist willing friends or associates to aid in the behavior; they can coerce, bribe, or intimidate unwilling associates or even trick the victim’s friends or family into gaining access to the victim (Logan & Walker, Citation2017; Purcell et al., Citation2010). For example, some proxy stalking scenarios reported by victims include occasions where proxy stalkers took the place of the primary stalker when they were not available, aided in the stalking activities, or helped the stalker circumvent protection orders (Melton, Citation2007b). Other cases can involve the stalker engaging unwitting service providers to find or track the victim, deliver goods, or unintentionally harass the victim (Mullen et al., Citation2000, Chapter 10). Victimization by proxy stalkers could represent heightened consequences for victims, as having multiple people watching, following, or harassing them could intensify their feelings of fear and worry, and indicates their primary stalker is committed to continuing the relationship (Logan & Walker, Citation2017; Purcell et al., Citation2010).
While a majority of both male (82.2%) and female (76.0%) stalking victims report being victimized by a single stalker throughout their entire lives, 17% of female and 9.6% of male victims reported being stalked by at least two different perpetrators (Black et al., Citation2011). Logan (Citation2019) found that around 30% of stalking victims regardless of the victim-offender relationship experience proxy stalking, suggesting that multiple perpetrator stalking is a significant issue among stalking victims, yet has received little attention in the literature. Gatewood Owens (Citation2017) found no significant gender differences in the likelihood of being stalked by a single perpetrator or multiple individuals. Though a study of stalking victimization among juveniles found that female stalkers were more likely than males to engage in proxy stalking (Purcell et al., Citation2010), and Melton (Citation2007b) observed this tactic is more likely to occur in cases where the stalker is an ex intimate partner. It is important to note, however, that victims can experience stalking by multiple perpetrators working independently of each other, and the presence of multiple perpetrators does not necessarily mean these individuals are working together. These victims could have multiple stalkers working independently at the same time, could have been stalked by multiple people at different times, or have not yet realized these incidents were connected.
The courts have recognized the severity of stalking by proxy, and at least two offenders have been convicted of felony stalking in the state of Washington. In 2001, while serving a prison sentence for violating a protection order against his ex-wife, Allan Parmelee distributed flyers among inmates with his ex-wife’s description and full address, requesting explicit letters and promising money, pictures, and a place to stay in exchange (State v. Parmelee, Citation2001). Parmelee was eventually found guilty of felony stalking and three misdemeanor counts of violating a protection order — and received a 12-month sentence for the felony conviction, and a 12-month sentence for each misdemeanor (State v. Parmelee, Citation2001). Even though the appeals court remanded two of the misdemeanor convictions, his felony conviction charge and one of the misdemeanor protection order violations were upheld. In a similar yet different case in 2008, the Washington Supreme Court upheld a stalking conviction of a defendant who used friends and associates to stalk his ex-girlfriend after she obtained a protection order that prohibited contact even through third parties (State v. Becklin, Citation2008). In this case, the defendant had associates submit written reports of their sightings of the victim and had friends drive his car in front of her house on numerous occasions (State v. Becklin, Citation2008). The defendant was found guilty by a jury after they asked the judge if “[you could] stalk a person through a third party” and the judge replied with “yes.” The decision was initially overturned by the Court of Appeals on the grounds that “stalking could not be accomplished through a third person,” but it was later upheld by the Washington Supreme Court (State v. Becklin, Citation2008). The majority determined that the “course of action” described in the stalking statute applied to actions carried out by others, and concluded that “the plain language of the statute is broad enough to encompass the act of directing third parties to follow and intimidate a victim” (State v. Becklin, Citation2008).
As our knowledge about stalking, its nature, and victim consequences increases, proxy stalking has emerged as a tactic employed by some stalkers to further exert power and control over their victims (Brewster, Citation2003; Melton, Citation2007b). As explained by Wagers et al.(Citation2015), power is having the ability to affect or influence another’s actions or behavior, while control is exercising that ability. Brewster (Citation2003) concluded that stalking is an extension of the power and control that was exercised during the relationship and represents the stalkers’ attempt to reassert that power after the relationship ended. While results from Melton (Citation2007b) suggest that the inclusion of proxy stalkers or multiple perpetrators by intimate partner stalkers could be an attempt to extend that power and control.
