Upload
ravibunga4489
View
215
Download
0
Embed Size (px)
Citation preview
7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
1/24
http://jiv.sagepub.com/Violence
Journal of Interpersonal
http://jiv.sagepub.com/content/22/1/3The online version of this article can be found at:
DOI: 10.1177/0886260506294994
2007 22: 3J Interpers ViolenceHeather C. Melton
by Domestic ViolencePredicting the Occurrence of Stalking in Relationships Characterized
Published by:
http://www.sagepublications.com
On behalf of:
American Professional Society on the Abuse of Children
can be found at:Journal of Interpersonal ViolenceAdditional services and information for
http://jiv.sagepub.com/cgi/alertsEmail Alerts:
http://jiv.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://jiv.sagepub.com/content/22/1/3.refs.htmlCitations:
What is This?
- Dec 6, 2006Version of Record>>
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/content/22/1/3http://jiv.sagepub.com/content/22/1/3http://www.sagepublications.com/http://www.apsac.org/http://www.apsac.org/http://jiv.sagepub.com/cgi/alertshttp://jiv.sagepub.com/cgi/alertshttp://jiv.sagepub.com/subscriptionshttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsPermissions.navhttp://jiv.sagepub.com/content/22/1/3.refs.htmlhttp://jiv.sagepub.com/content/22/1/3.refs.htmlhttp://jiv.sagepub.com/content/22/1/3.refs.htmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://jiv.sagepub.com/content/22/1/3.full.pdfhttp://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://jiv.sagepub.com/content/22/1/3.full.pdfhttp://jiv.sagepub.com/content/22/1/3.refs.htmlhttp://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsReprints.navhttp://jiv.sagepub.com/subscriptionshttp://jiv.sagepub.com/cgi/alertshttp://www.apsac.org/http://www.sagepublications.com/http://jiv.sagepub.com/content/22/1/3http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
2/24
Predicting the Occurrenceof Stalking in RelationshipsCharacterized byDomestic ViolenceHeather C. MeltonUniversity of Utah
A high correlation has been found between domestic violence and stalking.
However, very few studies have examined what factors predict the occur-
rence of stalking in relationships characterized by domestic violence. Using
in-depth interviews with victims of domestic violence whose cases have gone
through the criminal justice system, this article explores this issue. It was
found that experiences of stalking by their abusers were very prevalent in this
group of domestic violence victims. In terms of predicting stalking, domes-
tic violence victims who were not in a relationship with their abuser, whoseabusers had an alcohol or drug problem, who experienced more controlling
behaviors by their abusers, and who had experienced prior stalking by their
abusers were at the greatest risk of experiencing more severe stalking.
Implications for intervention are discussed.
Keywords: stalking; domestic violence; victims of stalking
Researchers, law enforcement, and the press have paid increasing atten-
tion to the problem of stalking throughout the late 20th century and
early 21st century. Although this is partly because of a number of high-
profile cases involving celebrities, we now know that the majority of stalking
incidents involve individuals who are or were intimates or acquainted, and
a high correlation exists between stalking and domestic violence (e.g., see
Baldry, 2002; Coleman, 1997; Davis, Ace, & Andra, 2000; Logan, Leukefeld,
& Walker, 2000; McFarlane et al., 1999; Mechanic, Uhlmansiek, Weaver,
& Resick, 2000; Mechanic, Weaver, & Resick, 2000; National Institute of
Justice [NIJ], 1997; Roberts & Dziegielewski, 1996; Tjaden, 1997; Tjaden& Thoennes, 2000; U.S. Department of Justice [USDOJ], 1998; White,
Kowalski, Lyndon, & Valentine, 2000). Given this relationship, identifying
the factors that predict whether or not stalking will occur within a certain
Journal of Interpersonal
Violence
Volume 22 Number 1
January 2007 3-25
2007 Sage Publications
10.1177/0886260506294994http://jiv.sagepub.com
hosted at
http://online.sagepub.com
3
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
3/24
domestic violence relationship is particularly important. This article explores
this issue.
The term stalking is used to describe the willful, repeated, and maliciousfollowing, harassing, or threatening of another person. Estimates indicate that
stalking may affect anywhere between 200,000 and 1.4 million people annu-
ally (Roberts & Dziegielewski, 1996; Tjaden, 1997; Tjaden & Thoennes,
1998; USDOJ, 1998). Victims are most often the current or former spouses
or intimate partners of their stalkers (Baldry, 2002; Burgess et al., 1997;
Coleman, 1997; Guy, 1993; NIJ, 1996; Roberts & Dziegielewski, 1996;
Tjaden & Thoennes, 1998; USDOJ, 1998). Some estimate that as many as
80% of all stalking cases involve people who were or are intimately involved(Coleman, 1997; Roberts & Dziegielewski, 1996). Previous research has
demonstrated the link between domestic violence and stalking (see Coleman,
1997; Davis et al., 2000; Logan et al., 2000; McFarlane et al., 1999; Mechanic,
Uhlmansiek, et al., 2000; Mechanic, Weaver, et al., 2000; NIJ, 1997; Tjaden
& Thoennes, 2000; White et al., 2000). For example, Tjaden and Thoennes
(1998) reported that 80% of the victims of stalking in their study reported
having been physically assaulted by the partner that later stalked them. It is
estimated that between 29% and 54% of all female murder victims are bat-tered women, and in 90% of these cases, stalking preceded the murder (Guy,
1993; USDOJ, 1998). This has led many to conclude that stalking in intimate
relationships is a form of intimate partner violence.
Very few studies have examined what factors predict stalking within
domestic violence contexts. Given this relationship between the two prob-
lems, predicting which relationships characterized by violence may also
involve stalking becomes important. In one of the few studies that attempts
to predict stalking, Mechanic, Weaver, et al. (2000) found that stalking was
more correlated with psychological abuse than physical abusethe emo-
tional abuse variables in the study better predicted stalking than the physi-
cal abuse variables. Moreover, in another study, Mechanic, Uhlmansiek,
et al. (2000) found that severe stalking in domestic violence relationships is
highly correlated with more severe physical, sexual, and emotional abuse
the more severe were those types of abuses, the more severe was the stalk-
ing. Brewster (2002) found that the best predictor of violence during
stalking incidents was evidence of prior verbal threats.
In addition, research has explored the significance of the duration of therelationship between the parties. Mechanic and her colleagues found that
stalking in a relationship escalated among women who left their partners.
Logan and colleagues (2000) also concluded that stalking is a continuation
of intimate partner violence after the relationship has ended. Thus, prior
4 Journal of Interpersonal Violence
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
4/24
research has identified the type of abuse experienced, the severity of abuse
experienced, and the time in the relationship as significant predictors of
stalking in domestic violence relationships.This article explores the issue of stalking in the context of domestic vio-
lence. The specific hypotheses and research questions examined are as fol-
lows. First, it is expected that the experience of stalking will be prevalent
among the population of domestic violence victims being studied. Prior
research that has identified this correlation (Coleman, 1997; Davis et al.,
2000; Logan et al., 2000; McFarlane et al., 1999; Mechanic, Uhlmansiek,
et al., 2000; Mechanic, Weaver, et al., 2000; NIJ, 1997; Tjaden & Thoennes,
2000; White et al., 2000). Second, given prior research on predicting stalk-ing, it is expected that more severe stalking will occur in relationships with
more severe domestic violence (Mechanic, Uhlmansiek, et al., 2000). Last,
it is expected that the victim-offender relationship is linked to the victims
stalking experiences. This research attempts to broaden our knowledge of
stalking in the context of domestic violence by understanding predictors of
stalking to determine the appropriate response to this problem.
Method
The study was funded by an NIJ grant (No. 98-WT-VX-0024) titled
A Longitudinal Study of Battered Women in the System: The Victims and
Decision Makers Perceptions. The data for the study were collected
through intensive and longitudinal interviews with female victims of domestic
violence whose cases had at least entered the criminal justice system (i.e.,
the abuser was arrested) in one of three jurisdictions in the United States: a
medium-size Midwestern city, a large Western metropolitan area, and a
Western, rural college county. The respondents were recruited from the
District Attorneys (DA) offices in these three jurisdictions. After their case
closed, they were mailed a flyer briefly describing the research with a
phone number to call for more information as well as a stamped return
postcard. Cases were closed for a number of reasons. The case may have
been closed because the charges were dismissed, the defendant may have
reached a plea bargain, or the case may have gone to trial. Only women
who contacted the project by either calling or returning the card wereincluded in the study. Because of the fact that personnel from the DAs and
prosecutors offices controlled the mailing of the flyers, the overall
response rate is unknown (i.e., they did not keep track of how many flyers
they handed or sent out).
Melton / Predicting Stalking in DV Relationships 5
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
5/24
Women were interviewed right after their case closed, 6 months later,
and 1 year later. The retrospective time frames covered in each interview
are as follows: at Interview 1, the 6 months before the arrest; at Interview 2,
the 6 months between the case closing and the second interview; and atInterview 3, the 6 months between the second and third interview. They were
paid $40 for the first interview and $50 for each of the two subsequent inter-
views. The final sample included 178 women at Time 1 (right after the case
closed), 160 at Time 2 (6 months later), and 148 at Time 3 (1 year later).
Measures
Dependent Variable
Stalking was measured in a number of ways and included a wide variety
of behaviors. Stalking was operationalized for this study using the Stalking
Behavior Checklist adapted from Coleman (1997; also see Table 1).
The respondents chose from the following responses: 0 (never), 1 (once),
2 (rarely), 3 (sometimes), and 4 (often). A composite score for stalking for
each respondent was created by adding the responses to each individual
stalking variable. Table 2 provides descriptive information for the items
comprising the composite. This measures the level of severity of stalkingexperienced. The composite score (i.e. for each respondent) ranged from
0 to 40 at Time 1 and Time 2, and 0 to 36 at Time 3. The composite range
does not meet its maximum high because no one respondent experienced all
eleven behaviors often. It went down by Time 2 and Time 3.
6 Journal of Interpersonal Violence
Table 1
Measures of Stalking
How often in the 6 months before the assailant was arrested has he:
Came unwanted to your home, work or school?
Broken or destroyed something important to you?
Broke into or attempted to break into your home or car?
Stole or read your mail?
Made unwanted calls to your home or work?
Left you unwanted phone or pager messages at home or work?
Checked up on you?
Followed or watched you?
Sent you unwanted gifts, photos, or letters? Threatened, harmed, or attempted to harm your new partner (if you have one)?
Source: Coleman (1997).
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
6/24
7
Table2
StalkingTypesa
ndFrequenciesReportedatTime1,Time2,an
dTime3
Time1(N=
178)
Time2(N=
160)
Time3(N=
148)
Variables
%a
Meanb
%a
Meanb
tTest
%a
Meanb
tTest
Checkedupon
76.4
13
6
2.52
51.9
83
1.55
6.32**
*
45.9
68
1.40
1.36
Brokesomething
important
65.7
11
7
1.79
24.4
39
0.57
8.64**
*
25.7
38
0.61
0.30
Unwantedcalls
60.7
10
8
1.95
43.1
69
1.22
4.65**
*
37.8
56
1.17
0.47
Followed/watche
d
59.0
10
5
1.73
36.3
58
0.99
5.23**
*
29.7
44
0.90
0.71
Cameunwanted
57.9
10
3
1.75
31.9
51
0.67
6.94**
*
30.4
45
0.85
1.31
Unwantedmessages
52.8
9
4
1.68
28.1
45
0.86
5.36**
*
33.8
50
0.95
0.87
Stole/readmail
43.3
7
7
1.27
15.0
24
0.38
6.62**
*
18.2
27
0.46
0.75
Broke/attempted
tobreak
35.4
6
3
0.82
14.4
23
0.25
5.33**
*
12.2
18
0.32
0.91
intohome/car
Sentunwantedgifts,photos,
27.5
4
9
0.73
16.3
26
0.39
3.09**
13.5
20
0.38
0.18
letters
Threatened/harmednewpartnerc
14.0
2
5
0.44
15.6
25
0.47
0.28
16.9
25
0.47
0.09
Compositeofstalkingd
1
5.17
7.76
8.53***
7.55
0.51
Experiencedany
stalking
92.1
17
8
0.93
56.3
90
0.57
8.84**
*
58.1
86
0.59
0.39
Compositeofphysicalviolence
1
8.45
4.27
10.61***
4.45
0.15
Experiencedany
violence
96.6
17
2
0.97
38.1
61
0.38
14.68**
*
34.9
51
0.35
0.67
a.Thisrepresentsthepercentageoftherespond
entswhoexperiencedanyofth
estalkingbehaviorsatleastonceinthetimeperiodpresented.
b.Therespondentswereaskedhowfrequentlyth
eyexperiencedeachtypeofsta
lkingbehavior.Themeanspresentedhereareforonlythe148
respon-
dentswhoparticipatedineachinterview.Thescalewasnever(0),once(1),rarely(2),sometimes(3),andoften(4).
c.Ifonlywomen
whothisisapplicableto(e.g.
theyhaveanewpartner)aree
xaminedthepercentagesare71.4%atTime1;48.1%atTime2;and
43.1%atTime3
.Moreover,thereisasignificantdifferencebetweentheme
ansofTime1andthemeans
ofTime2itdecreasessignificantly.
However,thereisnotasignificantdifferencebe
tweenTime2andTime3.
d.Thecomposite
wasdevelopedbyaddinguptheresponsestoeachofthestalkingvariables.Itrangedfrom0to40attime1;0to40attime2;and
0to36attime3.
*p
.05.**p.01.***p.001.
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
7/24
The analysis also compares the relationship between physical violence
and stalking. Thus, the study measures whether the subject experienced
any stalking. Given that stalking is a repeated behavior, any respondent
who experienced one of the above stalking behaviors sometimes or more in
the given time period was coded as having experienced stalking.
Independent Variables
A similar process was used for measuring control and physical violence.
First, experiences with control were measured. Control was defined as any
actions taken by one that is an attempt to have control over another, whether
it is financially and/or emotionally. These measures were taken from the
Index of Psychological Abuse (Sullivan, Parisian, & Davidson, 1991) and
the Violence Assessment and Index of Controlling Behaviors (Dobash &
Dobash, 1998). Control was measured by the following variables at the first
interview, the 6 month interview, and the year interview (see Table 3).
The respondents could answer 0 (never), 1 (once), 2 (rarely), 3 (some-
times), or 4 (often) to each of these eight items. Once again, a composite
scale was created from the responses to each of these controlling vari-
ables. It ranged from 0 to 32 at Time 1, 0 to 25 at Time 2, and 0 to 32 at
Time 3. As with stalking, this composite scale of control was very reliable.
The alpha at Time 1 was 0.81, at Time 2, 0.78, and at Time 3, 0.84. Thismeasures severity of control experienced by the respondents at each of the
three time periods.
A composite measure of physical violence was created by adding the
scores that respondents provided for an extended Conflict Tactic Scale
8 Journal of Interpersonal Violence
Table 3
Measures of Control
How often in the 6 months before he was arrested (or since the end of the court case or the
6 months since the end of the court case depending on which interview it is) did assailant do
any of the following to annoy or hurt you:
Tried to control your activities?
Tried to control you money?
Discouraged your contact with family or friends?
Left you somewhere with no way to get home?
Threatened to end the relationship if you didnt do what he wanted?
Tried to force you to leave your home?
Acted like he owned you? Forbid you from leaving your home?
Source: Sullivan et al. (1991), Dobash & Dobash (1998).
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
8/24
(Straus, 1979). The scale is comprised of the 23 items shown in Table 4.
The respondents answered never (0), once (1), once a month or less (2),
2 or 3 times a month (3) once or twice a week(4), 3 or 4 times a week(5),
5 or 6 times a week(6) or every day (7). For physical violence, a compos-
ite was created. The scores on this measure ranged from 0 to 109 at Time
1, from 0 to 47 at Time 2, and from 0 to 70 at Time 3. This composite rep-
resents the severity of physical violence at each of the time periods.
This scale was highly reliable, with alphas of 0.92 at Time 1, 0.92 at
Time 2, and 0.94 at Time 3. This composite represents severity of physical
violence at each of the time periods.The respondents and the abusers relationship status was measured to
determine the relationship between victim-abuser relationship and stalking.
At several points in each interview, the respondents were asked what their
relationship was with the abuser (see Table 5).
Melton / Predicting Stalking in DV Relationships 9
Table 4
Measures of Physical Violence
How often did assailant:
Break your glasses or tear your clothing?
Push or shove you?
Grab you?
Slap you with an open hand?
Pull your hair?
Bite you?
Hit you with a fist?
Kick you?
Throw something at you? Hit you with an object (aside from throwing something)?
Try to hit you with an object?
Twist your arm or leg?
Drove recklessly to scare or hurt you?
Choke you or try to smother you (including drowning)>
Burn you (i.e., cigarette burns)?
Tie you up or physically restrain you somehow?
Beat you up?
Force sexual activity?
Threaten you with a knife? Threaten you with a gun?
Stab you?
Shoot you?
Commit other violence?
Source: Straus (1979).
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
9/24
Respondents were coded as either together (1) or separated (0).
Respondents reporting that they are married, living together; girl/boyfriend,
living together; girl/boyfriend, not living together; dating but not girl/boyfriend;
or another relationship that implies they are intimately involved, were coded
as together (1). Respondents reporting that they are married and sepa-
rated (not living together), divorced (not living together), ex-girlfriend/
ex-boyfriend, or in another relationship that implies they are not intimatelyinvolved were considered separated (0). They were asked this for the
6 months before the arrest, the 6 months after the court case, and the 6 months
since the 6 month interview. Another variable was also created for whether
or not they lived together at the different time periods0 = did not live
together and 1 = lived together.
Additional variables that were measured and included in the analyses
were victim and abuser race, income, age, education, site, and whether the
abuser had an alcohol or drug problem. Victim and abuser race were
dummy variables, with White as 0 and non-White as 1. Victim income was
measured by the dollar amount of her monthly income. Age was measured
by her age in years at the time of the first interview. For the univariate and
bivariate analyses, education consisted of a number of different categories
based on her educational attainment: less than high school, high school
graduate or some college, and college graduate and beyond. For the multi-
variate analyses, a dummy variable of education was created comparing
respondents who graduated high school or attended some college and respon-
dents who graduated college or beyond, with respondents who had less thana high school education. Site was measured for the univariate and bivariate
analyses by 1 for respondents from the midsized city, 2 for respondents
from the rural college county, and 3 for respondents from the large, metro-
politan city. For the multivariate analyses, a dummy variable was created of
10 Journal of Interpersonal Violence
Table 5
Measures of Type of Relationship
How would you describe your relationship with [abusers name]? Would you say that you are:
Married, living together
Married, separated (not living together)
Divorced (not living together)
Girl/boyfriend, living together
Girl/boyfriend, not living together
Ex-girlfriend/ex-boyfriend
Dating but not girl/boyfriend
Other
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
10/24
site comparing the midsized city and the rural college county to the omitted
group of respondents from the large, metropolitan city. Finally, whether the
abuser had an alcohol or drug problem was measured by 1 (yes) and 0 (no).
Limitations
The study has several limitations. First, it is not known whether the case
from which the respondents were recruited was the victims 1st, 2nd, or
10th interaction with the criminal justice system. Thus, the victims experi-
ences prior to the 6 months before this arrest are not accounted for. Second,
this study only uses the victims perception of the events. Third, this studyonly explores victims of domestic violence and stalking who have had con-
tact with the police. There may be qualitative differences between these vic-
tims and those who have never had police contact. Furthermore, as noted
previously, the response rate to the study is unknown. Thus, the respondents
in this study may not be representative of all women whose cases go
through the criminal justice system. The sample is self-selected. Regardless
of the limitations, this article offers a unique examination of the problem of
stalking in the context of domestic violence.
Findings
Univariate Findings
Demographic characteristics of the sample are presented in Table 6.
A total of 178 women who had been abused by their partners or ex-partners
and whose cases had gone through the criminal justice system in either themidsized city, the rural college county, or the large, metropolitan city were
recruited and interviewed at Time 1. Ninety-two of these women were from
the large, metropolitan city, 48 from the rural college county, and 38 from the
midsized city. At Time 2, the retention rate was 90%, with 36 respondents
from the midsized city, 46 from the rural college county, and 78 from the
large, metropolitan city. At Time 3, the retention rate fell to 83.1%; 34 respon-
dents from the midsized city, 44 from the rural college county, and 70 from
the large, metropolitan city. The drop in retention was most significant in thelarge, metropolitan citythe most populous of the three sites in the study.
More than half of the respondents (55.1%) were White, 22.2% were African
American, 16.3% were Latina, and 8.4% were categorized as Other
including persons who were bi- or multiracial, Native American, Asian
Melton / Predicting Stalking in DV Relationships 11
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
11/24
12 Journal of Interpersonal Violence
Table 6Demographic Characteristics of the Respondents
Time 1 Time 2 Time 3
Variables (N= 178) (N= 160) (N= 148)a
Site % n % n % n
Midsize city 21.4 38 22.5 36 23.0 34
Rural college county 27.0 48 28.8 46 29.7 44
Large city 51.7 92 48.8 78 47.3 70
Race
White 55.1 98 57.5 92 57.4 85
African American 20.2 36 18.1 29 18.2 27
Latina 16.3 29 15.6 25 16.9 25
Other 8.4 15 8.8 14 7.4 11
Ageb
18 to 29 44.4 79 43.1 69 41.9 62
30 to 44 39.3 70 43.8 70 43.9 65
45-plus 13.5 24 13.1 21 14.2 21
Education
Less than high school 14.0 25 13.8 22 12.9 19
High school graduate 25.3 45 23.8 38 23.8 35
Trade school 5.1 9 5.6 9 6.1 9
Some college 30.3 54 31.8 50 30.6 45
Associates degree 5.1 9 3.8 6 4.1 6
College graduate 15.2 27 16.3 26 16.3 24
Professional degree 5.1 9 5.6 9 6.1 9
Incomec
$0.00-$499.99 12.4 22 10.6 17 11.5 17
$500.00-$999.99 19.2 34 18.8 30 17.6 26
$1,000.00-$2,999.99 48.6 86 52.5 84 49.3 73
$3,000.00-plus 19.7 35 15.6 25 20.9 31
Number of children
0 27.5 49 25.0 40 23.6 35
1 to 3 60.1 107 64.4 103 64.2 95
4-plus 12.4 22 10.6 17 12.2 18
Relationship with assailant
Married 9.6 17 10.6 17 8.8 13
Separated 14.0 25 10.6 17 11.5 17
Divorced 10.1 18 12.5 20 15.5 23
(continued)
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
12/24
American, and European immigrant (noncitizen). The average age of the
respondents was 32, with a range from 18 to 60 years old. The majority of
the respondents either had a high school diploma or attended some college
by the time of the first interview. The majority of the respondents (48.6%)
reported a household income level between $1,000.00 and $2,999.99 per
month at the first interview. One eighth (12.4%) made less than $500.00,
19.2% made between $500.00 and $999.99, and 19.7% made $3,000.00 amonth or more. The income levels remained fairly constant throughout the
three interviews. Most of the respondents had a least one child (72.5%),
with a range from no children to six children. In regards to relationships
with the abuser, close to 70% of the respondents reported that they were no
longer with the abuser at each of the three interviews.
Summarized in Table 2 are the frequencies of stalking behaviors
reported by the respondents. Clearly, stalking was a common occurrence
among this group of abused women. It should be noted that the violenceexperienced by this group of abused women was fairly serious and wide-
spread, with almost 90% of the women reporting some violence at Time 1
and more than one third of the women experiencing violence at Times 2 and 3.
Melton / Predicting Stalking in DV Relationships 13
Table 6 (continued)
Time 1 Time 2 Time 3Variables (N= 178) (N= 160) (N= 148)a
Site % n % n % n
Girlfriend/boyfriend 13.5 24 12.5 20 10.1 15
Dating 1.1 2 0.6 1 2.7 4
Ex-girlfriend/ex-boyfriend 46.1 82 46.3 74 47.3 70
Other 5.6 10 6.9 11 4.1 6
a. Significant demographic differences between women who participated in the study at allthree time periods and women who dropped out were only found for the variable site
women were significantly more likely to drop out from the large city than from the rural col-
lege county or the midsize city.
b. The mean age for Time 1 was 32.8, with a range from 18 to 60 years old; the mean age for
Time 2 was 33.1, with a range from 18 to 60 years old; and the mean age for Time 3 was 33.4,
with a range from 18 to 60 years old.
c. The mean income at Time 1 was $2042.00, with a range from $0.00 to $20,000.00 a month;
the mean income at Time 2 was $1,813.00, with a range from $0.00 to $8,000.00; the mean
income at Time 3 was $1,956.00, with a range from $0.00 to $10,000.00.
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
13/24
14
Table7
Distinguishing
StalkedfromNonstalk
edDomesticViolenceV
ictims
Time1Stalking
Time2Stalking
Time3Stalking
Experiencea
Experience
Experience
Yes
No
Yes
No
Yes
No
Variables
%
n
%
n
Chi-Square
%
n
%
n
Chi-Square
%
n
%
n
Ch
i-Square
Relationshipwith
abuserattime
ofeachinterview
Notinarelationship
3.1
4
96.9
123
8.84**
5.3
2
94
.7
36
18.92***
6.1
2
93.9
31
4.64*
Inarelationship
15.9
7
84.1
37
54.5
12
45
.5
10
27.8
5
72.2
13
Whetherabuser
hadanalcohol/
drugproblem
attimeofeach
interview
No
16.0
8
84.0
42
10.63**
56.3
9
43
.8
7
12.44***
50.0
6
50.0
6
21.08***
Yes
2.5
3
97.5
117
11.9
5
88
.1
37
0.0
0
100.0
37
Compositeofcontrolc
Leastcontrol
41.9
13
58.1
18
12
.23**
31.3
5
68.8
11
6.25*
Somecontrol
0.0
0
100.0
21
6.9
2
93.1
27
Mostcontrol
16.7
1
83.3
5
0.0
0
100.0
6
Compositeof
0.303*b
0.383*b
physicalviolencecd
a.Thisisstalkinge
xperiencecontrollingforexperienc
ewithphysicalviolence.Onlywom
enwithexperiencewithphysicalviolenceateachtimeareincluded.Thus,a
noindicatesthew
omenexperiencedphysicalviolencewithoutstalkingandayesindicatesthewomenreportedphysicalviolenceandstalking.
b.Cautionshouldbeusedbecauseoformorecellshav
elessthan5.However,bivariatecorrelationalsoshowedasignificant
relationshipbetweenthesevariables.
c.Thesevariablesa
remeasuredatandusedforeachtimeperiod.
d.Bivariatecorrelationanalysisindicatesasignificant
relationshipbetweenwhetherorno
tsheexperiencedmorephysicalviolenceandmorestalkingthemor
esevere
thephysicalviolencethemorestalkingsheexperience
d,controllingforexperiencingphysicalviolence.
*p
.05.**p
.01.***p
.001.
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
14/24
Regarding stalking, 92.1% of the respondents reported experiencing at least
one of the stalking variables at Time 1. Although stalking decreased during
the three time periods, more than half of the respondents reported experi-encing at least one of the stalking variables at Time 2 (56.3%) and Time 3
(58.1%). Thus, even after criminal justice intervention, more than half of
the respondents continued to experience some form of stalking.
Examining the specific stalking variables, the most common stalking
behavior reported at all three time periods was checking up on (76.4% at
Time 1, 51.9% at Time 2, and 45.9% at Time 3). At Time 1, the following
stalking behaviors were reported at least once by more than half of the
respondents: The stalker broke something important to them or arrivedunwanted, or the victim received unwanted calls or messages, was fol-
lowed, or was watched. More than two fifths (43.3%) reported that their
abusers stole or read their mail. More than a third (35.4%) reported that
their abuser broke or attempted to break into their home or car. More than
a fourth (27.5%) reported that they received unwanted gifts, photos, or
letters from their abuser. Finally, 14.0% of the respondents reported that a
new partner was threatened or harmed by their abuser. However, this last
group includes all respondents not controlling for whether the woman hada new partner. When controlling for whether the respondent had a new part-
ner, 71.4% of the respondents reported that their new partner was threat-
ened or harmed by their abusermaking it the second-most-frequent
stalking behavior experienced at Time 1.
Although the stalking experiences significantly decreased in Times 2
and 3, some of the stalking variables continued to be prevalent in these later
periods. For example, at Time 2 checking up on was still reported by
more than half of the respondents (51.9%), unwanted calls was reported
by more than two fifths of the respondents (43.1%), and followed or
watched was reported by more than a third of the respondents (36.3%). At
Time 3, almost half (45.9%) of the respondents reported being checked up
on, almost two fifths (37.8%) reported receiving unwanted calls, and almost
30% reported being followed or watched (29.7%). Thus, at all three time
periods, abused women whose cases have gone through the criminal justice
system were still quite likely to be subjected to stalking.
It is important to note that using the ttest to compare means indicates
significant differences between the percentages of women experiencingviolence at Time 2 and the percentages of women experiencing stalking at
Time 2. About 38% of the respondents reported experiencing any physical
violence at Time 2; Table 2 indicates that almost three fifths (56.3%) of
the respondents reported experiencing any stalking at Time 2. Thus, this
Melton / Predicting Stalking in DV Relationships 15
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
15/24
suggests that women were more likely to continue to experience stalking
than they were violence between Times 1 and 2. This may indicate that
although domestic violence abusers stop their violence after criminal jus-tice intervention, they may continue to stalk.
A Typology of Domestic Violence Stalkers
In an effort to further understand the difference between domestic vio-
lence abusers who stalk and those who do not, cross-tabulations were con-
ducted with any stalking experience at each of the three times, controlling
for experiences with physical violence. Thus, women who had only experi-enced physical violence without stalking were compared to women who
experienced both physical violence andstalking. The significant results are
presented in Table 7. For all women who experienced physical violence at
Time 1, women who were not in a relationship with their abusers (96.9%)
were significantly more likely to experience stalking than women who were
still in a relationship (84.1%;p .01). Women whose abusers had an alco-
hol or drug problem were significantly more likely to experience stalking
(97.5%) than women whose abusers did not have an alcohol or drug prob-lem (84.0%) (p .01). These two findings also held for Times 2 and 3. In
addition, at Times 2 and 3, two additional variables were significantly
related to whether domestic violence victims who experienced physical
violence also reported stalkingcontrol and physical violence levels. First,
women who reported the most control, also reported the most stalking. The
same held true for physical violence. The more physical violence reported
at Times 1 and 2, the greater the likelihood of experiencing stalking at
Times 2 (p .05) and 3 (p .05). Thus, compared to domestic violence
abusers who do not stalk, domestic violence abusers who stalk were more
likely to (1) no longer be in a relationship with their victim; (2) have an
alcohol or drug problem; (3) exhibit more controlling behaviors in general;
and (4) be engaging in more physical violence abuse against their victims
(refer to Table 2). This may have important implications for criminal jus-
tice practitioners in their attempt to identify those at most risk for stalking
and will be discussed in the discussion.
Multivariate Findings
Numerous multivariate regression models were created for each time
period. These findings are reported in Tables 8 through 10. Table 8 presents
an OLS multiple regression at Time 1, conducted with the composite of
16 Journal of Interpersonal Violence
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
16/24
stalking as the dependent variable. Independent variables included race of
the victim, race of the abuser (0 for White and 1 for non-White), income,
age, victim-abuser relationship at the 6 months before the arrest (0 = not ina relationship, 1 = in a relationship), victims educational attainment, site,
and whether the abuser had an alcohol or drug problem (0 = no and 1 = yes).
Significant relationships were found for the variables victim-abuser rela-
tionship 6 months before the arrest, education, site, and whether the abuser
had an alcohol or drug problem. Specifically, for the variables victim-
abuser relationship at 6 months before the arrest, women who reported that
they were not involved in a relationship with the abuser during this time
period reported significantly higher rates of stalking than women who werestill intimately involved with the abuser (p .001). Table 8 also indicates
that a womans level of education was negatively related to the levels of
stalking. Women with a high school education (and possibly some college)
reported lower levels of stalking than those without a high school education
(p .01). Women with the most education reported the lowest levels of
stalking (p .05). As indicated by the bivariate analysis, abusers who were
reported as having alcohol or drug problems were also reported to have
committed higher severity levels of stalking than abusers without alcohol ordrug problems (p .05). Finally, the study site was significantly related to
the severity of stalking. Compared to respondents from the large, metro-
politan city, respondents from the rural college county were significantly
less likely to experience more severe and more frequent stalking.
Three multiple regression models were performed at Time 2 (see Table 9).
First, a regression of stalking and demographic variables (Model 1) was
conducted at Time 2, revealing a significant relationship between the victim-
abuser relationship andthe severity of stalking. Respondents who were no
longer involved with their abusers at Time 2 were more likely than women
involved to report more severe and frequent stalking (p 0.05). All the other
variables were found to be nonsignificant.
A second regression model was developed for stalking at Time 2 (see
Table 9, Model 2). In addition to the demographic variables, Time 1 abuse
characteristics were added to the equation. These included the respondents
experiences with control at Time 1 and the respondents experiences with
violence from the abuser at Time 1. Notably, of the demographic variables
the only one significantly related to stalking severity was whether theabuser had an alcohol or drug problem. Respondents who reported that
their abusers had alcohol or drug problems were more likely than those who
did not to report more severe stalking. Regarding the abuse variables, the
only Time 1 variables significantly related to the severity of stalking women
Melton / Predicting Stalking in DV Relationships 17
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
17/24
reported at Time 2 were experiences with control. More specifically, the
higher the levels of abusers control reported at Time 1, the more severe the
levels of stalking reported at Time 2 (p 0.01). The severity of violence
experienced did not significantly predict the severity of stalking experi-
enced at Time 2.
A final regression model was developed to explain stalking experiences
at Time 2 (see Table 9, Model 3). In addition to Time 1 characteristics
added in the last model, one final Time 1 characteristic was addedtherespondents experiences with stalking at Time 1. The addition of the stalk-
ing variable of Time 1 negated the significance of the control variable from
Time 1. However, stalking experiences at Time 1 became a highly significant
predictor of her stalking experiences at Time 2, with a positive relationship
18 Journal of Interpersonal Violence
Table 8
Multiple Regression of Variables Related to Predicting
Stalking Severity at Time 1
Model 1
Victim Characteristics B SE
Race (1 = non-White) 0.006 1.691
Income 0.000 0.000
Age 0.006 0.074
Educationb
High school/some college 6.794** 2.115College graduate or more 7.491** 2.557
Abuser characteristicsc
Alcohol/drug problem (1 = yes) 3.525* 1.622
Victim-abuser characteristics
Victim-abuser relationship (1 = involved) 8.820*** 1.684
Situational characteristics
Sited
Midsize city 4.552* 1.779
Rural college county 7.099*** 1.763
r2
0.306
a. Stalking severity was measured by a composite computed from the individual stalking vari-
ables. The composite ranged from 0 to 40 at Time 1.
b. Dummy variable with less than high school education as the omitted group.
c. Offender age and race could not be included in any of the regression models for it is too
highly correlated with the victim age and race.
d. Dummy variable with the large city site as the omitted group.
*p .05. **p .01. ***p .001.
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
18/24
between the severity of stalking at Time 1 and the severity of stalking at
Time 2 (p .001). A review of the standardized coefficient reveals that
stalking is the strongest predictor compared to the other variables. Thus, stalk-
ing experience at Time 1 becomes the most significant predictor of stalking
experience at Time 2. The race of the victim and the abuser, their relation-ship, income, age, education level, site, whether or not the abuser had an
alcohol or drug problem, and the respondents experiences with control and
physical violence at Time 1, all had no significant predicting value for
stalking at Time 2.
Melton / Predicting Stalking in DV Relationships 19
Table 9
Multiple Regression of Variables Related to Predicting Stalking
Severity at Time 2
Model 1 Model 2 Model 3
Victim Characteristics B SE B SE B SE
Race (1 = non-White) 1.761 1.823 2.390 1.848 2.185 1.743
Income 0.000 0.001 0.000 0.001 0.000 0.000
Age 0.005 0.078 0.002 0.079 0.003 0.075
Educationb
High school/some college 4.129+
2.131 3.197 2.290 1.666 2.188College grad or more 3.135 2.602 1.789 2.773 0.377 2.636
Abuser characteristics
Alcohol/drug problem 3.041 1.728 3.565* 1.677 2.919+ 1.595
(1 = yes)
Victim-abuser characteristics
Victim-abuser relationship 3.715* 1.849 0.154 1.718 2.195 1.695
(1 = involved)
Situational characteristics
Sitec
Midsize city 0.582 1.851 0.178 1.918 0.792 1.822Rural college county 2.762 1.770 2.381 1.840 0.810 1.777
Time 1 characteristics
Composite of control 0.244* 0.108 0.029 0.115
Composite of violence 0.001 1.694 0.003 0.051
Composite of stalking 0.390*** 0.090
r2 0.140 0.145 0.251
a. Stalking severity was measured by a composite that was computed from the individual stalk-
ing variables at Time 2. The composite ranged from 0 to 40 at Time 2.
b. Dummy variable with less than high school education as the omitted group.c. Dummy variable with the large city site as the omitted group.
*p .05. **p .01. ***p .001. + Approaching significance.
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
19/24
20
Table10
MultipleRegressiono
fVariablesRelatedtoPredictingStalkingSeverityatTime3
VictimCharacteristics
Model1
Model2
Model3
Model4
Model
5
Race(1=
non-White)
2.142
2.053
2.154
2.062
2.565
1.921
1.192
1.790
3.616*
1.560
Income
0.000
0.001
0.000
0.001
0.000
0.001
0.000
0.000
0.000
0.000
Age
0.025
0.081
0.009
0.083
0.013
0.077
0.027
0.071
0.017
0.063
Educationb
Highschool/somecollege
3.256
2.508
2.378
2.609
1.079
2.443
1.736
2.188
0.823
1.857
Collegegradormo
re
3.636
2.980
2.421
3.092
1.119
2.912
3.802
2.588
3.309
2.210
Abusercharacteristics
Alcoholordrugproblem
4.310*
1.918
4.490*
2.022
3.500+
1.916
3.298*
1.688
3.535*
1.479
(1=
yes)
Victim-abusercharacteristics
Relationshipin6m
os.
