GEng 198 – Methods of Research

SPECIAL STUDIES/THESIS/RESEARCH PROPOSAL

Special Study Title: GIS-Based Mapping and Identification of Theft Crime Hotspots in
Poblacion Areas of Baybay City, Leyte
Student Name: Aninao, Mhejoy A.
Contact Number: +639 488610917
Email Address: aninaomejoy@gmail.com
Address: Brgy. Macanip Jaro, Leyte
Adviser: Rianel April A. Palo Signature: ______________

1 INTRODUCTION

1.1 Background of the study

Crime refers to actions that violate the laws and regulations of a society or governing
body. These actions can range from theft, physical injury, and murder (known as index crimes)
to violations of specific laws (known as non-index crimes). Theft, in its legal definition, is the
act of intentionally and unlawfully taking the property of another person without their consent,
with the intent to deprive them of it permanently (Black's Law Dictionary, 2023). It essentially
boils down to the misappropriation of another's possessions for personal gain or other
motives.

Theft is a serious crime that can occur in different settings, whether it is within
commercial establishments or private homes, and it is important to understand the legal
implications of each type of theft. Commercial theft occurs within establishments like shops
or offices, targeting merchandise or cash through methods like shoplifting, employee theft,
burglary, or robbery (Act No. 3815, Articles 308-312). Residential theft, on the other hand,
involves taking personal belongings from private dwellings through burglary, home invasion,
package theft, or mail theft, violating the security of our homes (Revised Penal Code, Articles
308-309).

Crime can be influenced by various factors, such as socioeconomic conditions,

neighborhood characteristics, and the availability of security measures. Socioeconomic data
refers to quantitative and qualitative information that describes the economic and social
characteristics of individuals, groups, and communities. It encompasses a broad range of
factors that influence their well-being, opportunities, and life outcomes (United Nations
Development Programme, 2023). These factors can contribute to the prevalence of theft in
certain areas and impact the likelihood of individuals committing this type of crime. According
to Ansari and Kale (2014), there are many reasons why crime occurs, including illiteracy,
poverty, or revenge. A wide range of socioeconomic variables, such as household income,
educational achievement, employment status, poverty status, and demographic variables,
such as age, religion, and race, are also associated with crime. By examining the relationship
between these variables and the type of crime committed, we can better understand crime in
a specific area (Mitchell, 2011).

Special Studies Proposal Template version 1.2.2023

 Crime does not exhibit an even distribution; instead, it tends to cluster in specific
regions, and this clustering is observable in various environments (Eck, 2005). Specifically,
individuals in urban and densely populated areas demonstrate heightened alertness
compared to those in tranquil, less crowded locations. This observation underscores a
nuanced awareness among people regarding crime hotspots, where the likelihood of criminal
activities is perceived to be higher. According to Eck (2005), the term "hotspot" refers to
regions with a high concentration of criminal activity and is crucial in crime mapping and
analysis. Understanding the dynamics behind the emergence of these hotspots is crucial for
effective crime prevention and law enforcement strategies. Several studies have underscored
the significance of identifying and studying crime hotspots to decrease crime rates and
enhance the effectiveness of law enforcement services. Crime mapping, an essential
component of crime analysis, provides a platform for identifying and visualizing crimes,
including their locations, time, and other relevant data. Spatial distribution analysis can
provide security agencies with detailed information on vulnerable areas, enhancing security
and addressing specific issues (Fredrick, 2023).

Crime is a significant concern for every society, and accurate information

dissemination is critical in addressing this issue. Maps can be powerful tools in this regard,
as they can convey a significant message in a simple and effective manner. Traditionally,
crime mapping involves using physical maps and pins to mark the locations where crime
incidents occur. Over time, as technology has advanced, the utilization of Geographic
Information Systems (GIS) has become widespread in crime mapping and analysis.

A Geographic Information System (GIS) is a computer-based tool that captures, stores,

analyzes, manages, and presents data with a geographic component (Longley et al., 2021).
Essentially, it allows us to connect location-based information with other relevant data to
understand and visualize patterns, relationships, and trends across space. In recent years,
Geographic Information System (GIS) technology has become increasingly prominent as a
beneficial tool for crime analysis. Multiple studies have examined the effectiveness of
Geographic Information Systems (GIS) in reducing crime rates and identifying areas with high
crime activity. The study area can be analyzed using spatial and statistical techniques, such
as neighborhood and correlation analysis, available in GIS technology, to create maps that
identify locations with a high likelihood of criminal activity (Olajuyigbe et al, 2016).

Murray et al. (2001) highlight the significant advantage of geographic information

systems (GIS) in integrating geographical data with other data types, allowing for various
interpretations, viewing, and interpretation of data through maps, globes, reports, and charts.
Mohammed et al. (2019) utilized GIS technology to create a map of crimes, examine areas
with high crime rates, and evaluate the spatial data of crimes. Additionally, other studies have
utilized GIS technology to pinpoint areas with high crime rates and examine patterns in space
and time. In a prior study, Wang (2012) emphasized the potential of GIS technology in
analyzing crime. Similarly, Ratcliffe (2005) emphasized the significance of mapping crime
hotspots to comprehend and tackle crime patterns. These studies indicate that GIS has
proven to be a valuable tool for mapping and analyzing crime.

Special Studies Proposal Template version 1.2.2023

 This study aims to identify and map the areas in Poblacion, Baybay City Leyte, where
theft crimes occur frequently from 2018 to 2023. By analyzing the data, researchers aim to
draw conclusions about the spatial patterns of theft crimes in the Poblacion areas. The study
will examine if specific locations have a higher incidence of theft crimes and explore the
potential factors contributing to the concentration of theft crimes in these hotspots. This
research will also contribute to a broader understanding of crime patterns while
demonstrating the effectiveness of GIS in crime analysis. The findings of this study can
provide valuable insights for future studies and policy-making decisions.

1.2 Rationale

In the year 2022, the Philippine National Police (PNP) announced that Mindanao had
the highest decrease in crime occurrences from July to December 2022, with a decline of
12.29 percent, while Luzon recorded a decrease of 7.35 percent. However, there was a 5.65
percent increase in crime incidents in the Visayas area over this period (Caliwan, 2023). In the
same year, about 13 thousand cases of theft, classified as an index crime, were reported in
the Philippines (Statista, 2023).

The Philippine National Development Plan (PNDP) 2023–2028, specifically in Chapter

13, highlights the need to maintain peace and security. Outcome 2 of Subchapter 13.1 of the
mentioned PNDP emphasizes the government's commitment to protecting the quality of life
by reinforcing interactions between the police and the community, increasing the police
presence in communities, and improving crime and emergency response systems. These
measures aim to enhance general security and safety, fostering a feeling of protection and
assurance among the general public.

This study aims to contribute directly to the objectives of the Philippine National
Development Plan (PNDP) by identifying and mapping crime hotspots, focusing on theft
crimes. Identifying crime hotspots allows law enforcement to strategically allocate limited
resources, focusing patrols and preventative efforts on areas with the highest crime rates
(Braga, 2019). This targeted approach can be significantly more effective than randomly
spreading resources across a city. The use of crime mapping and analysis is a valuable tool
that plays a crucial role in understanding, preventing, and reducing crime. Through the
visualization and analysis of crime data spatially, valuable insights are gained that go beyond
simple numbers and statistics.

The aim of this research is to lay the groundwork for future research projects
supporting the Sustainable Development Goals and the Philippine National Development Plan.
Additionally, this study seeks to showcase the significant impact of modern technology,
specifically the Geographic Information System (GIS). Geodetic engineering traditionally
focuses on analyzing, interpreting, and measuring spatial data on the Earth's surface. By
integrating GIS in crime mapping and analysis, this research adopts a unique interdisciplinary
approach that blends geodetic engineering principles with criminology and public safety.
Apart from improving public safety, several studies demonstrate how GIS can enhance
emergency response plans and crime-fighting methods effectively. The government can use
GIS to refine strategies and adopt data-oriented techniques that align with national

Special Studies Proposal Template version 1.2.2023

development objectives, ultimately increasing the overall standard of living for the people.

1.3 Statement of the problem

In 2023, the Philippine National Police (PNP) reported that theft, rape, and physical
injury have been the most prevalent crimes since the start of President Ferdinand "Bongbong"
Marcos Jr.'s administration. The PNP observed a 5.9 percent increase in theft events reported
from January 2021 to July 2022 compared to the same period in 2023 (Pinlac, 2023).

Research indicates that crime, especially theft, tends to occur in specific places called
"hotspots". The prevalence of theft in these areas creates a sense of insecurity among the
population, which can impede economic growth and development. Therefore, it is crucial to
address this issue to ensure the safety of the people and promote progress in the country.

The government acknowledges that it is necessary to address challenges related to

crime prevention and response. However, the current methods are not always effective and
may require more data-driven decisions. One of the issues faced by the City Police Office of
Baybay is a lack of data-driven decision-making. Most data stored by them are still in non-
digitized formats, which makes it difficult for them to access easily and analyze the data
efficiently. As a result, the police force is struggling to identify and respond to crime patterns
in the city. To enhance crime prevention and response, this study proposes utilizing
Geographic Information System (GIS) technology. However, realizing the full potential of this
technology requires creating and executing comprehensive data-driven plans.

1.4 Objectives of the study

Generally, this study aims to determine and analyze areas identified as theft crime
hotspots. Specifically, this study aims to:

1. Identify clusters and crime hotspots of theft crime in the Poblacion Area of Baybay
City, Leyte, using data from 2018 to 2023;
2. Generate a map to analyze the potential influence of road networks and police
presence on theft crime locations.
3. Examine the correlation between socio-economic factors and the prevalence of crime
in the identified hotspots; and
4. Suggest surveillance measures in areas where crimes are clustered.

1.5 Scope and limitation of the study

The purpose of this study is to locate the areas in Poblacion 1-23, Baybay City, Leyte
that are more susceptible to theft. Furthermore, this study will evaluate the distance between
these hotspots and road networks (minor and major roads), police stations, and security
outposts to analyze how these factors can influence theft crimes. In addition to the
geographical aspects, this study will also examine the socioeconomic factors that can
contribute to the higher incidence of theft in certain areas.

Special Studies Proposal Template version 1.2.2023

 However, it is essential to acknowledge that this study has inherent limitations. The
scope of this study covers only the Poblacion Areas of Baybay City, Leyte, which comprises
of the 23 urban zones. Moreover, this study focuses solely on thefts that are classified as
commercial and residential thefts, which refers to thefts that occur within establishments,
buildings, or houses. Due to the limited availability of data from the police station, this is the
only type of theft that is covered in this study. As a result, other forms of criminal activity or
types of theft that occur outside of these settings are not included in this study. Additionally,
criminal activity is constantly changing, so law enforcement agencies must adapt their
strategies to address new trends and behaviours.

The data for this study will be sourced from official crime reports. However, it is
important to note that such reports may not cover all theft crime incidences due to
underreporting or other contributing factors. Additionally, the study results may also be
influenced by the availability of data collected from different agencies.

2 REVIEW OF RELATED LITERATURE

2.1 Crime Mapping and Analysis

In recent years, crime mapping using GIS has evolved significantly, with advances in
analytical capabilities leading to the development of geospatial tools and software for crime
mapping and analysis (Hunt, 2019). GIS and Remote Sensing technologies can be valuable
tools for visualizing data, analyzing facts, and making informed decisions (Guta et al., 2012).
Traditional crime analysis often relies on aggregate statistics and reports, which, while
informative, lack the ability to visualize and analyze spatial patterns. This is where GIS
emerges as a powerful tool. The use of GIS software in crime mapping and analysis has
enabled law enforcement agencies to analyze and correlate data sources, creating a detailed
snapshot of crime incidents and related factors within a community or geographical area
(National Institute of Justice).

Crime mapping and analysis is being applied in the Philippines. Several studies have
been conducted on this topic. One study by Vio Jianu Mojica and A. C. (2017) focused on
using GIS to identify crime hotspots in CAMANAVA, which includes Caloocan, Malabon,
Navotas, and Valenzuela. The researchers used the MarkerClusterer algorithm to analyze
crime data from 2017 and successfully identified hotspots for various types of crime, such as
robbery, theft, and drug-related offenses. The study highlighted the potential of GIS-based
crime mapping to inform targeted police patrols and resource allocation to potentially reduce
crime in these areas.

Another study focused on developing a crime mapping prototype for the province of
Cavite. By analyzing crime data from 2014 to 2016, the researchers identified hotspots for
various types of theft-related offenses, including burglary and robbery. The prototype
demonstrated the feasibility of using crime mapping to inform crime prevention efforts at the
provincial level (Baculo, M. J. C., & Marzan, C. S., 2017).

Special Studies Proposal Template version 1.2.2023

 Similarly, a study focused on identifying high crime areas in Metro Manila using spatial
analysis techniques. The study identified significant spatial clustering of theft and robbery
incidents, particularly in areas with high population density and limited security infrastructure.
The findings emphasized the need for targeted interventions and improved security measures
in these hotspots (Trepanier, E., 2014).

2.2. Crime Hotspots Theories

Crime hot spot maps can offer practical and consistent guidance to law enforcement
agencies when informed by various crime theories, Eck (2005) emphasizes that maps without
theoretical underpinnings may give law enforcement officials inaccurate and potentially
misleading information. The book of Eck et.al (2005) delves into various theories related to
crime hotspots. These theories include Place theory, Street theory, Neighborhood theory, and
Repeat Victimization theory.

Place theories concentrate on specific locations, trying to understand why crimes

occur at particular street corners or addresses. By analyzing these events, these theories
identify patterns such as where burglaries happen and where they don't. Using maps with dots
to represent these hotspots, police actions like warrants targeting specific addresses align
closely with this level of precision.

Street theories shift the focus slightly upwards, examining crimes happening on small,
elongated areas such as streets or blocks. This level of analysis includes prostitution strolls,
which prompt questions such as which streets do prostitutes operate on. Streets, segments,
and highway sections become the units of analysis, represented on maps as lines. While
police action remains relatively precise, concentrated patrolling and traffic alterations become
more prevalent here compared to the pinpoint precision of place-level interventions.

Moving further up the scale, neighborhood theories tackle crime patterns across large
areas such as square blocks, communities, or census tracts. Questions such as which areas
are gang territories dominate at this level. Maps utilize two-dimensional shapes like polygons
to depict hotspots. Due to the larger areas involved, precise police action such as
concentrated patrolling becomes less effective. Instead, initiatives encouraging resident
involvement in collective action against crime and disorder gain traction. If offenders roam
freely within a large area, deterring them also occurs at this broader level.

Finally, repeat victimization theories, independent of geographic scale, explore why

certain individuals become targets for repeated offenses. Although applicable to all levels
(dots, lines, or polygons), not all repeat victimization patterns are directly visualized on maps.

2.3 Methods and Techniques in Identifying Crime Hotspots

ArcMap provides various tools to create heat maps, hot spots, and interpolation
surfaces. The appropriate tool to use depends on the input data and the desired output. The
tools available are listed in the table below (Environmental Systems Research Institute, n.d.):

Special Studies Proposal Template version 1.2.2023

 Type of
Toolbox Toolset Tools
Surface
● Kernel
Density
Density ● Line Density Heat Map
● Point
Density
● IDW
● Kriging
● Natural
Spatial Neighbor
Analyst ● Spline
● Spline with
Interpolation Barriers Interpolation
● Topo to
Raster
● Topo to
Raster by
File
● Trend
● IDW
● Kriging
● Natural
Neighbor
● Spline
● Spline with
Raster
3D Analyst Barriers Interpolation
Interpolation
● Topo to
Raster
● Topo to
Raster by
File
● Trend
● Diffusion
Interpolation
with Barriers
● Empirical
Bayesian
Geostatistical Kriging
Interpolation Interpolation
Analyst ● Global
Polynomial
Interpolation
● IDW
● Kernel
Interpolation

Special Studies Proposal Template version 1.2.2023

 With
Barriers
● Local
Polynomial
Interpolation
● Radial Basis
Functions
● Cluster and
Outlier
Analysis
(Anselin
Local
Moran's I)
● Hot Spot
Mapping Analysis
Spatial Hot Spot
Clusters (Getis-Ord
Statistics Gi*)
● Optimized
Hot Spot
Analysis
● Optimized
Outlier
Analysis

Table 1. ArcMap tools to create heat maps, hot spots, and interpolation surfaces

The Hot Spot methods in Crime analysis can be broadly classified into three
categories, namely: Spatial Analysis Methods, Interpolation Methods and Mapping Cluster
or Spatial Interpolation Methods (Chang, 2012).

2.3.1 Spatial Analysis

Spatial analysis is a process that involves examining the attributes, locations, and
relationships of various features in spatial data. This is done using overlay and other analytical
methods to gain helpful knowledge and answer specific questions (Ansari & Vale, 2014). By
exploring data from a geographic point of view, spatial analysis can help solve complex
location-based problems, detect, and quantify patterns, determine relationships, assess
trends, and make predictions and decisions (Environmental Systems Research Institute, n.d.).
Different techniques under this category are Kernel Density Estimation (KDE), Point Density,
and Line Density.

2.3.2 Interpolation

Interpolation is a method that involves using known values or sampled points to

estimate unknown values at other points. It is used to predict values for geographic point data,

Special Studies Proposal Template version 1.2.2023

such as rainfall, elevation, chemical concentrations, noise levels, and more. While
interpolation is typically performed as a raster operation, it can also be carried out as a vector
operation by using a TIN (Triangulated Irregular Networks) surface model (GIS Sources, n.d.)
There are various interpolation methods available, including Inverse Distance Weighted (IDW),
Kriging, Natural Neighbor, and Spline, each with its own strengths and weaknesses. The
choice of method depends on the type and quality of the data, as well as the intended use of
the interpolated surface (Environmental Systems Research Institute, n.d.)

Interpolation Techniques

Kriging, IDW, Natural Neighbor, and Spline are commonly used interpolation methods
in spatial analysis. Unlike other interpolation methods, kriging is a geostatistical method that
considers both the distance and spatial correlation of data points. It is suitable for datasets
with complex spatial structures or irregularly spaced observations because it incorporates
both local variations and broader trends. Kriging is also useful because it provides estimates
of uncertainty for the interpolated values, which can help assess the reliability of the results.
In contrast, IDW is a deterministic method that estimates cell values based solely on the
distance to the prediction location. Natural Neighbor and Spline are good for data with
complex spatial patterns, but may not provide the same level of accuracy as Kriging (Wu et.a.l,
2016).

Kriging

Kriging is a powerful statistical interpolation method co-developed by Georges

Matheron, a French geomathematician, and D.G. Krige, a South African mining engineer. The
primary use of Kriging was to analyze geochemical data to discover gold deposits, and it
proved successful in doing so.

The primary difference between ordinary interpolation in ArcView and Kriging is that
with Kriging, an initial structural, spatial analysis is carried out to evaluate the variance and
autocorrelation of the data. The results of this analysis are then used to select the
interpolation model and carry out the interpolation itself (Oliver, 1999)

Kriging is an estimation method associated with the acronym BLUE, which stands for
Best Linear Unbiased Estimator. It is a linear method since the estimated values are weighted
linear combinations of available data. It is unbiased because the mean of the error is zero, and
it aims to minimize the variance of the errors, making it the best method for optimal prediction.
Therefore, Kriging is synonymous with optimal prediction (Food and Agriculture Organization,
n.d.).

Co-kriging

A more advanced use of GIS technology involves merging crime incident maps with
other social and economic factors to examine the correlation between high crime
concentrations and spatial variation (Gisela and Johnson, 2001). This can be achieved by
applying the co-kriging technique to interpolate crime data onto a grid, and subsequently

Special Studies Proposal Template version 1.2.2023

combining this grid with other spatial datasets. The approach allows for a more in-depth
analysis of crime patterns, which can be useful for law enforcement agencies seeking to
understand the underlying causes of crime and develop effective crime prevention strategies
(Murray, et. al. 2001).

Co-kriging is a popular geostatistical method that combines data from multiple

correlated variables to enhance the accuracy of spatial predictions. (Goovaerts, 1997; Isaaks
and Srivastava, 1989). While kriging excels at estimating a single variable of interest based on
its spatial autocorrelation, cokriging leverages the additional knowledge provided by
secondary variables, often referred to as covariates (Chiles & Delfiner, 2009). This approach
can be particularly advantageous when the main variable is sparsely sampled, but there is a
strong spatial correlation between it and the abundant secondary variables.

Notably, cokriging boasts enhanced prediction accuracy for crime rates, pinpointing
high-risk susceptibility zones through spatial insights, and facilitating data integration among
relevant stakeholders. These benefits pave the way for targeted interventions and resource
allocation based on the root causes of crime, not just its surface patterns (Yang et al., 2020;
Cho et al., 2010).

Several studies have demonstrated the effectiveness of cokriging in various contexts,

such as urban planning and public safety. One such study by Kannan (n.d.) employed Co-
Kriging to analyze the spatial relationships between crime and socioeconomic factors. The
study demonstrated the applicability of this geostatistical method in analyzing and predicting
crime patterns and their relationship to sociodemographic and socioeconomic factors. Yang
et al. (2020) used cokriging to predict theft and robbery rates in China by incorporating
offender movement data as a secondary variable, achieving improved prediction accuracy
compared to traditional methods. Cho et al. (2010) applied cokriging to analyze the spatial
distribution of violent crimes in Los Angeles, considering multiple demographic and socio-
economic variables, identifying high-risk areas for targeted prevention effort. Similarly, Ansari
et al. (2017) employed cokriging to study crime patterns in Ajmer City, India, using socio-
economic factors like illiteracy and non-working population, providing insights for crime
prevention strategies.

2.3.3 Mapping Clusters or Spatial Interpolation Methods

Mapping Cluster also known as Spatial Autocorrelation is an amount of the degree to

which a set of spatial features and the data values associated with it. It can be clustered
together in Space (positive spatial autocorrelation) or Scatter Widely (negative spatial
autocorrelation) (Environmental Systems Research Institute, n.d.). Various methods under
Spatial Autocorrelation, such as Anselin Local Moran's I and Getis-Ord Gi*, provide tools to
assess the spatial patterns inherent in the data.

Moran's I and Getis-Ord Gi* are used to detect clustering of similar values and separate
clusters of high and low values, respectively. These methods play a crucial role in
understanding crime patterns and developing effective strategies for crime prevention.

Special Studies Proposal Template version 1.2.2023

 In a comprehensive study by Ansari & Vale (2014), the precision of specific methods
within Spatial Analysis, Interpolation, and Spatial Autocorrelation was highlighted. The
research identified that Moran’s I can be used for both point and polygon fea-tures and can
detect the presence of the clustering of similar values. Moreover, the The Moran’s I and Gi*
both also calculate the z value for the analysis purpose

In essence, integrating Spatial Analysis, Interpolation, and Spatial Autocorrelation

methodologies provides a robust framework for extracting, predicting, and understanding
spatial information, contributing significantly to the field of geographic Analysis. The selection
of specific techniques, guided by studies such as that of Ansari and Vale (2014), ensures the
precision and relevance of the analytical processes applied to spatial data. Thus, it is essential
to use these methodologies effectively to derive meaningful insights from spatial datasets.
Tables below shows the different methods for Hot Spot Analysis.

Table 2. Different Methods for Crime Analysis (Ansari & Vale, 2020)

Table 3. Usefulness of the methods

2.4 Euclidean Distance

Euclidean Distance is a commonly used metric in spatial analysis as it measures the

straight-line distance between two points in an Euclidean space. It is often used to determine

Special Studies Proposal Template version 1.2.2023

the relationship between road networks and total crime in a specific area. This measure of
proximity is simple and intuitive, allowing researchers to assess the distance between crime
locations and road networks. This helps to understand how road networks influence crime
patterns, including the impact of accessibility and connectivity on criminal behavior
(Environmental Systems Research Institute, n.d.).

On the other hand, other distance measures such as Path Distance and Cost Distance
consider additional factors such as terrain, land cover, and travel impedance, making them
more suitable for analyzing the actual travel distance or cost between locations. However,
when assessing the spatial relationship between crime locations and road networks,
Euclidean distance is often a suitable and convenient measure.

Most geographical crime analysis applications use Euclidean distance to represent

spatial separation between points (Tompson et al., 2009). A study by Kannan et al. (n.d.) used
the Euclidean Distance method in order to determine if the road networks are associated with
the total crime in the city of Ajmer. The study revealed that maximum crime spots were
nearest minor and major roads in the Euclidean distance map. Additionally, densely populated
old city regions with narrow lanes and overcrowded commercial zones were found to be the
areas where significant crimes occurred. Similarly, a study by Mohammed et al. (2019),
highlights the use of Euclidean distance as the method for calculating the distance among all
crimes on the map. The study also focused on the variance to determine the optimal distance
between the points and achieve the best visualization of the fishnets grid. Unlike the weighted
interpolation method Inverse Distance used in another study, Euclidean distance was found
to be the most suitable measure for this analysis.

