Global Journal of Social Sciences Studies

Volume 2, Number 3 (2016) pp 131-143 doi 10.20448/807.2.3.131.143 | Research Articles

 

Analysis of Criminal Cases in Adamawa State, Nigeria

Akinrefon A. A 1Adeniyi O.I 2 Adejumo A. O 2 Olawale A. O 4Ubong B.A 5
1 Department of Statistics and Operations Research, Modibbo Adama University of Technology, Yola
2 Department of Statistics, University of Ilorin, Kwara State
4 Department of Statistics, Osun State Polytechnic, Iree, Nigeria
5 Department of Statistics and Operations Research Modibbo Adama University of Technology, Yola

Abstract

The Occurrence of Criminal act worldwide is on the increase, curbing the menace of Crime effect to humanity has been a thing of concern, Adamawa state-Nigeria not excluded. The results show that Crime rate is on the increase and that the Age group that most involved in criminal act is between the Ages 16 – 35. Crime incidence in the state varies over the years given the age-group (i.e. Crime depends on year of occurrence, so also the criminals’ age group). Further analysis show that the all possible two-way interactions between crime, year of occurrence and gender (CY.CG.YG) in the first case is most appropriate for assessing crime cases in the state since it has the least AIC value for the first analysis while the model of association between crime cases and age-group independent of gender (CA.G) has the least AIC, and thus, is accepted to be the best model.

Keywords: Crime rates, Log-linear, Criminal cases, Akaike Information criterion (AIC),Crime incidence.

DOI: 10.20448/807.2.3.131.143

Citation | Akinrefon A.A; Adeniyi O.I; Adejumo A.O; Olawale A.O; Ubong B.A (2016). Analysis of Criminal Cases in Adamawa State, Nigeria. Global Journal of Social Sciences Studies, 2(3): 131-143.

Copyright: This work is licensed under a Creative Commons Attribution 3.0 License

Funding : This study received no specific financial support.

Competing Interests: The authors declare that they have no competing interests.

History : Received: 8 June 2016/ / Revised: 2 July 2016 Accepted: 16 September 2016/ Published: 29 October 2016

Publisher: Online Science Publishing

1. Introduction

Nigeria has one of the alarming crime rates in the world Financial, 2011 In Yola, Adamawa state, crime is challenging social vice. Crimes like armed robbery, breaking of homes and industries, rapping, arson, killing, and so on, are the events of the now. Others also include; burglary, pick pocket, destruction of properties, looting, assassination, examination malpractices and mostly recently, the insurgence of bombing and kidnapping by Boko Haram in North-Eastern parts of Nigeria of which Adamawa state is inclusive. The negative effect of these include death of the inhabitants of Adamawa state, increase in the rate of internally displaced persons, living in serious tension, loss of businesses and investment opportunities, and farm products etc. The question is: why is crime getting worse on a daily basis?

Crime is committed by persons when both psychological and social reasoning combine to create such intense feeling as to throw away any hope they have left and release their anger in a devastating way. Some causes of crime are; lack of parental care or neglect, needs, social environment, peer group influence, spiritual undertone etc. But the reason behind crime varies on the background of the mental disorder, traumatic events, individual or group interest etc. Every individual differs in looks, minds, personality and ideas. And this makes it impossible to detect any two or more persons, and also difficult to know who is a criminal. Other reasons may include high levels of unemployment, low labor utilization, high level income and asset inequality, high levels of poverty, immigration etc.

There are evidences, according to Farrington that children from broken homes may not get proper monitoring and that those from poor homes may not get proper feeding and get necessary provisions for day to day living including school needs and hence they tend to find alternative means to provide these necessities, as the result, the alternative means to them are to involve in criminal offences. Ahmed (2000 observed that crime results from economic structure of the society, he also opined that the gap between the rich and the poor in any society creates crime mostly against property. By this, one needs to re-orientate the society against the syndrome of materialism.

Adamawa State has witnessed a substantial increase on crime rate in recent times. Crime has affected humanity and properties greatly; such as murder, robbery, kidnapping, arson, rape, killing, fraud and what have you. Incessant murdering and bombing by Boko Haram in Adamawa state and its environs is an important issues to monitor because unlike other crimes, the number of reported murders is in a very high rate compared to other crimes in Adamawa state.   

This study thus seeks to understand the prevalence of crime cases in Adamawa state via gender distribution, year of incidence, and age dependencies of criminals; and to fit a model that best describes association of these factors with crime statistics.

2. Methodology

Data for this research work covered crime situations in Adamawa state from 2010 to 2014 and the data used are data from Yola prison. These are secondary data extracted from the records of the Nigerian Prison, Yola, Adamawa state on criminal cases. A total number of 1672 criminal cases were recorded. The analysis were carried out using R package. The variables considered are Crime, Year and Gender, and Crime, Age group and Gender.

Ratio: A ratio is another statistical analysis method used to show any numerator-denominator relationship between two numbers e.g.

Crude Rates: The crude rates are obtained by dividing the number of events (i.e. criminal cases) of a specified type occurring within an interval (which is usually a year) by the size of the population within which the events occurred.

