Bethel University Distance Analysis Journey to Crime Analysis Essay 1. For this unit’s Complete assignment, write a comprehensive scholarly essay (minimum

Bethel University Distance Analysis Journey to Crime Analysis Essay 1.

For this unit’s Complete assignment, write a comprehensive scholarly essay (minimum 1500 words) in which you analyze, explain, and apply these concepts in the context of a law enforcement crime mapping issue. You must incorporate and cite, using correct APA citation format, at least four different scholarly research sources.

Be sure your essay demonstrates a comprehensive understanding of the READ and ATTEND sections from this unit. In-line citations must be used in the body of your essay, and all research sources must be fully cited at the conclusion of your essay. Correct APA citation formats must be used.

After reviewing the material from the READ and ATTEND sections for this week write a comprehensive essay that includes the following elements:

Choose one of the five types of distance analyses and apply it successfully to a particular criminological event of interest in your jurisdiction or region. Demonstrate that your choice of that type of distance analysis is applicable to the information you wish to find.
Identify and describe two salient issues in the varied efforts of mapping crimes for two distinct audiences, their purposes and recommended types of presentation.
Discuss why are tactical maps so important? What are some of the barriers to creating effective tactical maps?
What is spatial autocorrelation, and why is it important in crime mapping and spatial-behavioral studies? Provide a practical example of how it might affect crime mapping. © Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
CHAPTER
Distance Analysis
LEARNING OBJECTIVES
W
Chapter 9 begins with a discussion of several typesO
of distance analysis. Each
type of distance analysis has strengths and weaknesses that must be understood
O be able to:
prior to interpretation. After studying this chapter, you should
• Identify the common types of distance analyses usedD
in crime mapping.
• Explain the appropriate uses of distance analysis. S
• Understand the strengths and limitations of distance, analysis.
9
KEY TERMS
Euclidean Distance
Manhattan Distance
S
H
All analyses that rely on distance measures are distance analyses.
A in Chapter 10 could
Thus, the hot spot analyses that will be examined
N determination of a
also be viewed as distance analyses in that the
clustering of events requires that individual events
are located closer
I
together than we would expect based on random chance. In this case,
A criminal incidents
the distances measured are strictly those between
Introduction
(distances between homicides in a city, for example). This chapter discusses those analyses where distance is measured
1 in relation to another
point or in efforts to find another point (the mean center of a distribution, an offender’s home, predicted next target,5or some other point of
importance such as a school, bar, or pawn shop).
9 The distance analyses
to be discussed in this chapter include mean center
analysis, journey
5
to crime analysis, spider distance analysis, proximity analysis, distance
T
between hits analysis, and distance and time analysis.
There are several different methods of calculating
distance. Each
S
has its strengths and limitations in analysis. In crime analysis, the
two most common methods of distance calculation are Euclidean and
Manhattan. Euclidean distance is measured by measuring the distance
between two points. Often referred to “as the crow flies” measurement,
it is the shortest distance between two points on a map. The problem
is that there rarely is a road that leads directly from point A to point B
335
55751_CH09_335_370.pdf 335
3/23/09 3:20:04 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
336
CHAPTER 9
Distance Analysis
Point B
Dark outer lines and inner staircased line
are equal.
Light gray diagonal line is shorter than the
other 3 lines.
Point A
Figure 9–1 Distance Illustration
W
O
O
D
S
,
on a map (have you ever navigated Boston?), and travelers usually must
take a series of twists and turns on their destination. Thus, Euclidean
calculations of distanceSare typically smaller than the path actually
traveled by an individual.
H Manhattan distance or “street” distance,
as it is often called, is calculated by using right angles to get from one
A
point to another. In practical terms, individuals do not always travel
in right angles to get to N
their destination (again, have you ever driven
in Boston?). ManhattanI calculations are always larger than Euclidean calculations because the shortest distance between two points is
9–1).
a straight line (see FigureA
You may be wondering at this point which is the best calculation
method to use. The answer
1 depends on the type of analysis you are
performing and the data you are using (including the street layout of
5 distance is much easier to calculate; howthe study area). Euclidean
9
ever, Manhattan more closely
approximates the distance traveled by
individuals. Rossmo et al.
(2005)
observe that:
5
Research has shown that Manhattan distance gives the most accurate
T
result in the greatest number of cases, while not being significantly worse
S the entire spectrum of cases—a finding true
than other methods across
for both North America and Great Britain . . . As long as these specific
exceptional cases can be identified and recognized through training and
experience, the most reliable and practical method involves the use of
Manhattan distance. (p. 111)
The best method of distance calculation is dependent on the type
of analysis that is to be performed. Groff and McEwen (2005) found
55751_CH09_335_370.pdf 336
3/23/09 3:20:08 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
Distance Analyses
337
a strong correlation between Euclidean and street distances and thus
argued that one could use the coefficients of Euclidean measures in a
regression model to estimate street distances.