2.5 Socioeconomic Factors and Theft Crime

Numerous studies have explored the complex relationship between socio-economic

factors and crime rates, revealing several consistent associations. Studies conducted across
various countries (World Bank, 2006; Levitt, 2004) indicate strong correlations between low
income and unemployment rates with higher theft rates. Desperation and limited legitimate
opportunities are seen as driving factors behind such trends.

The "rational choice" perspective (Becker, 1968) argues that individuals engage in
crime when the perceived benefits outweigh the costs. In situations of economic hardship,
the potential gains from theft become relatively more attractive. High unemployment rates
are often linked to increased theft crimes (Raphael & Winter-Ebinger, 2001). Lack of jobs,
particularly for young men, can contribute to criminal activity as a means of income
generation. Lower levels of education are associated with higher crime rates, possibly due to
limited employment prospects and a weaker attachment to social norms (UNESCO, 2016).

Social disorganization theories (Shaw & McKay, 1942) highlight factors like residential
instability, community deterioration, and weak social control mechanisms as fostering
environments conducive to criminal activity. According to the Social Disorganization Theory,
some neighborhoods are unable to control deviant behavior due to frequent residential
turnover and outmigration which disrupts social networks responsible for maintaining social

Special Studies Proposal Template version 1.2.2023

control. The absence of such networks leads to higher levels of deviancy. Additional factors
such as poverty and racism have also been identified as contributing to this problem.

A study by Mitchell (2011) in Pittsburgh, Pennsylvania, using GIS to collect data on

crime occurrences and socioeconomic and demographic variables revealed significant
relationships between unemployment, household income, and violent and property crime
occurrences. Similarly, Kannan et al (n.d.) study highlighted the influence of
sociodemographic and economic factors on the likelihood of crimes in a region, particularly
the high correlation between crime, illiteracy, and non-working population. These studies show
that repeat victimization theories and crime patterns can be better understood through the
analysis of socioeconomic and demographic variables using GIS.

2.6 Weighted Overlay

Weighted Overlay (WO) in GIS unveils the most suitable areas for specific purposes by
integrating and analyzing multiple data layers (Longley et al., 2021). Each layer represents a
factor contributing to the suitability assessment, such as land cover, infrastructure, or
socioeconomic indicators (Bonham-Carter, 1994). The core of WO lies in assigning weights
to each layer, reflecting their relative importance in influencing the desired outcome (Eastman,
2012). Weights can be determined through diverse methods, including expert assessments,
statistical analysis, or data-driven approaches (Malczewski, 2023). Once weighted, the layers
are seamlessly combined using algorithms like weighted sum or product, ultimately
generating a suitability map highlighting areas with the highest potential for the chosen
purpose (Bonham-Carter, 1994). In the context of crime analysis, Weighted Overlay can be
used to integrate various factors such as crime history, socioeconomic indicators, and
infrastructure elements to identify high-risk areas and guide resource allocation and policing
efforts (Kannan, et.al., n.d.).

2 METHODOLOGY

2.2 Study area

Baybay is a coastal city in Leyte province in Eastern Visayas, Philippines. It spans a

land area of 46,050 hectares and consists of 92 barangays. It is comprised of twenty-three
(23) urban zones and sixty-nine (69) rural barangays. According to the 2021 census, the City's
total population is 125,347 individuals (City Profile - City of Baybay, Leyte, 2023).

The study area for this research is Poblacion 1-23, classified as the urban zones of
Baybay City, Leyte. The said area covers approximately 391.809 hectares.

Special Studies Proposal Template version 1.2.2023

 Figure 1. Study Area: Poblacion 1-23 Baybay City, Leyte

Baybay City has three police stations. The main station is located in Brgy. Gaas, near
Poblacion 23, with the coordinates 10°40'15.50"N, 124°48'52.52"E. The other two substations
are situated in Poblacion 15 and Brgy. Guadalupe, with the coordinates 10°40'20.35"N,
124°48'1.37"E and 10°44'32.30"N, 124°47'43.57"E, respectively.

Special Studies Proposal Template version 1.2.2023

 Figure 2. Baybay City Police Stations

The Poblacion areas are the central and most populated regions of a city. In Baybay
City, the Poblacion 1-23 barangays form the urban core, accommodating a significant portion
of the City's population and commercial activities. These areas have a mix of residential,
commercial, and institutional establishments, making them crucial hubs for economic and
social activities in the City. With no previous studies conducted in Baybay, initiating such
research in the City becomes even more imperative. Undertaking this study locally holds
significance as its outcomes benefit the residents and law enforcement agencies in Baybay.

2.3 General Methodology

Figure 3 outlines the methodology for this study. Each method will be further discussed in this
chapter.

Special Studies Proposal Template version 1.2.2023

 Spatial Data Non-spatial Data

Police Station Crime Type of Crime Socioeconomic

and security Location Data
outposts (Primary Overlay
locations Data)
Theft

Data
Preprocessing and
Thematic Mapping

Spatial Analysis of
Crime Patterns

Hotspot Cluster Euclidean Interpolati

Analysis Analysis Distance on (co-
of Crime Kriging)
from
Police
Stations
and
Roads

AHP

Results and Suggestions

Figure 3. Research Methodology Adopted in the Study

Special Studies Proposal Template version 1.2.2023

 2.3.1 Data Collection

Primary Data

The primary data in this study will be the geographic locations of crime incidences,
police stations, and security outposts. The researcher will gather the exact geographic
location of crime through firsthand observation using GPS in the field. This is necessary
because the data from Baybay Police Station does not contain the exact geographic location
of the crime but only the description of where the crime occurred.

Secondary Data

This study will utilize secondary data such as socioeconomic data and road networks.
The socioeconomic data will be sourced from the Barangay or the Local Government Unit of
Baybay. This dataset will contain variables such as population, employment status (number
of working and non-working population), education (number of literate and illiterate
population), housing, and other available socioeconomic data. Meanwhile, the road networks
will be obtained from OpenStreetMap.

2.3.2 Mapping Technique

This study will use thematic mapping technique in order to visualize the analysis in
this study. Thematic mapping is a powerful analytical tool that is used to study the spatial
patterns of crime point data. It involves the application of different colors, symbols, or other
visual cues to represent various types of crime incidents. Through this technique, researcher
can obtain a better understanding of the geographic distribution of crime activity and identify
any patterns or trends that may exist.