The mid-interval population is often taken as the best estimate for calculating crude rates, which involves calculating crude crime rate in a given area (i.e. Adamawa) in a given year by the ratio of crimes committed during a 12th month period to the population of the area within the same 12th month.

Specific Rates: Specific rates are similar to crude rates but are computed on the basis of events (i.e. criminal offences) in or to a particular sub-population (i.e. age).

Therefore,

The rate also is used to determine the crime by crime and the years (2010-2014), the gender and the age groups.

2.1. Log-Linear Models

Log-linear models are used to determine whether there are any significant relationships in multi-way contingency tables that have three or more categories variables and/ or to determine if the distribution of the counts among the cells of a table can be explained by a sampler underlying structure. The models specify how the expected count depends on levels of the categorical variables for that cell as well as associations and interactions among those variables. The purpose of log-linear modeling is the analysis of association and interaction patterns. The log-linear analysis assume that the response observations are counts having Poisson distributions (Lawal, 2003). The log-linear model is one of the specialized cases of generalized linear models (GLMs) for Poisson and multinomial distributed data. The variables investigated by log-linear models are all treated as “response variables”. In other words, no distinction is made between independent and dependent variables. If one or more variables are treated as explicitly dependent, others as independent, then logit or logistic regression should be used instead. Also, if the variables being investigated are continuous and cannot be broken down into discrete categories, logit or logistic regression would again be the appropriate analysis.

2.2. Statistical Independence

2.3. Independence Model

2.4. Interpretation of Parameters

Log-linear models for contingency tables are GLMs that treat the N cell counts as independent observations of a Poisson random component. Log-linear GLMs identify the data as the N cell counts rather than the individual classifications of the n subjects. The expected cell counts link to the explanatory terms using the log link. As equation (***) illustrates, of the cross-classified variables, the model does not distinguish between response and explanatory variables. It treats both jointly as responses, modeling {µij} for combinations of their levels. To interpret parameters, however, it is helpful to treat the variables asymmetrically.

2.5. Model Selection in Categorical Data Analysis

In itself, the value of the Akaike’s Information Criterion (AIC), for a given data set has no meaning. It becomes interesting when it is compared to the AIC of a series of models specified   a priori, the model with the lowest AIC being the (best) model among all models specified for the data at hand. After having specified the set of plausible models to explain the data and before conducting the analyses (e.g., log-linear model), one should assess the fit of the global model, defined as the most complex model of the set. AIC judges a model by how close its fitted values tend to be to the true values, in terms of a certain expected value. Even though a simple model is farther from the true model than a more complex model, it may be preferred because it tends to provide better estimates of certain characteristics of the true model, such as cell probabilities. Thus, the optimal model is the one that tends to have fit close to reality.

AIC= -2(Maximum Likelihood – number of parameter in the model).

This penalizes model for having many parameter. We generally assume that if the global model fits, simpler models also fit because they originate from the global model (Cooch and White, 2001; Anderson et al., 2002)

If H0 holds, A will be large (i.e. near 1) and G2 will be small. Thus, this means that H0 will be rejected for large G2.

2.6. Chi-Square Test (χ²)

The chi-square test is usually applied when considering qualitative or count data. It is also used to test independence of two variables. In this case, comparison is made between a set of observed frequencies and a set of expected frequencies.. The row (r) is the cases of criminal offences committed and the column (c) is the five years duration (i.e. 2010-2014).

Test statistic,

3. Results

Table-1. frequency of Criminal Offences by Years (percentages are in parenthesis)
Source: Records of the Nigerian Prison, Yola, Adamawa state on criminal cases, 2010-2014.

In table 1, it is seen that the highest and the lowest percentages of Murder offence was 36.667% in the year 2014 and 5.000% in the year 2010 respectively. The highest and the lowest percentages of Rape/assault was 34.247% in the year 2012 and 6.849% in both 2010, 2011 respectively. The highest and the lowest percentages of Robbery was 36.007% in the year 2013 and 10.160% in the year 2010 respectively.

Fig-1. Component Bar Chart Showing Years with various Crimes
Source: Records of the Nigerian Prison, Yola, Adamawa state on criminal cases, 2010-2014.

The highest and the lowest percentages of Drugs was 31.750% in the year 2012 and 10.500% in both 2010, 2013 respectively. The highest and the lowest that falls in the category of “Other crimes” was 46.296% in the year 2014 and 3.241% in the year 2011 respectively, these are as shown in Fig I Above:

Table-2. Table of Crime Cases by Age Groups in Percentage
Source: Records of the Nigerian Prison, Yola, Adamawa state on criminal cases, 2010-2014.

From the above table, it is seen that age-groups 16-20, 21-25 26-30 and 31-35 have high counts. People within these age groups may be graduates, and may be involved in crime because of unemployment, thus, they may think that crime becomes the only option in the absence of job opportunities. This is illustrated in Fig. II below:

Fig-2. Percentage Bar Chart showing the Age Group that Committed the Highest Crime
Source: Records of the Nigerian Prison, Yola, Adamawa state on criminal cases, 2010-2014.