2.350
2.007
0.359
1.806
2.491
1.746
2.346
1.632
1.815
1.514
betweeninterviews
(1=
involved)
Situationalcharacte
ristics
Sitec
Midsizecity
3.114
2.009
3.154
2.033
1.443
1.927
2.557
1.781
2.642
1.559
Ruralcollegecounty
5.021*
1.968
4.130*
2.035
2.329
1.935
3.106+
1.726
2.280
1.488
Time1characteristics
Compositeofcontrol
0.027
0.122
0.270*
0.126
Compositeofviolence
0.106+
0.059
0.077
0.056
Compositeofstalking
0.449***
0.098
Time2characteristics
Compositeofcontrol
0.758***
0.146
0.084
0.156
Compositeofviolence
0.009
0.106
0.009
0.090
Compositeofstalking
0.709***
0.098
r2
0.154
0.169
0.291
0.383
0.573
a.Stalkingseverityismeasuredbyacompositethatwascomputedfromtheindividualstalk
ingvariablesatTime3.Itrangedfrom0to36atTime3.
b.Dummyvariable
withlessthanhighschooleducationastheomittedgroup.
c.Dummyvariable
withthelarge,metropolitancitysiteastheomittedgroup.
*p
.05.**p
.01.***p
.001.
+
Approachingsignificance.
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
20/24
Four regression models were created for stalking severity at Time 3 (see
Table 10). First, a model was run with the demographic variables as the
only independent variables (Model 1). In this model, whether or not theabuser had an alcohol or drug problem and the site of rural college county
were significant. More specifically, an abuser who was reported to have an
alcohol or drug problem was also more likely to be reported as committing
higher levels of stalking severity (p .05). Also, respondents from the rural
college county were less likely to report more severe stalking than respon-
dents from the large, metropolitan area (p .05).
A second model was created with the addition of Time 1 abuse charac-
teristics (see Table 10, Model 2). The same variables significant in Model 1continue to be significant with the addition of the composite of control and
violence (control and violence experienced at Time 1 do not significantly
predict stalking severity experienced at Time 3). With the addition of the
respondents experiences with stalking at Time 1, the strongest, significant
variable in Model 3 (of the models predicting stalking at Time 3) was stalk-
ing at Time 1: The more severe the stalking reported by the respondents at
Time 1, the more severe the stalking reported by the respondents at Time 3
(p
.001). The composite measure of control also reaches significance.Three additional regression models were estimated predicting stalking
severity at Time 3 (see Table 10, Models 4 and 5). Time 1 and Time 2 abuse
and stalking variables could not be run simultaneously because they are too
highly correlated. In these models, some Time 2 abuse variables were
added. Criminal justice variables were added as well but did not reach sig-
nificance and are addressed in another paper. For Model 4, with the addi-
tion of Time 2 abuse characteristics, except the severity of stalking
experiences, experiences with control was the most significant predictor of
stalking at Time 3. Women who reported more severe control at Time 2
experienced more severe stalking at Time 3 (p .05). With the addition of
stalking severity at Time 2, stalking severity at Time 2 continued to have a
positive relationship with stalking severity at Time 3. Women who reported
more severe stalking at Time 1 continued to report more severe stalking at
Time 3 (p .05). Finally, the strongest predictor of stalking severity at Time
3 was stalking severity at Time 2. Experiencing more severe stalking at
Time 2 increased the likelihood of experiencing more severe stalking at
Time 3 (p .001). Comparing the standardized coefficient for these signif-icant variables in this model (not presented in the tables) also illustrated
that experience with stalking was the strongest and most significant predic-
tor of severity of stalking experience at Time 3.
Melton / Predicting Stalking in DV Relationships 21
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
21/24
Discussion
Clearly, stalking by their current or former abuser is a prevalent problemamong this population of domestic violence victims. At the time of the first
interview, more than 90% of these women had experienced some form of
stalking. By the Time 2 and Time 3 interviews, more than 50% of the
respondents continued to experience some stalking behaviors. This con-
firms prior research on this issue (Coleman, 1997; Davis et al., 2000; Logan
et al., 2000; McFarlane et al., 1999; Mechanic, Uhlmansiek, et al., 2000;
Mechanic, Weaver, et al., 2000; NIJ, 1997; Tjaden & Thoennes, 2000;
White et al., 2000). Stalking appears to be a behavior intimately related todomestic violence. This has clear implications for how we deal with domes-
tic violence. Those victims must be given information regarding stalking,
as well as domestic violence, given that they are at high risk for experienc-
ing this behavior.
It is important to note that most of the physical violence and stalking
behaviors decreased during the three time periods (most significantly
between Time 1 and Time 2). The significant occurrence between those
time periods was criminal justice intervention. Although it is not known forcertain, it appears that criminal justice intervention decreases these behav-
iors. This issue is addressed in another paper with these data.
The next question becomes how we differentiate between domestic vio-
lence abusers who engage in physically abusive behaviors from those who
also engage in stalking. In terms of predicting stalking, experiencing prior
stalking behaviors is the best predictor. In other words, victims who have
experienced stalking within their relationships characterized by domestic vio-
lence are at greatest risk for experiencing more stalking (by their abuser) in
the future. This makes sense given that abuse is a fairly stable behavior (e.g.,
prior physical abuse will be the best predictor for future abuse and so on).
This indicates how important it is to intervene in stalking cases to stop future
stalking.
The multivariate models reveal several important findings regarding the
prediction of stalking in the context of domestic violence beyond prior
stalking experiences. First, regarding demographic variables, only the victim-
abuser relationship and the presence of an alcohol or drug problem by the
abuser appear to have any sort of significance. Without controlling for thelevel of prior abuse, only women who were no longer in relationships with
their abusers and those women whose abusers had alcohol or drug problems
were more likely to experience more severe stalking over time. Thus, abusers
may be more likely to stalk once they are no longer in a relationship with
22 Journal of Interpersonal Violence
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
22/24
their victims, supporting previous research (Logan et al, 2000; Mechanic,
Uhlmansiek, et al, 2000). Stalking may be a method of continuing the
domestic violence against the victim after the abuser has lost physicalaccess to them.
Also important, abusers who have alcohol or drug problems may be
more likely to stalk than abusers who do not have such problems. This may
be because of the fact that alcohol/drug abuse may disinhibit the impulse to
stalk. Given this correlation, it is of great import that abusers receive help
for alcohol and/or drug abuse as well as for the violence. Moreover, although
little prior literature has addressed the link between stalking and drug and
alcohol use, it has linked drug and alcohol use with more severe domesticviolence (Pan, Neidig, & OLeary, 1994; USDOJ, 1994). This study high-
lights that drug and alcohol use might even be of greater import to stalking
in the context of domestic violence than to domestic violence alone. This
could illustrate that stalking in the context of domestic violence represents
a more severe form of domestic violence. Regardless, this link between
drug and alcohol use and stalking in the context of domestic violence mer-
its greater study.
Regarding the abuse variables independent of prior stalking experiences,the victims experiences with controlling behaviors appears to have the
greatest predictive power. Women who experience more severe controlling
behaviors were more likely to experience more severe stalking behaviors
over time. This, in part, supports prior research (Mechanic, Weaver, et al.,
2000). This may illustrate that stalking is a method of maintaining control
over the victim after the abuser perceives a loss of that control.