Distance Analyses
A variety of distance measures can be used in crime analysis and
mapping, and in reality, distance and how features are related to each
other on a map are what crime mapping is all about. The very technical
term “Spatial Autocorrelation” has a fairly simple explanation to crime
mappers, which is that those features that areW
closest to one another
are likely more related to other features near O
it than to those further
away. If we think about this within our own lives, we might often find
our neighbors, friends, and co-workers nearer O
to us in geography than
Doccupation, interests,
persons who live further away or have no similar
or residence. The first type of distance analysis
S we’ll discuss is mean
center analysis, which is used for tactical analysis purposes.
Mean Center Analysis
,
The mean center of any distribution is, very simply, the point
S of Y (longitude) of
at which the mean of X (latitude) and the mean
events meet on a map. It represents the averageH
or the center of gravity
of a spatial distribution (Levine, 2002). The problem
with performA
ing a mean center analysis is that it is sensitive to outliers (recall from
Chapter 8 that the mean itself is sensitive toN
outliers and thus cannot be relied upon in skewed distributions).IIn addition, the mean
center of distributions that are multimodal or odd shaped (such as
A
in an L-shaped distribution) may be placed at a point where very few
crimes actually occur. Thus, the utility of mean center analyses is not
in finding a point on the map to throw more resources
at or the apex
1
of crime in an area. Rather, it is a reference point to be used in further
5
analyses, such as in the comparison of two different distributions (same
9 same time), and as a
crime but different time, or different crime but
starting place to begin prioritizing places and persons
of interest (such
5
as performing a standard deviation ellipses analysis; recall that stanT
dard deviation was discussed in Chapter 8). There are several different
S perform. CrimeStat
types of mean center analyses that an analyst can
performs several different mean center analyses, including the mean
center, the harmonic mean, and the geometric mean. Note that when
viewing Figure 9–2 at a smaller scale, the three mean centers appear to
be located at the same point.
However, in Figure 9–3, a larger scale reveals that the three mean
center analyses are indeed three separate points.
55751_CH09_335_370.pdf 337
3/23/09 3:20:09 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
338
CHAPTER 9
Distance Analysis
W
O
O
D
S
,
S
H
A
N
I
A
1
5
9
5
T
Figure 9–2 Mean Center Analysis, Burglaries 2003, Glendale, Arizona, Small Scale
Source: CrimeStat and ArcGIS
S
What does this all mean to the analyst? If we remember that maps
are simply representations of the real world and our goal is to be as
accurate as possible when doing analyses, then deciding which distance
method to use is just part of that well-planned analytical process.
55751_CH09_335_370.pdf 338
3/23/09 3:20:09 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
Distance Analyses
339
W
O
O
D
S
,
S
H
A
N
I
A
1
5
9
5
Figure 9–3 Mean Center Analysis, Burglaries 2003, Glendale, Arizona, Large Scale
T
Source: CrimeStat and ArcGIS
S
Journey to Crime Analysis
A journey to crime analysis is one type of distance analysis and
is used primarily in investigations of a crime series that is thought to
be attributed to an individual or group of persons acting together. It
is conducted in hopes of prioritizing areas in which the offender or
offenders are most likely to live or work. Recall from Chapters 3 and 4
55751_CH09_335_370.pdf 339
3/23/09 3:20:10 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
340
CHAPTER 9
Distance Analysis
TABLE 9–1 Average Distance to Crime in Glendale, Arizona
Crime Type
Aggravated assault
Arson
Number of
Crimes
8,526
Standard 68% of Crime
Furthest
Mean Crime Deviation Trips Within
(Miles)
Distance
(Miles)
Trip (Miles)
1.18
2.86
4.04
6.91
62
1.32
2.68
4
6.67
Auto theft
1,058
2.88
3.55
6.43
9.97
Burglary
1,356
2.34
3.73
6.07
9.79
164
1.85
3.44
5.28
3.39
5.66
9.06
4.42
6.8
11.23
3.9
6.32
10.22
2.66
4.27
6.92
2.79
4.14
6.94
Curfew/loitering
567
W 1.59
2.27
O 2.38
O 2.42
D 1.61
S 1.35
, 3.23
4.21
7.44
11.65
Runaway
5,958
0.19
1.29
1.48
2.77
Theft
5,139
3.18
4.28
7.46
11.75
Drug offenses
Murder
Other miscellaneous
Other sex offenses
Rape
Robbery
3,970
70
18,237
404
85
S
H
that offenders are fairlyA
routine in their travels for both criminal and
noncriminal behavior (see TABLE 9–1).