Victim consequences
Even though the courts have shed some light on the criminal liability of proxy stalking, less is known regarding the implications for victims. Incidents with multiple perpetrators can have a heightened negative effect on victims, and findings by Salter (Citation2014) suggest that “multiple perpetrators and multiple incidents of interpersonal violence are associated with negative mental and physical health outcomes for women in comparison to solo offenses” (p. 103). Findings suggest there are different consequences for victims of interpersonal violence that were victimized by multiple perpetrators than for those who suffered revictimization by the same offender (Matlow & DePrince, Citation2013). This is in line with research that demonstrates that crimes committed by groups of co-offenders are more likely than those perpetrated by single offenders to involve weapons or result in an injury (Lantz, Citation2018).
Churcher and Nesca (Citation2013) found that approximately one-third of stalkers engage in violent behavior towards their victims, and a reported 29% of victims suffer physical harm as a result of these incidents. Stalking tactics generally meant to keep the stalker informed of the victim’s daily activities are associated with a higher risk of physical violence and homicide among stalking victims, and it is normally where proxy stalking takes place (Logan & Walker, Citation2017; NCVC, Citation2007). Victimization consequences of stalking can include emotional and psychological distress, physical injuries, financial costs, identity theft, and property damage among others (Churcher & Nesca, Citation2013; Ngo & Paternoster, Citation2016). The tactics used by stalkers push victims to employ a variety of protective measures in the hopes of ensuring their safety and regaining peace and normalcy. These can go from changing e-mails or phone numbers, taking time away from work or school, staying with friends or family, getting a weapon, or taking self-defense lessons (Baum et al., Citation2009). Recent research by Villacampa and Pujols (Citation2019) found that stalking victims take an average of 2.35 measures as a result of the incidents.
A common protective measure taken by victims is the use of protection orders to keep their stalkers away. However, these can be hard to obtain as both the courts and law enforcement have difficulty identifying stalking, and it can be difficult for victims to collect enough evidence to prove stalking (Logan & Walker, Citation2018). Furthermore, protection orders meant to limit contact between the abuser and the victim do not protect against harassment by third parties on behalf of the stalker. So even when one is secured by the victim, proxy stalking can be a tool for abusers to circumvent these protection orders, making it nearly impossible for victims to escape their stalker despite their legal attempts (Melton, Citation2007b; State v. Becklin, Citation2008; State v. Parmelee, Citation2001). The ability of a third party to harass or keep tabs on victims can create further uncertainty and eliminate the sense of safety provided by the protection order, making multiple-perpetrator stalking a significant concern for victim safety planning (Logan & Walker, Citation2018).
Underreporting and police enforcement
Some factors contributing to the low number of arrests and prosecutions of stalking include widespread underreporting by victims, subscription to stalking myths that romanticize and trivialized stalking behaviors, inadequate law enforcement training in identifying and investigating stalking, as well as difficulties obtaining enough evidence to meet the legal requirements of stalking. Unlike other crimes clearly defined and easily identified by the law, stalking needs to be acknowledged by the victim and reported to the police for the criminal justice process to start (Campbell & Moore, Citation2011). Underreporting of crime is an issue in general; however, it is especially problematic in stalking as police involvement by victims seems to be an effective way of interrupting stalking behaviors (Baum et al., Citation2009; Campbell & Moore, Citation2011; Melton, Citation2007a). Using surveys, Baum et al. (Citation2009) found that 16% of victims attributed the stalking cessation to police warning the perpetrator. Similarly, interview results from Melton (Citation2007a) suggest that criminal justice intervention is significantly linked to decreases in physical violence and stalking behaviors among victims of domestic violence. Failure to report stalking incidents to law enforcement puts victims at risk of continued victimization, violence escalation, and deprives them of support services that could help mitigate some of the harmful consequences of stalking (Fohring, Citation2014).
