The city of Chicago in Illinois has the highest per capita homicides in the United States, ahead of Los Angeles and New York City. The Chicago Police Department and other law enforcement agencies have a challenge especially on how to manage the situation. However, it is noteworthy that the number of homicides in Chicago in 2011 is less than half the number in 1991 (CPD, 2012). The homicides figure in 2011 was a 30-year low for the city. One more homicide is still one too many a life lost.
Homicide and murder are two terms that people use interchangeably in daily conversation. However, legal definitions differentiate the two terms (CPD, 2006). In this case, I will use the two terms interchangeably as is the case in the Chicago Police Department website. The website lists several motives for murder as per the records. The motives include street gang altercation, gangland altercation, armed robbery, and domestic altercation (CPD, 2012).
An analysis of the main causes of homicides will assist policy makers to decide which initiatives would have the greatest impact at reducing homicides in Chicago. In particular, gang related crime is rife in the city of Chicago. It is quite easy to associate the high homicide figures in Chicago with gang activity and gang violence. In fact, gang violence accounted for 20% to 40% of all homicides in very large cities, in the United States, in the period 1996 to 2009 (Howell et al., 2011).
Design and hypothesis
Statistical analysis provides insight and may demonstrate a correlation between two variables. This would be the case in the issue at hand. The data I use in this analysis comes from the Chicago Police Department database, and the research department of the CPD used the data to prepare the 2011 Murder Analysis Report. The report is in the police department website portal.chicagopolice.org. This report contains the summarized tables for common murder motives and murder victims by gender between 1991 and 2011 (Table 2 and table 3, appendix).
The data gives the annual number murders in Chicago as well as four common motives for murder. The Chicago Police Department collects the data using a reliable system that systematically codes to capture all reported crime and classify it. The system also provides information to the residents by mapping crime on the Chicago area. The data is quite reliable given that it captures the activity of the police day by day.
The main contention is that homicides in the city of Chicago are mainly due to gang violence. Consequently, reducing gang violence should be the main focus for the Chicago Police Department. A statistical analysis requires that I state the problem as a hypothesis:
H0: Homicides in Chicago are not mostly caused by Gang Violence
H1: Homicides in Chicago are mostly caused by Gang Violence
The analysis takes homicides in Chicago as the dependent variable while homicides due to gang violence as the independent variable. The homicides due to gang violence are a summation of homicides due to street gang altercation and gangland narcotics. All the analyses will be in EXCEL.
The implicit issue is whether there exists a correlation between the two variables. Therefore, an analysis using Spearman's rank test will establish the strength of the connection between the variables. A 5% significance level is the critical value for accepting the hypothesis (Myers and Well, 2003). A regression equation will determine if homicides are dependent on the gang violence homicides. In excel, a plot of the trend line is useful to deduce dependence of the variables.
The analysis involves carrying out tests using the homicide values for gang violence and those for domestic violence. Only then can a comparison be made and an interpretation of the result in a useful manner. This is in line with the contention that gang violence is the major cause of Homicides.
A preliminary inspection of the data shows a general decrease in the number of homicides from 1991 up to 2011. Homicides due to domestic violence decrease almost uniformly over the years while homicides due to gang violence decrease but have spikes in 1993, 2001, and 2008. There is a corresponding increase in the number of total homicides in Chicago whenever gang violence homicides increase. On average, gang violence contributed to about 32.8% of the homicides in the years 1991 to 2011. On the other hand, 7.6% of murders were due to domestic violence in the same period (Table 4, appendix).
Graph 1: Homicides as a percentage of the total number of Homicides
In performing in depth statistical analysis, I calculated the correlation coefficient r, the slope, and intercept. Then I performed the Spearman's rank test on the data while giving the confidence interval for deciding whether to accept the alternate hypothesis. I plotted a scatter plot and inserted a trend-line which was to be consistent with the slope and intercept obtained by calculation.
The initial calculation of the correlation coefficient produced 0.5800 for gang violence homicides and 0.9129 for domestic violence homicides. These were simply to give direction on whether there is a positive or negative correlation. It is apparent that a positive correlation exists. This provides guidance while doing Spearman's rank test since a one tailed test is sufficient.
Table 1: initial values calculated.
The spearman's rank test for gang violence homicides produced a correlation coefficient of 0.5734 at a significance level of 1%. On the other hand, the test produced a correlation coefficient of 0.8601 at a significance level of 1% for domestic violence homicide. Gang violence homicides, therefore, have a slight positive correlation with the total number. Domestic violence homicides have an even stronger positive correlation with the total number of homicides.
A scatter plot gave the visual impression for both sets of variables. I compared the variables to the initial value of slope obtained from calculation. The trend line is the regression line and gives the extent to which one value is dependent on another.
Graph 2: Scatter plot for gang violence homicides and annual total homicides
Graph 3: Scatter plot of domestic violence homicides and annual total homicides
While analyzing the scatter plot, the initial calculated values of slope are necessary (Venkateshan, 2010). Graph 2 (gang violence) has a regression equation y=1.47x + 340.78 while graph 3 (domestic violence) has a slope of y= 7.03x + 288.27. The regression coefficients are 0.34 for gang violence homicides and 0.83 for domestic violence homicides. Therefore, total homicide show dependence on domestic violence while it displays little dependence on Gang violence.
The analysis shows that gang violence homicides represent a third of all homicides in the city of Chicago. On the other hand, less than ten percent of all murders are due to domestic violence (graph 1). However, analysis of correlation and dependence has shown that domestic violence return stronger values for correlation and dependence to the total number of homicides. The focus here is on gang violence homicides, which show a positive connection with the entire number of homicides. However, the total number of homicides shows little dependence on the gang violence homicides.
The alternate hypothesis holds. Homicides in Chicago are mostly caused by gang violence. This is because there is a slight positive correlation to the total number of homicides and a slight dependence. Furthermore, gang violence homicides account for a large fraction of all homicides in the city of Chicago as per the preliminary analysis.
There are a number of issues that come to the fore in this analysis. First, the analysis demonstrates that correlation does prove a causal relationship. Secondly, the underlying social issues are complex and further analysis of correlation between gang violence and other variables is necessary…