N generally tend to travel greater distances for
In addition, offenders
property crimes than they
I do for violent crimes, and the likelihood of
offenders to commit anyAcrime dissipates as they get farther away from
their home. (Also recall that if offenders commit crime in relation to
an anchor point that is not their home—for example, if the anchor
point is their place of employment—distance
decay measures from
1
home to crime should not be used if more accurate work to home
5
distances are available.)
9research traditionally has examined the route
The journey to crime
undertaken by offenders
5between their home and place(s) where the
crime was committed. However, this ignores the victim’s journey to
T
his or her victimization and also ignores other nodes where offenders
S school, parole office; Costello & Leipnik,
maybe traveling from (work,
2003). Groff and McEwen (2005) found “clear differences in travel
behavior between victims and offenders” in their study of homicides
in Washington, DC (p. 60). They calculated both Euclidean and street
distances for both offenders and victims and found that for homicides, victims traveled a median of 0.69 street miles (0.54 Euclidean
55751_CH09_335_370.pdf 340
3/23/09 3:20:10 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
Distance Analyses
341
miles). Median distances were chosen due to the skewed nature of the
data set. (Both victims and offenders of homicide tended to stick very
close to home.) The distance traveled for both victims and offenders
varied according to the originating motivation for the homicide. For
robbery-motivated homicides, victims were killed about 1 mile from
their homes. For retaliatory, drug, and gang-related homicides, victims
were killed a median of 0.67 miles from their homes. Male victims traveled farther to their murders than did female victims in every category
of homicide (with the exception of domestic violence homicides).
You may be wondering why we included W
a discussion of victims’
travel to crime. The answer is straightforward. Crime, in spatial analysis, must be viewed as a criminal event, and O
thus the behaviors and
travel patterns of victims are equally important
O to the behaviors and
traveling patterns of offenders. Although crime
D is largely opportunistic, and thus any number of targets may be equally desirable in an
S in serial crimes), it
offender’s eyes (although this not always the case
is important to understand how victims and ,offenders interact with
their environments to gain a better picture of how a crime unfolds
from multiple perspectives. You will also find that law enforcement
S victims, and it may
agencies do not often ask these questions of crime
take considerable effort to change the vision ofH
information collection
to include collection of travel and behavior of victims in less serious
A
crimes.
N
I
Spider distance analysis draws lines from each point in a distribuA helps to answer sevtion to its centroid, or mean center. Spider analysis
Spider Distance Analysis
eral questions in crime analysis. First, in a crime series, is the offender
likely to be a poacher or a hunter? Poachers or marauders exhibit fairly
1 committing crimes
predictable patterns in their offending, usually
5
short distances around their central base (typically
their homes, but
not always). Hunters or commuters, on the 9other hand, are much
less predictable in their offending patterns. Spider distance analyses
5 a series is expanding
can provide clues to analysts to whether or not
T
outward or shrinking inward.
Buffer Analysis and Queries (Theme Selection)
S
There are two general types of proximity analysis: buffer analysis
and queries. A buffer zone analysis is completed by drawing circles
around points of reference, such as pawn shops or city parks, at distances determined by the analyst. Essentially, the points (locations of
pawn shops or parks) serve as centers (or centroids) for the buffers
55751_CH09_335_370.pdf 341
3/23/09 3:20:11 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
342
CHAPTER 9
Distance Analysis
with a radius set at a distance desired by the analyst (this can be done
in miles, meters, feet, and so on). For example, an analyst might wish
to draw 0.5-mile buffers around all pawn shops in his jurisdiction.
He then may wish to add to a map layer of known burglaries within
its jurisdiction and visually scan to see those burglaries that occur
within a half mile of a pawn shop. At other times the analyst may want
to analyze crime around city parks. A specific park may be the local
hangout of transients, and the analyst may want to look at the crime
around the park to determine if the crime rate in and around that park
is higher than other parks
Win similar neighborhoods. This analysis may
assist with planning efforts and enforcement of no camping ordinances
O surround high transient population areas.
and other issues that often
In this case a point is not
Othe center of the buffer, but the limits of the
park polygon would actD
in this manner, and then the buffer would be
drawn, following the park outlines outward by 0.5 miles. Buffer zone
S of distance analysis and is easily interpreted.
analysis is a common type
See Figure 9–4 as an example
, of a buffer analysis.
An analyst may also want to query incidents that are within a
distance of some other type of point data, such as schools, liquor
establishments, or pawnSshops. For example, an analyst may be called
upon to produce a mapHof crimes that are within a given distance of
schools. Another example might be to query store robberies that are
A
within a given distance of a freeway entrance. The benefit to using
N
a query over a buffer analysis
is that points that are not within the
specified distance are removed
from the map, lessening clutter and
I
improving map clarity. Another related query is one that queries the
A
number of points within a polygon, such as a police precinct or beat.