The first step in deciding to report any criminal incident to the police requires victims to recognize that a crime has been committed. Insufficient knowledge and understanding of stalking could hinder a victim from recognizing the contacts they have been experiencing as a crime. Some victims may not deem the incidents serious enough to report to the police. In contrast, others may hesitate to label their experiences as criminal victimizations to avoid the stigma and blame that comes with being recognized as a victim of crime (Fohring, Citation2014). Stalking research suggests that victimization acknowledgment is positively correlated to the perceived seriousness of the incidents and the fear experienced by the victim (Gatewood Owens, Citation2017; Ménard & Cox, Citation2016). In turn, labeling the event as stalking has been found to influence victims’ formal and informal help-seeking behaviors — such as contacting law enforcement or reaching out to friends or family (Reyns & Englebrecht, Citation2010, Citation2014).
Even when victims recognize the seriousness of stalking and report the incidents to the police, the likelihood of an arrest or formal legal action is very unlikely. One of the many hurdles in the implementation of stalking laws is the inability of many officers to recognize and investigate stalking incidents. Brady and Nobles (Citation2017) found that of 3,756 stalking calls for service made to the Houston Police Department over eight years, there were only 66 stalking-related incident reports filed, and only 12 stalking-related arrests were made. Lynch and Logan (Citation2015) examined police officers’ perceptions and opinions regarding charging and investigating stalking cases and found that many officers (37%) were generally unfamiliar with stalking laws or considered their implementation challenging. Police officers’ lack of experience with stalking cases and the misunderstood nature of the crime can result in stalking reports being dismissed or trivialized. Even though isolated stalking incidents could be perceived as harmless or even romantic (McKeon et al., Citation2015), these repeated behaviors over time can have significant consequences for victims and need to be assessed within the context in which they occur and not as isolated events (Logan & Walker, Citation2017). Korkodeilou (Citation2015) found that stalking victims were dissatisfied with the help and advice offered by law enforcement, and their reports were met with disbelief, downplayed, and dismissed when victims contacted the police. Victims reported that officers placed great emphasis on physical harm rather than the more common psychological and emotional forms of abuse found in stalking (Korkodeilou, Citation2015). These negative experiences with law enforcement could negatively affect the likelihood of victims contacting the police to report similar incidents in the future (Ngo, Citation2019), yet, contact and collaboration between victims and police officers are vital to collecting evidence and building a stalking case.
Law enforcement challenges identifying stalking could be even more problematic for victims of proxy or multiple perpetrator stalking. Victim reports of multiple stalkers could be received with even more disbelief and inaction by police, as victims may be deemed paranoid and unreliable. Further, results from Ngo (Citation2020) suggest that police intervention by talking to or warning the perpetrator is effective at disrupting the stalking situation, highlighting the need for police officers to understand multiple perpetrator stalking and be able to adequately help these victims. As a result, it is necessary to explore the nature and risks of multiple perpetrators stalking to educate victims, law enforcement, and victim service providers of its prevalence and victim outcomes, otherwise, proxy stalking will continue to be a tactic employed by stalkers to circumvent protection orders.
Current study
The current project aims to expand the stalking literature by providing a better understanding of the nature, risks, and consequences of multiple perpetrator stalking. Although several studies have identified instances of proxy stalking, no work employing a nationally representative sample has been conducted to further our understanding of this issue. The majority of studies that touch on proxy stalking use small qualitative samples of female domestic violence victims (Brewster, Citation2003; Melton, Citation2007a, Citation2007b), juvenile stalkers (Purcell et al., Citation2010), and only recently (Logan, Citation2019) has a study using a large sample of both male and female adults looked at proxy stalking. None of these works examine the presence of multiple perpetrators as the primary variable of interest; rather, it is noted as a tactic employed by stalkers. Even though a significant number of stalking victims are affected by proxy stalking (Black et al., Citation2011; Logan, Citation2019), the experiences and outcomes of this type of victimization remain unexplored. This is worrisome as proxy stalking can be used to circumvent protection orders (Logan, Citation2019; State v. Becklin, Citation2008; State v. Parmelee, Citation2001) and might put victims at higher risk of physical violence and other negative outcomes (Lantz, Citation2018; Salter, Citation2014).
The current project is not only interested in the differences among victims of solo stalkers and multiple stalkers, but also the different experiences of victims of multiple, independent stalkers and multiple stalkers working together. It is important to distinguish if collaboration among multiple perpetrators, not merely their presence, might result in more severe consequences for victims. To this end, the current project has three main objectives: first, to examine whether having multiple stalkers increases the number of protective measures taken by victims or the number of negative outcomes experienced. Second, to explore whether having multiple independent stalkers or multiple stalkers working together increases the number of negative outcomes and protective measures taken. Finally, to examine if certain victim-offender relationships are more likely than others to result in having multiple stalkers, or multiple stalkers working together.
Three different regression models are employed to examine these relationships using two different samples. The first sample includes all cases and is used to compare victims of solo stalkers and multiple stalkers. While the second sample only consists of a subsample of all cases, those with multiple stalkers, and is used to examine the experiences of victims of multiple, independent stalkers and victims of multiple stalkers working together. The first question of interest is whether the presence of multiple perpetrators or collaboration among stalkers affects the average number of negative outcomes experienced by victims. The second research question examines how the average number of self-protective measures taken by victims is affected by the presence of multiple stalkers or multiple stalkers working together. According to Salter (Citation2014), victimization by multiple perpetrators is associated with negative mental and physical health outcomes for women, and co-offending research suggests that crimes committed by groups of co-offenders are more likely than those perpetrated by single offenders to involve weapons or result in an injury (Lantz, Citation2018). As a result, it is hypothesized that victims of multiple perpetrators, on average, will experience more negative outcomes and engage in more protective measures than victims who reported only one stalker. Additionally, it is expected that victims of multiple stalkers working together on average will experience more negative outcomes and engage in more protective measures than those stalked by multiple, independent stalkers. Finally, although proxy stalking or multiple perpetrator stalking has been more commonly observed in cases of ex-partner stalking (Brewster, Citation2003; Melton, Citation2007b), Logan (Citation2019) found that around a third of stalking victims regardless of the victim-offender relationship report proxy stalkers. The current study seeks to examine this relationship with a nationally representative sample, and it is expected that victims will be more likely to experience multiple perpetrator stalking and have multiple stalkers working together when their stalker is an intimate or ex-intimate partner.
Method
Sample
Data for the project are drawn from the 2006 Stalking Victimization Supplement (SVS), of the National Crime Victimization Survey (NCVS). The SVS was administered to NCVS respondents who were at least 18-years old and met a set of screening questions that determined that they had experienced stalking. A total of 65,272 people completed the SVS, 82.9% of all eligible NCVS respondents. One hundred and seventy cases were flagged as spam in the SVS and were removed from the sample. Additionally, cases in which the victims could not identify the number of perpetrators involved (n = 45) or those with missing answers (n = 5) were removed. The sample for this study consists of 898 stalking victims who reported feeling frightened, concerned, angered, or annoyed as the result of experiencing at least one of seven stalking or harassing behaviors on more than one occasion during the 12 months before the survey (Catalano, Citation2012). Stalking or harassing behaviors included receiving unsolicited or unwanted phone calls, letters, e-mails, or other forms of written communication; following or spying on the victim; waiting outside or inside places for the victim; showing up at places where they were even though they had no business being there; leaving unwanted items, presents, or flowers, and or posting information or spreading rumors about them on the internet, in public, or by word of mouth (Catalano, Citation2012).
Two samples are used in this study. The first sample includes all stalking victims (n = 898) and compares victims of solo stalkers (n = 607) and multiple stalkers (n = 291). The second sample only includes cases with multiple stalkers and compares victims who reported multiple, independent stalkers (n = 159) and victims of multiple stalkers working together (n = 111). Of the 291 cases with victims who reported multiple stalkers, 21 could not determine if their stalkers were working together or independently and were removed from the second sample, for a total of 270 cases.
Measures
Dependent variables
Three dependent variables were used in the analyses: perpetrator type, number of negative outcomes experienced by victims, and number of protective measures taken by victims. The number of perpetrators and the relationship among them was measured using two questions in the survey. Respondents were asked how many people engaged in the stalking behaviors reported, and those who reported multiple stalkers were asked if they believed these people were working together or independently. In this study, the perpetrator type is examined using two dichotomous variables: stalker type and working together. Stalker type was coded as 1 (multiple stalkers) and 0 (solo stalker), while working together was coded as 1 (multiple stalkers working together) and 0 (multiple, independent stalkers). It is important to note that using these data it was not possible to determine if victims who reported having multiple perpetrators were victimized by these individuals during the same time period, or at different times during the 12 months preceding the interview. As presented in Table 1, almost 70% of victims were stalked by only one person (n = 607), while a third of the sample reported being stalked by multiple individuals (n = 291). Over half (55%) of all victims stalked by multiple perpetrators believed the stalkers were working alone (n = 159), 38% believed they were stalked by multiple perpetrators working together (n = 111), and 7% (n = 21) could not identify if their stalkers were working together or independently.
Table 1. Frequencies for all variables by stalker group
Download CSVDisplay Table
Negative outcomes is a count variable with possible values ranging from 0 to 6. It was computed by combining six different stalking outcomes victims could have experienced — emotional or psychological distress, an attack or attempted attack, other crimes, cyberstalking, identifying incidents as stalking, and reporting incidents to the police. Victims could experience none, some, or all of the negative outcomes, but on average, they experienced 2.5 negative outcomes across the sample (SD = 1.4). Most victims experienced one (20%), two (25%), or three (22%) negative outcomes — and only 2% of all victims reported experiencing all six outcomes. Frequencies for all variables are reported in Table 1. Although not all the outcomes included as part of the variable represent a direct consequence of stalking like experiencing an attack, or other crimes, they are all being used as indicators of how intense the stalking incident was. For example, research suggests that self-labeling of the incident as stalking is positively correlated to the perceived seriousness of the incidents and the fear experienced by the victim (Gatewood Owens, Citation2017; Ménard & Cox, Citation2016), and victims were more likely to report the incident to law enforcement when they experienced threats, felt intimidated, and the perpetrator had a prior criminal record (Reyns & Englebrecht, Citation2010).
Protective measures taken by victims is a count variable with values ranging from 0 to 16. It was measured using sixteen questions that asked victims about protective measures taken as a result of the incidents. Protective measures included taking time off from work or school, changing or quitting a job or school, changing their commute to work or school, avoiding relatives, friends, or holiday celebrations, changing usual activities outside of work or school, staying with friends or relatives or vice versa, altering their appearance, taking self-defense or martial arts classes, getting pepper spray, getting a gun, getting any other kind of weapon, changing their social security number, changing their e-mail address or telephone numbers, installing caller id or call blocking systems, or changing or install new locks or a security system.
Although 16 different protective measures could be taken, victims in the sample only took between one and 13. Almost a third of all victims took some sort of protective measure (62%), 38% did not take any protective measures, and on average victims across the sample took 1.7 protective measures (SD = 2.1), see Table 1. Most of those who took action were likely to employ one (25%) or two (12%) protective measures and only 2% of all victims took between eight to 13 different actions.
Independent variables
In addition to the offender groups described above, the victim-offender relationship is the second predictor of interest. This variable was measured using the survey question “what was the relationship of the person who did (this/these things) to you when the contacts began” and answers were used to create four dummy variables coded as 1 (yes) and 0 (no). Known was the most common victim-offender relationship across the sample (n = 419, 47%), and it includes victims who were stalked by friends, housemates, acquaintances, schoolmates, neighbors, customers, students, or coworkers. Intimate stalkers made up the second-largest group (n = 252, 28%), followed by unknown stalkers (n = 110, 12%), which includes both strangers and individuals the victim could not identify, and finally, relatives (9%).