Table-3. Table of Crime Cases and the Years in Percentages and Degrees
Source: Records of the Nigerian Prison, Yola, Adamawa state on criminal cases, 2010-2014.

The table above shows the crime cases in proportion by year of occurrence.  2014 has the highest number of crime cases with about 33% of the total.

Table-4. A Table of Ratios of number of Females per 100 Males
Source: Records of the Nigerian Prison, Yola, Adamawa state on criminal cases, 2010-2014.

Table 4 shows that crime categorized under Other Crimes case with ratio (5.88) has the highest number of ratio, but it cannot be regarded as the highest because it is the accumulation of crimes like illegal arms, environmental sanitation, kidnapping, forgery etc. Rape/assault has the highest number of ratio of 4.29. This means that about 4 females per 100 males were involved in the Rape/assault cases. Next is murder cases with 3.45. Robbery has the lowest ratio with 2.94. This also means that about 3 of females per 100 of males were involved in Robbery within the years (2010-2014).

Table-5. Table of Crime Rate for the years 2010 – 2014
Source: Records of the Nigerian Prison, Yola, Adamawa state on criminal cases, 2010-2014.

3.1. Results on Log-Linear Analysis

An algorithm was written in R program to evaluate the models with its respective likelihood ratio statistic (G2) as well as the Akaike Information Criteria (AIC) for each log-linear model stated above. The results are summarized below.

3.2. Chi-square Test of Independence

Table-6ai. Partial Associations for Crime cases vs Year vs Gender
Source: Result output using R statistical package

Table-6aii. Summarized results for table 6ai (C.Y.G)
Source: Result output using R statistical package

Table 6aii shows, based on the results obtained, model 6 which is the saturated model has the least AIC as expected but we select the next model with the least AIC which is model 5 with AIC of (294.53) -the all two way interactions, as the best model on the basis of parsimony.

3.2. Chi-Square Test of Independence

Table-6bi. Partial Associations on Crime cases vs Gender vs Agegroup
Source: Result output using R statistical package

Table-6bii. Summarized results for table crime, Age group and gender
Source: Result output using R statistical package

Table 6bii shows the values of AIC for each model. Comparing the values of AIC for each model on the results obtained, model 3, (CA.G) when considering association between Crime-type and Age group independent of Gender with AIC (316.30) as the best model.

4. Summary And Conclusion

From the various analyses conducted, crime incidence differs and is on the increase over the years. The year 2014 has the highest criminal cases with robbery being the highest crime committed throughout the interval of five years. The results also show that persons within the age-group 16-20, 21-25, and 26-30 were mostly involved in crime while those that are 41+ years committed the least crime. The analyses showed also that males (about 96%) are more involved in crime offences than the females (4%).

The results show also that in understanding the pattern of criminal cases committed in Adamawa state, the year of occurrence and the criminals’ age group (especially youths between 16 -35 years) should be considered. When considering Crime type vs. Year vs. Gender, it is seen that model 5, the all two way interactions is considered as the best model for the first analysis based on the results on the likelihood ratio G2 and Akaike Information Criteria (AIC). Also model 3 when considering Crime type vs. Age group vs. Gender, association between Crime-type and Age-group independent of Gender is considered as the best model.

Also the year 2014 recorded the highest crime rate of (0.173) and that crime rates from 2010 to 2014 is on the increase. A total crime rate of 0.528 for the five years interval with crude crime rate of 0.528 is observed.

References

Adejumo, A.O., 2005. Modelling generalized linear (Log-linear) models for rater agreement measure with complete and missing cases. Frankfurt am Main: Peter Lang.

Agresti, A., 2002. An introduction to categorical data analysis. New York: Wiley.

Ahmed, L., 2000. The political economy of criminality.

Anderson, D.R., K.P. Burnham and G.C. White, 2002. Kullback-leibler information in resolving natural resource conflicts when definitive data exist. Wildlife Society Bulletin, 29: 1260-1270.

Cooch, E. and G. White, 2001. Program mark: Analysis of data from marked individuals, "a gentle introduction".

Farrington, D., 2007. Understanding and preventing youth crime. Cambridge: Cambridge University, Joseph Rouwtree Foundations.

Financial, 2011. Nigeria crime. Financial Times, 7/11/ 2011.

Knoke, D. and P.J. Burke, 1980. Log-linear models. Newberry Park California, U.S.A: Sage Publications, Inc.

Lawal, B., 2003. Categorical data analysis with SAS and SPSS application. London: Lawrence Erlbaum Associates.

About the Authors

Akinrefon A. A
Department of Statistics and Operations Research, Modibbo Adama University of Technology, Yola
Adeniyi O.I
Department of Statistics, University of Ilorin, Kwara State
Adejumo A. O
Department of Statistics, University of Ilorin, Kwara State
Olawale A. O
Department of Statistics, Osun State Polytechnic, Iree, Nigeria
Ubong B.A
Department of Statistics and Operations Research Modibbo Adama University of Technology, Yola

Corresponding Authors

Akinrefon A. A