In this study, the severity of the violence experienced did not predict
whether or not a victim experienced more severe stalking over time, which
contradicts prior research (Mechanic, Uhlmansiek, et al., 2000) and high-
lights that prediction of stalking within relationships characterized by domes-
tic violence may be more complex than previously asserted (i.e., dependent
on the type of abuse).
Thus, both the bivariate and the multivariate analysis reveal that, compared
to domestic violence abusers who do not stalk, domestic violence abusers who
stalk were more likely to (a) no longer be in a relationship with their victim,
(b) have an alcohol or drug problem, and (c) exhibit more controlling behav-
iors in general. This may have important implications for criminal justice prac-titioners in their attempt to identify those at most risk for stalking.
This has important implications for dealing with stalking in the context
of domestic violence. Given the high correlation between stalking and
domestic violence, victims who enter the system experiencing domestic
Melton / Predicting Stalking in DV Relationships 23
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
23/24
violence must be given resources regarding stalking. Particularly important
are resources regarding how to deal with stalking once a couple has broken
up. Information on protective orders, restraining orders, no-contact orders,shelters, and the criminal justice process must be given to these victims to
protect their safety. Treatment programs for offenders must address stalk-
ing as well as domestic violence. Given that many of the offenders may stop
physically abusing their partners after criminal justice intervention, but
continue to stalk their partners, attention must be paid to this behavior.
Furthermore, given that drug/alcohol abuse may be a risk factor for domes-
tic violence abusers becoming stalkers, prevention and intervention must be
focused on that behavior as well. All in all, helping practitioners identifypotential domestic violence stalkers may aid them in their treatment of both
victims and offenders, and this is an important step in our effort to deal with
the serious, criminal problem of stalking and domestic violence.
References
Baldry, A. C. (2002). From domestic violence to stalking: The infinite cycle of violence. In
J. C. W. Boon & L. Sheridan (Eds.), Stalking and psychosexual obsession: Psychologicalperspectives for prevention, policing, and treatment(pp. 83-104). Chichester, UK: Wiley.
Brewster, M. P. (2002). Stalking by former intimates: Verbal threats and other predictors of
physical violence. In K. E. Davis & I. H. Frieze (Eds.), Stalking: Perspectives on victims
and perpetrators (pp. 292-311). New York: Springer.
Burgess, A. W., Baker, T., Greening, D., Hartman, C. R., Burgess,A., Douglas, J. E., et al. (1997).
Stalking behaviors within domestic violence.Journal of Family Violence, 12(4), 359-403.
Coleman, F. L. (1997). Stalking behavior and the cycle of domestic violence. Journal of
Interpersonal Violence, 12(3), 420-432.
Davis, K. E., Ace, A., & Andra, M. (2000). Stalking perpetrators and psychological maltreat-
ment of partners: Anger-jealousy, attachment insecurity, need for control, and break-upcontext. Violence and Victims, 15(4), 407-425.
Dobash, R. E., & Dobash, R. P. (1998).Rethinking violence against women. Thousand Oaks,
CA: Sage.
Guy, R. A. (1993). The nature and constitutionlity of stalking laws. Vanderbilt Law Review,
46(4), 991-1029.
Logan, T. K., Leukefeld, C., & Walker, B. (2000). Stalking as a variant of intimate violence:
Implications from a young adult sample. Violence and Victims, 15(1), 91-111.
McFarlane, J. M., Campbell, J. C., Wilt, S., Sachs, C. J. U. Y., & Xu, X. (1999). Stalking and
intimate partner femicide.Homicide Studies, 3(4), 300-316.
Mechanic, M. B., Uhlmansiek, M. H., Weaver, T. L., & Resick, P. A. (2000). The impact ofsevere stalking experienced by acutely battered women: An examination of violence, psy-
chological symptoms and strategic responses. Violence and Victims, 15(4), 443-458.
Mechanic, M. B., Weaver, T. L., & Resick, P. A. (2000). Intimate partner violence and stalk-
ing behaviors: Exploration of patterns and correlates in a sample of acutely battered
women. Violence and Victims, 15(1), 55-72.
24 Journal of Interpersonal Violence
by RAVI BABU BUNGA on October 23, 2011jiv.sagepub.comDownloaded from
http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/http://jiv.sagepub.com/7/29/2019 3.FullPredicting the Occurrence of Stalking in Relationships Characterized by Domestic Violence
24/24
National Institute of Justice (NIJ). (1996).Domestic violence, stalking, and antistalking legis-
lation: An annual report to Congress under the Violence Against Women Act. Washington,
DC: Author.Pan, H., Neidig, P., & OLeary, K. (1994). Predicting mild and severe husband-to-wife physi-
cal aggression.Journal of Clinical Psychology, 62, 975-981.
Roberts, A., & Dziegielewski, S. (1996). Assessment typology and intervention with the sur-
vivors of stalking.Aggression and Violent Behavior, 1(4), 359-368.
Straus, M. (1979). Measuring intrafamily conflict and violence: The Conflict Tactics (CT)
Scale.Journal of Marriage and Family, 41(1), 75-88.
Sullivan, C. M., Parisian, J. A., & Davidson, W. S., II. (1991, August).Index of Psychological
Abuse: Development of a measure. Paper presented at the American Psychological
Association Annual Meeting
Tjaden, P. (1997). The crime of stalking: How big is the problem. Washington, DC: U.S.Department of Justice, National Institute of Justice.
Tjaden, Patricia and Nancy Thoennes. 2000. The Role of Stalking in Domestic Violence
Crime Reports Generated by the Colorado Springs Police Department. Violence and
Victims 15(4):427-41.
U.S. Department of Justice (USDOJ). (1994). Bureau of Justice statistics selected findings:
Violence between intimates. Washington, DC: Author.
USDOJ. (1998). Stalking in America: Findings from the National Violence Against Women
Survey. Washington, DC: Author.
USDOJ Violence Against Women Grants Office. (1998). Stalking and domestic violence: The
third annual report to Congress Under the Violence Against Women Act. Washington, DC:Author.
White, J., Kowalski, R. M., Lyndon, A., & Valentine, S. (2000). An integrative contextual
developmental model of male stalking. Violence and Victims, 15(4), 373-388.
Heather C. Melton, PhD, is an assistant professor at University of Utah in the Department of
Sociology. Her research focuses on violence against women and the criminal justice response
to it. She earned her doctorate in sociology at the University of Colorado at Boulder.
Melton / Predicting Stalking in DV Relationships 25