In fact, any concept of space and relationships in space can be queried,
such as within a distance
1of, totally within, intersected by, or adjacent
to something else in a map layer in GIS. The analyst can even use
5
graphic objects drawn on the map display to query points, lines, and
polygons in relation to 9space and distance. In addition, one has the
option with queries to create
5 a new theme and save it as its own map
layer. However, buffer analysis can be more visually appealing, such as
T because one can see the physical proximity
in courtroom illustrations,
S the point of reference. (For example, a map
between the incident and
could have one point to represent a school and another to represent
the home of a known sex offender.)
Distance Between Hits Analysis
A distance between hits analysis is the calculation of the distance
between each crime in a crime series (crimes committed by the same
55751_CH09_335_370.pdf 342
3/23/09 3:20:11 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
Distance Analyses
343
W
O
O
D
S
,
S
H
A
N
I
A
1
5
9
5
T
S
Figure 9–4 1000-foot Buffer of Glendale Schools
Source: ArcGIS
55751_CH09_335_370.pdf 343
3/23/09 3:20:11 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
344
CHAPTER 9
Distance Analysis
offender(s)) to determine the most likely distance the offender may
choose from the most current hit. The concept here comes from the
same mean and standard deviation calculations we learned in Chapter 8.
We place the crimes in a series on the map and measure the distance
from the first crime to the second, the second crime to the third, and
so on. We then calculate the average and the standard deviation of
those distances. Let’s say the following were the distances we found
between five hits in a crime series:
•
•
•
•
First to second: 2.0 miles
Second to third: W
1.5 miles
Third to fourth: O
1.25 miles
Fourth to fifth: 0.75 miles
O
By reviewing the distances between crimes in this series, we can
D
determine a few things. First, the average distance this offender traveled
S deviation of 0.52 miles. This means that
was 1.38 miles with a standard
68% of the time the offender
traveled from 0.85 to 1.9 miles between
,
hits, and 95% of the time the offender traveled from 0.33 to 2.42 miles
between crimes. Second, by looking at the distances the offender traveled between crimes, weScan see that a pattern is emerging between
each hit in that the distance
H from one hit to the next decreased over
time. Common sense would lead us to the conclusion that the offender
will probably travel less A
than 0.75 miles from the fifth hit in this series
to the next crime he willNcommit in the future. Depending on the circumstances of the crimes,
I the targets or victims that are chosen, and the
modus operandi, we might be able to use this information efficiently
A might go in the future.
to predict where the offender
Another side benefit of this analysis for tactical purposes is that
a directional pattern might
1 also be revealed. In Figures 9–5 and 9–6 we
can see the same crimes where there is no visible pattern and one
crime where there is a 5
specific directional pattern to the offender’s
activities.
9
Distance and Time Analysis
5
Distance and time are
T also very often related. When an analyst
begins looking at the distribution of crimes based on how far away
S
they are from another crime, a specific location, or a geographic region,
he or she should also consider the tempo of the events. In the simple
examples illustrated in Figures 9–5 and 9–6, if we looked at the time
element, we might also see a pattern where the longer the offender
waits between hits, the farther he or she travels between each crime
in the series. This may relate to the amount of money the offender got
55751_CH09_335_370.pdf 344
3/23/09 3:20:12 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
Online Resources
345
W
O
O
D
S
,
Figure 9–5 Path and Direction Not Clear
Source: ArcGIS
S
H
A
N
I
A
1
at each crime site and how much money that offender needs to satisfy
5
a drug habit, for example.
9
5
ArcGIS Spatial Analyst: Distance Analysis:
T http://www.esri.com/
software/arcgis/extensions/spatialanalyst/about/distance.html
S
CrimeStat III: http://www.icpsr.umich.edu/CRIMESTAT/
Online Resources
•
•
• CrimeStat III User Workbook. This entire workbook is extremely
helpful: http:// www.icpsr.umich.edu/CRIMESTAT/workbook/
CrimeStat_Workbook.pdf
• Spatial Predictive Analysis Crime Extension: http://www
.bairsoftware.com/space/help.html
55751_CH09_335_370.pdf 345
3/23/09 3:20:12 PM
© Jones & Bartlett Learning, LLC. NOT FOR RESALE OR DISTRIBUTION
346
CHAPTER 9
Distance Analysis
W
O
O
D
S
,
S
H
A
N
I
A
Figure 9–6 Path and Direction Clear
…
Purchase answer to see full
attachment