The process of thematic mapping can also help reveal any spatial hotspots where
crime is more prevalent. Furthermore, thematic mapping can assist in identifying any
correlations or patterns between different types of crimes, which can provide valuable insights
into the underlying causes of criminal behavior.

2.3.3 Preprocessing

Data Projection

Data projection refers to the process of transforming the spatial reference system
(SRS) or coordinate reference system (CRS) of a vector dataset. It is crucial to ensure that all
layers in a project are in the same coordinate reference system for accurate spatial analysis
and visualization

Special Studies Proposal Template version 1.2.2023

 2.3.4 Data Processing

Euclidean Distance

In distance analysis, Euclidean distance is calculated from the center of the source
cell to the center of the surrounding cells. The Euclidean algorithm works by calculating the
hypotenuse of the triangle with x_max and y_max as the other two legs of the triangle to
determine the distance of each cell to each source cell. This calculation provides the true
Euclidean distance, not the cell distance. The output raster assigns the value of the shortest
distance to a source cell to the cell location if it is less than the maximum distance specified.

Figure 4. Euclidean Distance Calculation

The tool for Euclidean distance raster produces floating-point distance values. If a cell
is equidistant to two or more sources, then the cell is assigned to the first source encountered
during the scanning process. The actual algorithm uses a two-scan sequential process, which
makes the tool's speed independent from the number and distribution of the source cells and
the maximum distance specified. The size of the raster is the only factor that affects the speed
of the tool, and the computation time is directly proportional to the number of cells in the
Analysis window.

In this study, the Euclidean Distance tool in ArcGIS will be used to calculate the
distance between crime locations and the nearest major and minor roads and police stations
and security outposts. Previous researches has shown that crime locations are often
associated with the road network. The Euclidean Distance tool will be used to determine the
proximity of the crime locations to the road network by calculating the distance for each cell.

Special Studies Proposal Template version 1.2.2023

 Figure 5. Euclidean Distance in ArcGIS

The input raster or feature source data for crime data analysis will be the Road
Networks (Major Roads) and the Police Stations and Security Outposts.

To determine the proximity of crime locations to major roads, it is necessary to input

the maximum distance. This maximum distance will serve as the upper limit of the
classification. Based on this input, the class interval will be calculated. The distances will then
be classified as Nearest, Near, Near Medium, Medium, Far Medium, Far and Farthest.

Similarly, the same method will be used to determine the proximity of crime locations
to police stations and security outposts. The maximum distance between police stations will
be inputted as the maximum distance to classify the distances between crime locations and
police stations/outposts

Hotspots Analysis and Cluster Analysis (Anselin Local Moran’s I)

Hotspot analysis is a process used to identify clusters of high or low values of a

particular variable within a given geographic area. The process involves obtaining the data to
be analyzed, checking for spatial clustering using Moran's I statistic, aggregating the data to
create a count of incidents within a given range, running a Cluster and Outlier Analysis (Anselin
Local Morans I), and visualizing the results using a map.

Cluster and Outlier Analysis (Anselin Local Moran's I) is a spatial statistical technique
used to identify areas where crime rates are either unexpectedly high (hotspots) or low
(coldspots), alongside isolated instances of significant deviation from surrounding trends
(outliers). This analysis relies on the Anselin Local Moran's I statistic, a measure of spatial
autocorrelation.

The "local Moran's I" tool is used to determine whether a feature (such as a point or
polygon) is part of a cluster or an outlier. A positive value for "I" means that the feature has
neighbors with similar attribute values, indicating that it is part of a cluster. A negative value
for "I" means that the feature has neighbors with dissimilar values, indicating that it is an

Special Studies Proposal Template version 1.2.2023

outlier. However, the p-value for the feature must be small enough to be considered
statistically significant. The output feature class shows the local Moran's I index, z-score, p-
value, and COType for each feature. The COType field distinguishes between a statistically
significant cluster of high values (HH), cluster of low values (LL), outlier with a high value
surrounded by low values (HL), and outlier with a low value surrounded by high values (LH).
When the tool is run, the output feature class is automatically added to the table of contents
(TOC) with default rendering applied to the COType field.

Figure 6. Cluster and Outlier Analysis (Anselin Local Moran's I) in ArcGIS

Interpolation

Kriging is a widely used interpolation method in spatial analysis. It estimates values at

unobserved locations based on the values observed at nearby locations, taking into account
the spatial correlation between data points. This creates a more accurate representation of
the entire study area. In this study, Kriging will be used because it is considered the Best Linear
Unbiased Estimator (BLUE). This means that it provides the most unbiased and efficient
estimates compared to other interpolation methods. If the Kriging results indicate a strong
correlation between total crime and other variables, Co-kriging can be employed to include
additional variables in the prediction model.

To determine the output value for each location, the Kriging tool fits a mathematical
function to a specified number of points or all points within a specified radius. Kriging involves
several steps, including exploratory statistical analysis of the data, variogram modeling,
creating the surface, and (optionally) exploring a variance surface. Kriging is suitable when
there is a spatially correlated distance or directional bias in the data.

Special Studies Proposal Template version 1.2.2023

 Figure 7. Kriging Formula

Figure 7. Kriging in ArcGIS

The input point feature will be the Crime Location points and the z value filed will be
the GZfiled computed in the Spatial Autocorrelation.

Special Studies Proposal Template version 1.2.2023

 Co-Kriging

Co-kriging is an extension of the Kriging technique that allows for the estimation of
two or more related variables at unsampled locations. To accomplish this, it is necessary to
estimate the autocorrelation for each variable and all cross-correlations. This process can add
more variability to the process, which means that ordinary Kriging will be used first to estimate
the autocorrelation for each variable individually. Once these autocorrelations are determined,
the cross-correlations between the variables can be estimated, making it possible to predict
values at unsampled locations more accurately.

Ordinary cokriging assumes the following two models:

The primary focus is on a variable called Z1. To improve predictions, autocorrelation

for Z1 and cross-correlations between Z1 and other variable types are taken into account.
Ordinary cokriging uses two models and considers two types of random errors- ε1(s) and
ε2(s). Each of these has autocorrelation and cross-correlation between them. Ordinary
cokriging tries to make better predictions for Z1(s0) using information from Z2(s) as well. In
this method, Z1 and Z2 are found to be autocorrelated. Also, when Z1 is below its mean µ1,
Z2 is often above its mean µ2, and vice versa. Therefore, Z1 and Z2 have negative cross-
correlation. Although each location s can have both Z1(s) and Z2(s), it is not necessary. Both
autocorrelation and cross-correlation are used to make better predictions for the main
variable of interest, Z1.

For this particular study, co-kriging will be utilized to generate a map that displays
regions prone to theft crimes. Co-kriging is employed to evaluate the autocorrelation among
crime locations and socioeconomic indicators. The cross-correlation between these variables
can provide a deeper understanding of the underlying factors that contribute to the occurrence
of crime.

Weighted Overlay

To ensure more accurate identification of crime hotspots, a combination of different

factors will be considered, such as distance to road networks, police presence, and socio-
economic data. The Weighted Overlay Analysis will be used in determining the hotspots
considering these different factors.

Special Studies Proposal Template version 1.2.2023

 The weights in the Overlay Analysis will be based on the ranking of the most correlated
factor to crime locations. To do this, the previous steps in co-kriging will be used. This will
help to produce a map with crime-susceptible regions based on socioeconomic data. Then,
using Band Collection Statistics, the different maps generated will be used to produce a
correlation matrix between the crime locations and each socioeconomic data. After the
correlation is established, the ranks based on correlation will be determined.

Once the weighting of the factors is done, the Weighted Overlay tool in ArcGIS will be
used to overlay the results of the various variables and generate a composite hotspot map.
This map will provide a comprehensive view of the areas with the highest concentration of
hotspots based on the combined influence of all the variables considered.

2.4 Timetable/work schedule/Gannt chart

3 EXPECTED OUTPUTS

1. Euclidean Distance Map of Crime Locations and Road Networks and Police/Security
Outposts - This thematic map will illustrate the distances between theft crime
locations and road networks & police presence, employing the Euclidean distance
measure. It aims to ascertain potential correlations between the proximity of theft
crime locations and the accessibility of these road networks and police presence.
2. Crime Hotspots Map - This map will exclusively highlight crime hotspots, disregarding
other factors and focusing solely on crime locations. It will provide a clear visualization
of areas with the highest concentration of crime incidents in the study area.
3. Table of Correlation Coefficient between Crime Volume and Non-Spatial Socio-
Economic Data - This table will illustrate the correlation between theft crime volume
and various socio-economic factors, including income levels, education levels, and
unemployment rates. By analyzing the correlation coefficients, it will be possible to
determine if there are any significant relationships between these variables and the
occurrence of theft crimes.
4. Susceptibility Map of Theft Crime - This map will delineate areas most susceptible to
theft crime based on socio-economic factors such as population, employment status,
and education.

Special Studies Proposal Template version 1.2.2023

 5. Comprehensive Crime Hotspots Map considering Euclidean Distance and Socio-
Economic Data - This map will identify areas with the highest concentration of theft
crimes by considering not only socio-economic factors but also the proximity to roads
using the Euclidean distance measure. By incorporating these additional factors, the
study aims to provide a more comprehensive understanding of theft crime patterns
and potential causes in different areas.

4 REMARKS

Adviser note (please check one):

Approved for proposal defense

Schedule: ______________________

Needs revision/not approved

5 REFERENCES

Ansari, S. M., & Kale, K. V. (2013). Methods for Crime Analysis Using GIS. International Journal
of Scientific & Engineering Research. https://www.ijser.org/researchpaper/Methods-
for-Crime-Analysis-Using-GIS.pdf

Ansari, S. M., & Vale, K. V. (2014). Mapping and Analysis of Crime in Aurangabad City using
GIS (pp. 67-76). IOSR Journal of Computer Engineering. https://doi.org/10.9790/0661-
16476776

Argun, Uğur, & Dağlar, Murat. (2016). Crime Mapping and Geographical Information Systems
in Crime Analysis. Journal of Human Sciences, 13(8), 2208.
https://doi.org/10.14687/ijhs.v13i1.3736

Baculo, M. J. C., & Marzan, C. S. (2017). Crime Mapping Approach for Crime Pattern
Identification: A Prototype for the Province of Cavite. Asia Pacific Journal of
Multidisciplinary Research, 5(8), 7-19
https://www.researchgate.net/publication/354659695_Crime_Mapping_Approach_fo
r_Crime_Pattern_Identification_A_Prototype_for_the_Province_of_Cavite

Special Studies Proposal Template version 1.2.2023

Barangays - City of Baybay, Leyte. (2023, November 16). City of Baybay, Leyte. Retrieved
December 8, 2023, from https://baybaycity.gov.ph/government/barangays/

Baybay, Province of Leyte. (n.d.). Retrieved from

https://www.philatlas.com/visayas/r08/leyte/baybay.html

Bernard, T. J. (2023, December 23). theft. Encyclopedia Britannica.

https://www.britannica.com/topic/theft

Braga, A. A. (2019). Hot Spots Policing of Small Geographic Areas Effects on Crime. Campbell
Systematic Reviews, 12(1), 1-45. doi:10.1002/cl2.1046

Britannica, T. Editors of Encyclopaedia (2023, December 29). burglary. Encyclopedia

Britannica. https://www.britannica.com/topic/burglary

Caliwan. (2023, January 9). PH Crime Rate Drops by 6.37% in 2nd Half of 2022: PNP. Philippine
News Agency. Retrieved December 8, 2023,from
https://www.pna.gov.ph/articles/1192262#:~:text=However%2C%20the%20Visayas
%20logged%20a,prevalent%20crimes%2C%E2%80%9D%20Azurin%20said

City Profile - City of Baybay, Leyte. (2023, November 15). City of Baybay, Leyte.
RetrievedDecember 8, 2023, from https://baybaycity.gov.ph/about-city/about/

Divya, G., Robinson, R., & Selvan, K. (2014). Suitability of Clustering Algorithms for Crime
Hotspot Analysis (pp. 231-234). International Journal of Computer Science
Engineering and Technology.
https://ijcset.net/docs/Volumes/volume4issue7/ijcset2014040704.pdf

F. Mohammed, A., & R. Baiee, W. (2020). IOP Conference Series: Materials Science and
Engineering, 928(3), 032071. doi:10.1088/1757-899x/928/3/032071

Fredrick, V. & Lazarus, V. & Emmanuel, M. & Bwala, I. (2023). Geospatial Appraisal of Crime
Hotspot Distribution in Bauchi Metropolis. Journal of Geographic Information System.
12. 43–50. 10.5923/j.ajgis.20231201.03.

Geospatial World. (2016). Crime analysis: GIS: A gateway to safe city. Retrieved from
https://www.geospatialworld.net/article/crime-analysis-gis-a-gateway-to-safe-city/

Gupta, R. (2012, February 5). Crime analysis: GIS: A Gateway to Safe City. Geospatial World.
https://www.geospatialworld.net/article/crime-analysis-gis-a-gateway-to-safe-city/

Guta R., Rajitha K., Basu S. and Mittal S.; Application of GIS in Crime Analysis: A Gateway to
Safe City, India Geospatial Forum, 2012.

Highlights of the Region VIII (Eastern Visayas) Population 2020 Census of Population and

Special Studies Proposal Template version 1.2.2023

 Housing (2020 CPH) | Philippine Statistics Authority | Republic of the Philippines.
(2021, August 23). Retrieved December 7, 2023from
https://psa.gov.ph/content/highlights-region-viii-eastern-visayas-population-
2020census-population-and-
housing2020#:~:text=The%20City%20of%20Tacloban%2C%20the,a%20population%2
0of%2025%2C881%20persons.

Hunt. (2019, July 10). From Crime Mapping to Crime Forecasting: The Evolution of Place-
Based Policing | National Institute of Justice. National Institute of Justice. Retrieved
December 11, 2023, from https://nij.ojp.gov/topics/articles/crime-mapping-crime-
forecasting-evolution-place-based-policing

Kannan, M., Rao, K. V., Raj, M., & Verma, A. Crime Mapping and Hot Spot Analysis using
Geospacial Techniques: A Case of Ajmer (p. 10). https://a-a-r-
s.org/proceeding/ACRS2017/ID_753_1652/999.pdf

Kan-tsung Chang; Introduction to Geographic Information System (4th Edition, Tata McGraw-
Hill, Eleventh Reprint 2012).

Mapping Crime: Understanding Hot Spots. (2005). In U.S Department of Justice, Office
ofJustice Programs (NCJ 209393). Retrieved December 3, 2023, from
https://www.ojp.gov/pdffiles1/nij/209393.pdf

MDPI. (2023). IJGI | Free Full-Text | Spatiotemporal Distribution and Influencing Factors of
Theft during the Pre-COVID-19 and COVID-19 Periods: A Case Study of Haining City,
Zhejiang, China. Retrieved January 13, 2024, from https://www.mdpi.com/2220-
9964/12/5/189

Mitchell, S. E. (2010). Using GIS to Explore the Relationship between Socioeconomic Status
and Demographic Variables and Crime in Pittsburgh, Pennsylvania. Semantic Scholar.
https://gis.smumn.edu/GradProjects/MitchellS.pdf

Mojica, V. J. (2017). Spatial Analysis of Violent Crimes in Metro Manila, Philippines.

International Journal of Geo-Information, 6(8), 140.
https://www.tandfonline.com/doi/abs/10.1080/01924036.2017.1398669

Murray, A. T., McGuffog, I., Western, J. S., & Mullins, P. (2001). Exploratory spatial data analysis
techniques for examining urban crime: Implications for evaluating treatment. British
Journal of criminology, 41(2), 309- 329.

Patel, K., P. Thakkar, L. Patel and P. Chandresh (2014). GIS-based decision support system for
crime mapping, analysis, and identifying hotspots in Ahmedabad city. International
Journal of Modern Engineering Research Vol. 4 (1), 32 – 35.

Philippines: Number of Focus Crimes by Type 2022 | Statista. (2023, September 19). Statista.
Retrieved December 8, 2023, from

Special Studies Proposal Template version 1.2.2023

 https://www.statista.com/statistics/1171040/philippines-number-of-index-
crimesbytype/#:~:text=In%202022%2C%20about%2013%20thousand,main%20types
%20of%20sch%20crimes.

Pinlac, B. (2023, January 10). PNP: Theft, Rape, and Physical Injury Most Common Crimes
Since Marcos Rule | Inquirer News. INQUIRER.net. Retrieved December 8, 2023, from
https://newsinfo.inquirer.net/1714254/fwd-pnp-theft-rape-physical-injury-most-
prevalent-crimes-since-start-of-marcos-administration#:~:text=Nation-
,PNP%3A%20Theft%2C%20rape%2C%20and%20physical%20injury%20most,common
%20crimes%20since%20Marcos%20rule&text=MANILA%2C%20Philippines%20%E2%
80%93%20Theft%2C%20rape,Philippine%20National%20Police%20(PNP).

Ratcliffe, J. H. (2005). Mapping Crime: Understanding Hot Spots. National Institute of Justice
Journal, 253, 10-23. Retrieved from https://www.ojp.gov/pdffiles1/nij/209393.pdf

Rubio, A. J. M., Ballera, M. A., & Gonzalez, D. B. (2018). Identifying Crime Hotspots in
CAMANAVA by Geographic Information System using Spatio-Temporal Analysis (pp.
82-90). Asia Pacific Journal of Multidisciplinary Research. http://www.apjmr.com/wp-
content/uploads/2018/07/APJMR-2018.6.3.11a.pdf

Santos, A. D., et al. (2018). Identifying Crime Hotspots in CAMANAVA by Geographic

Information System using Spatio-Temporal Analysis. Asia Pacific Journal of
Multidisciplinary Research, 6(4), 1-10. http://www.apjmr.com/wp-
content/uploads/2018/07/APJMR-2018.6.3.11a.pdf

Sharma, Ravi & Palria, Sarvesh & Bhalla, P. (2016). Crime Mapping & Analysis of Ajmer City-A
GIS Approach. Journal of Geomatics. 10. 96-101

Wang, F. (2012). Annals of GIS, 18(3), 159–171. doi:10.1080/19475683.2012.691900

Williamson, D., S. McLafferty, V. Goldsmith, J. Mallenkopf, and P. McGuire. 1999. “A Better

Method to Smooth Crime Incident Data.” Environmental Systems Research Institute
ArcUser Magazine, January–March, 1–5. https://www.Environmental Systems
Research Institute.com/news/arcuser/0199/crimedata.html

Special Studies Proposal Template version 1.2.2023