Control variables
Several demographic variables were used in the analyses, including the victim’s age, race, gender, and marital status, as well as the offender’s criminal history. Age is measured as a continuous variable, with victim ages ranging from 18 to 86 years old. The average age across the sample was 38 (SD = 13.8), and victims stalked by multiple stalkers working together were on average older than others in the sample (M = 40, SD = 14.2). Gender was measured using a dichotomous variable coded as 1 (females) and 0 (males). As observed in Table 1, a majority of victims in the sample were female (n = 684, 76%), and around a third were male (n = 214, 27%). Race was computed as a dichotomy, and victims were classified as 1 (white) and 0 (nonwhite). Victims in the sample were overwhelmingly white (n = 763, 85%), and 15% of all victims were nonwhite. Marital status was measured as a dichotomy and coded as 1 (yes) and 0 (no) — and around a third of victims reported being married (32%). Finally, offender criminal history was measured using a question that asked victims if to their knowledge the stalker had a criminal past. It was coded as 1 (yes) and 0 (no), and less than half of all victims (n = 350, 39%) reported their stalker had a criminal history. Descriptive statistics, means, and standard deviations for all variables can be found in Table 2.
Table 2. Descriptive statistics of variables in the analyses
Download CSVDisplay Table
Analytical approach
Data for the current study were examined using ANOVAs, chi-squares, and three different regression strategies. ANOVAs and chi-squares were used to explore the bivariate differences of variables across all stalker groups, while Poisson regression, negative binomial regression, and binary logistic regression were used to estimate models. All three regression models in the sample were estimated twice — first, using the full study sample (n = 898) and comparing victims of solo stalkers to victims of multiple stalkers, and second, using a sample of victims of multiple perpetrator stalking who could distinguish if the people involved were working together or independently (n = 270). Significance levels in this project are reported at p < .05.
First, Poisson regression was selected to analyze the first research question — whether the average number of negative outcomes experienced by victims was affected by the stalker type. This is an appropriate method of analysis for the model as it consists of a count dependent variable and both continuous and dichotomous independent variables. Additionally, the dependent variable had a Poisson distribution, and the data displayed equidispersion, making it a good fit for Poisson regression. Second, negative binomial regression was used to examine whether the average number of protective measures taken by victims was affected by the stalker type. This technique was selected because it is one of the few methods that allow the use of a count dependent variable displaying overdispersion, which was the case with protective measures in this model. Finally, binary logistic regression was used to test if victims of intimate partner stalkers were at higher risk of having multiple perpetrators than victims who were stalked by a relative, another known person, or someone unknown. This method of analysis was selected because the model uses a dichotomous dependent variable and both categorical and continuous independent variables. Binary logistic regression is one of the few methods that can accommodate a dichotomous dependent variable like stalker type. Multicollinearity was evaluated in the models and was not found to be a problem for the analyses.
Missing value analysis was conducted on all variables used in the analyses. Three variables had missing values of more than one percent: negative outcomes (5%), victim-offender relationship (4%), criminal history (4%). Little’s Missing Completely at Random test was conducted, and results suggest the data are not missing at random. Missing value indicator variables were computed for negative outcomes, victim-offender relationship, and criminal history, with missing values coded as 0 and all other values coded as 1. These variables were used to conduct t-tests and chi-squares to determine if there were any significant differences among the cases that were missing and those that were not. Missing values in negative outcome had a significant relationship to age (p = .04), missing values in the victim-offender relationship were significantly related to victim’s gender (p = .01) and stalker type (p
Though stalking is often portrayed as a solo offender crime, victims can also be targeted by multiple stalkers working together or separately. This project uses data from the 2006 Stalking Victimiz...
