Crime is one of the most significant threats to the mankind. Crime has always caused problems to people, and in the modern world, this issue becomes dangerous like never before. During the primitive days of human civilization, crime was a simple issue, and it wasn’t organized. Today it’s organized and much more complex. Therefore, this problem in the context of human existence developed throughout the history. According to Uche (2008) and Financial (2011), there is one of the worlds highest crime rates in Nigeria. Lagos noted that the increase in poverty caused a significant growth in numbers of armed robbery shoplifting, pickpockets, etc. At least twelve people were killed in attacks on police stations and banks in 2011 (Nossiter, 2011).
Maritz (2010) noted that Nigeria is no different than other countries regarding a percentage of crimes in metropolitan areas, where these figures are always higher. The government implements various solutions aimed to control this situation and prevent new crimes, but there are many complex problems. Every time when an offender is unable to commit a crime, he or she still has various types of displacement available. Innovative methods and modern scientific solutions provide possible solutions to such a problem. According to Kpedekpo and Arya (1981), and Printcom (2003), multivariate statistics helped in developing of various criminological explanations.
Another useful tool of crime analysis is PCA. This abbreviation stands for Principal Component Analysis. It allows criminalists to estimate the overall level of criminality in a particular area. This type of analysis helps in reducing the dimensionality of a big number of variables that are interrelated. At the same time, it retains as much of the variation as possible. The problem of the computation reduced to an eigenvalue-eigenvector problem. The researchers may want to turn numerous variables into a smaller amount of super variables, therefore creating a group of measures (Jolliffe, 2002). This problem appears when one tries to determine the relationship between crime incidents and various socio-economic factors. PCA develops a relatively small set of components using the correlation between variables. This set of components summarizes a correlation. PCA was used in a study that examined the relationship between socio-economic status and crime. The study was carried out in Ottawa and Saskatoon. A huge set of variables have been replaced with a small number of elements, which represent inter-correlated variables as close to the original data, as possible (Exp, 2008). The overall criminality can also be determined using principal component analysis. The first eigenvector shows relatively equal loadings of all variables, which allows PC to measure the overall rate of crimes. Being applied during the analysis of the crime data for 1997 US, this value was obtained from the first PC. Hardle and Zdenek (2007) have got the same results for the 1985 year. Along with the first PC, there is a second PC, which classifies crimes by types. Usman et al. (2012) carried out a research that showed how three Principal Components were obtained from seven components present at the beginning. This system included the Loading plot and the Scree plot, which indicated a correlation between crimes against property, and another category of crimes – crimes against persons. In Los Angeles, Yan Fang (2011) successfully used Component Analysis to extract five Principal Components from the original fifteen components. The variance of the original dataset equals 85%, which means that most of the information is retained. Another interesting source devoted to this topic is the study that has been carried out in Katrina State. It describes eight common crimes reported from 2006 to 2008. The list of crimes includes auto theft, rape, murder, robbery, store breakings, assault, and grievous hurt. Principal Component analysis and the correlation matrix helped explain the correlation between these kinds of crimes, illustrating the distribution of crimes in various areas.
Classification of Crime
In every country, there is a particular system of classification. In the US, the Federal Bureau of Investigation collects annual data on crimes in a form of reports (Uniform Crime Reports, UCR). All the crimes are classified according to the common law. Part 1 crimes (also called Index 1 crimes) are categorized again and classified either as violent crimes, or crimes against property. Violent crimes include homicide, rape, assault, and robbery. In turn, property crimes are theft, burglary, vehicle theft, and arson. Other crimes are classified as Part 2 crimes. The Police of Nigeria also sorts crimes depending on what law is violated. Nigeria Police Abstract of Statistics includes following categories of violation of laws:
- Offenses against persons: grievous hurt, manslaughter, child stealing, murder and attempted murder, rape, and assault.
- Property crimes: store and housebreakings, theft, armed robbery, forgery.
