Jason A. Bakas, Joe Ilsever, Patrick Neal, Alana N. Cook & P. Randall Kropp

thumbsshoesΛογισμικό & κατασκευή λογ/κού

11 Δεκ 2013 (πριν από 3 χρόνια και 10 μήνες)

300 εμφανίσεις

Jason A. Bakas
1,2
, Joe Ilsever
2
, Patrick Neal
2
, Alana N. Cook
3
& P. Randall Kropp
4


1
American Military University,
2
British Columbia Institute of Technology,
3
Simon Fraser University,
4
Forensic Psychiatric Services Commission
Contact: jason.bakas@mycampus.apus.edu

Abstract

Criminal analysts have a unique skill set that can be applied to help understand various forms of criminal behaviour, this
includes geographic analytic techniques. The current study filled a gap in the field by applying geographical analytic
techniques to a form of violent behaviour that was previously unexplored by these methods, stalking. The current results
showed that the distance of stalking behaviour and direct contact by offenders was not significantly related to risk of
continued stalking or risk for physical harm. Offenders who engaged in direct contact showed a significantly lower
average distance of stalking behaviour compared to cases with no direct contact. The offender’s residence was the
primary anchor point from which stalking behaviour originated, and
the victim’s home was
the primary anchor point in
which the stalking behaviour took place . The average distance between stalker and victim anchor points in both groups
was 5.9 km with a large range of distances. In addition to these findings, this study highlighted the limited relevant data
available for crime analysis techniques in understanding stalking. As such, a recommended stalking information sheet
was developed. Implications for crime analysts and investigators are discussed.

Geospatial Analysis

Criminal analysts have a unique skill set that can be applied to help
understand various forms of criminal behaviour. One analytic
technique is Geospatial Analysis. Brantingham and Brantingham
(1993) have demonstrated how both crimes and offender behaviour
are not random, but in fact discernible patterns which are greatly
influence by the geographical environment. Geospatial analysis is
the understanding of these patterns and of the offender’s
behavioural geography (spatial decision making process) and how
behavioural geography affects the way he or she offends.

One area of particular importance within behavioural geography is
the understanding of how one’s spatial knowledge impacts an
individual’s spatial activities, or “spatial behaviour” (Brantingham &
Brantingham, 1993; Regert, 1989). The theory of spatial behaviour
suggests that an offender’s propensity to commit criminal activity in
a given geographical area is reliant on the offender’s knowledge
and comfort in that area. Specifically, offenders will select targets in
a given area based on four facets: (1) an offender’s mental map
(i.e., highly detailed cognitive understating of the routes and
landmarks in a familiar area), (2) awareness space (i.e., the general
understanding of given area); (3) activity space (i.e., one’s habitual
geography made up of places routinely visited and there connected
routs); and (4) journey-to-crime (i.e., the distance one will travel to
commit an offence). Most offenders will travel only a relatively short
distance from home, or other anchor points, to commit the majority
of their crimes (Rossmo, 2000). Geospatial analysis has been
helpful in understanding and solving many crimes of a serial nature,
including sexual assault, robbery, and murder (Rossmo, 2000).
Procedure

The cases were already coded using a set of structured
professional guideline validated for use in assessing risk
for stalking violence - the
Guidelines for Stalking
Assessment and Management
, or SAM (Kropp et al.,
2008) - on the basis of file review in the data provided by
Dr. Kropp. Conclusory opinions of the case as rated on
the SAM were used the current study (i.e., Risk for
Continued Stalking [Low, Moderate, High] and Risk
Serious Physical Harm [Low, Moderate, High]). The cases
were also coded for motivations for stalking risk. These
included (M1) defence/distance, (M2) justice/honour, (M3)
gain/profit, (M4) control/change, (M5) status/esteem, (M6)
release/expression, (M7) arousal/activity, (M8) proximity/
affiliation.

Results

There was direct contact between the offender and victim in
82.9% of cases - all other had indirect forms of contact, such as
letters or phone calls (14.2%; 2.9% unknown). The offender
anchor points were primarily from the offender’s residence
(88.6%) and were infrequently from associates residence
(2.8%) and from the offender’s workplace (5.7%; 2.9%
unknown). The victim anchor points were primarily at the victims
home (57.1%), and follwing than from victim’s workplace
(34.3%), and least frequently from a victim’s associates
residence (5.7%; 2.9% unknown).
The average distance between stalker and victim anchor points
was 162.7 km (range, 0.1 – 4423.0;
SD
= 764.5). However,
when examined for outliers it was evident that five of the cases
were deviated from the sample. When outliers (n=5) were
removed the average distance between stalker’s and victim’s
anchor point was 5.9 km (range, .1 – 31.1,
SD
= 7.2).

A series of logistic regression analyses were conducted and
revealed that (1) the relationship of distance of stalking and risk
of continued stalking was not statistically significant,
F
(1,33) = .
989,
p
= .33; (2) the relationship of distance of stalking and risk
of physical harm was not statistically significant,
F
(1,33) = .16,
p

= .69; (3) direct contact and risk for continued stalking was not
statistically significant,
F
(1,33) = .07,
p
= .79; and (4) direct
contact and risk for physical harm was not statistically
significant,
F
(1,33) = .66,
p
= .42.
An independent samples
t
test was conducted to determine (5)
if offenders who engaged in direct contact had a significantly
lower mean distance in stalking behaviour compared to no
direct contact offenders. The test was significant,
t
(32) = -2.5,
p
<.05. Offenders who had direct contact with their victim (
M
=
35.20
, SD
=144.63)

on average had a closer distance in their
stalking behaviour than offenders who did not have direct
contact (
M
= 902.16,
SD
= 1968.28).

A step-wise linear regression analysis was conducted to
evaluate (6) motivators for risk of continued stalking and risk for
continued stalking. The regression was statistically significant,
F
(1, 34) = 2.81,
p
<.05. The stepwise model determine (M7)
arousal activity was the only motivator that predicted the model,
t
=13.71,
p
<.01.

Discussion

The current results showed that the distance of stalking behaviour
or that direct contact by offenders to their victim was not
significantly related to risk of continued stalking or risk for physical
harm. These findings may be indicative of the unique relationship
often found between stalkers and their victims that may be different
from other serial offenses.

Offenders who engaged in direct contact showed a significantly
lower average distance of stalking behaviour, and therefore traveled
a closer distance to stalk their victims, compared to cases with no
direct contact (i.e., letters or phone calls). The offender’s residence
was the primary anchor points from which stalking behaviour
originated, compared with associate’s residence or least frequently
offender’s workplace. The victim’s primarily anchor point, in which
the stalking behaviour took place was at the victim’s home
compared with the victim’s workplace or least frequently the victim’s
associate’s residence. The average distance for that total between
stalker and victim anchor point was 5.9 km.

In comparison, when using a non-geospatial model to examine for
risk of continued stalking and risk for physical harm there was
significant results. Motivating factors showed a relationship for risk
of continued stalking in addition to risk for physical harm.
Specifically, arousal activity demonstrated the greatest relationship
with risk for future stalking. Motivation may be playing a more
significant role is risk for future stalking and harm than geospatial
variables.

One limitation was the “dark figure” of unrecorded details or
instances in the stalking behaviour. The offender may have stalked
at multiple location points, and through multiple methods (i.e.,
phone calls and physical contact) not recorded.

Recommendations for Practice

The cases reviewed lack important information that may help
understanding stalking from any analytic perspective. This was not
known until after data collection began. For example, data
necessary to run geospatial analysis are not routinely collected or
recorded by arresting or attending officer in stalking cases.

To better understand the geospatial significance for future studies
and to outline the type and general history of the instance of
stalking for threat assessment and investigation, a check-list tool is
needed. This tool must outline routine information and history of the
stalking offence and provides a standard systematic methodology
for arresting officers to fill out. This would provide investigators and
crime analysts a comprehensive data list of information relating to
the crime. As well as allowing analysts to have a more complete set
of data to better understand offender pattern behaviour, and help
transform the current reactive approach of stalking enforcement to a
more proactive approach. For this reason the first author (Bakas)
developed a pilot check-list for attending patrol officers and
investigators to complete in Criminal Harassment cases:
The
Stalking Information & Analysis Sheet
(see Attached Sheet
).

The cases were also coded for the offender’s primary
anchor point in which the stalking behaviour was initiated
and the victim’s primary anchor point in which the stalking
behaviour occurred (i.e., residence, friends’ residence,
workplace/ school). See example above. The cases were
coded for the presence of direct contact. Direct contact was
defined as the action of physical, or in-person, contact
between the offender and victim.
Stalking Behaviour

Stalking is defined by Kropp and colleagues (2008)
as unwanted and repetitive interactions between an
offender and a victim. This behaviour can be
communication (e.g., phone calls, letters), contact
(e.g., attending victim’s workplace), or other
conduct that is deliberate or reckless to the point it
causes significant concern or fear for the victim’s
safety. Importantly, stalking is, by definition, a
repetitive
crime committed by the same offender,
targeting the same or similar victims, using a
common modus operandi, or one that is serial in
nature, lending itself appropriately for geospatial
analysis

Method

The sample consisted of 35 individuals who have
been convicted of stalking related offences under
the Criminal Code of Canada. The offenders were
all 18 years or older (M age = 35) and were mainly
male (n = 32; 91%). The victims were primarily
female (n = 29; 83%). The offender-victim
relationship included 23 cases (66%) of familiar
relationships (e.g., intimate partners), 7 cases
(20%) of acquaintances (e.g., employment related),
and 5 cases (14%) of stranger relationships.

Fictitious Example Geospatial Map of Stalker Anchor Points