- Crimes against lawful authority: gambling, corruption, forgery of current notes, breach of peace.
- Offenses against local act: liquor offenses, traffic offenses, etc.
Causes of Crimes
Usually, criminal behavior is caused by a number of different factors, since human behavior is determined by many cultural, social, environmental, and psychological factors. Different types of people commit different crimes, regardless of a place, time, or other circumstances (Danbazau, 2007). We will take a look at some most common causes:
Biogenetic factors. Criminologists express an opinion that criminal activity may be caused biologically (Pratt and Cullen, 2000). Lombrose (1911) states that people born, not become criminals, and notes that criminal behavior is dictated by a certain type of genetic constitution.
Social and environmental factors (Sutherland, 1939). Environment plays a very important role in human behavior. Criminal behavior is often increased in conditions of poverty, corruption, poor education, unemployment, child abuse, etc. According to Oyebanji (1982) and Akpan (2002), the current problem with crimes in Nigeria is caused mostly by industrialization, urbanization, and lack of education. Most Nigerians agree with Kutigi (2008), that ignorance and poverty are the key factors that cause the growth in crime rates in Nigeria. Ayoola (2008) also notes other factors, such as accountability in the public fund management. This factor is of extreme importance considering government structures that have been identified as the reason of the endemic corruption. Corruption is now present in all spheres of the Nigerian society.
The Nigerian Police
The police are the most important part of criminal justice. During the past three decades, the society hasn’t been satisfied with the work done by the police (Danbazau, 2007). Why is the police unable to prevent crimes? Most of all, such a problem is determined by professionalism, equipment, and inadequate manpower, as well as by low approval ratings (Okeroko, 1993), and corruption (Al-Ghazali, 2004).
Statistics of Crimes
There is one of the worlds highest crime rates in Nigeria. Murders are often committed along with burglaries. Rich people take care of their security themselves, in some states the Police have a right to shoot criminals “on sight” (Financial Times, 2009). Starting from the 1980s, crime rates grew so fast that Nigeria faced the real epidemic of crimes, especially in urban regions. This process was characterized by a number of negative side-effects, such as inadequate government service, social disorganization, inequality, etc. (Nigeria, 1991). Annual crime rates amounted to 200 for populations of 100,000, and starting from the 1960s, they increased up to 300. In 1981, the number of reported crimes rose up to 330,000 and increased to 355,000 in 1985. The number of robberies grew from 1,937 in 1990 to 2,419 in 1991. Crimes like assault were often reported along with crimes of other categories. According to Cleen (1993), these figured declined starting from 1994, with a significant decline to the level of 162,000 which have been registered in 2006.
Theory of Principal Component Analysis
It’s hard to decipher patterns of association with a big number of variables. Sometimes variables may repeat. Principal Component Analysis reduces the dimensionality of the data, therefore simplifying the whole process. Most of the information obtained from the data set is explained by the primary variables.
In other words, this type of analysis reduces the number of variables by taking linear combinations. Thus, principal components may be used to obtain information about the whole data set, using correlations among the principal components, and considering original variables. Loadings matrices or matrices of correlation show principal components that are associated with each variable. The linear combination of the highest variance determines the first principal component.
Given the principal component, we can estimate how it provides a necessary amount of information, by checking the data set. The linear combination that has the highest variance is designated as the second principal component after the first principal component is removed, in case there is no adequate information. This process is repeated until the necessary amount of information is collected. Each component defines the dimension, and then the process continues taking into account these dimensions.
Components which should be utilized are determined by correlations between the original variables, and the principal components. Such a decision-making process can be simplified by using a scree plot. This is a graph of the variance where each principal component is displayed in a descending order. Below the point designated as an “elbow”, the graph becomes more horizontal, and all the components below are discarded. The label of each particular component is determined by the original variables that show a high correlation level with each component.Date of public: January 10, 2018 Category: