Sunteți pe pagina 1din 13

Statistical analysis of the relation between Crime Rate, Education and Poverty: USA, 2009

Sonarika Mahajan 100076

Research Question In this research paper, analysis is done to conclude whether the level of education and poverty influence the total crime rate in the United States of America. Using descriptive statistics such a mean, standard deviation, variance, histograms, scatter diagrams and simple linear regression analysis performed upon both independent variables separately, it can be analysed till what extent do these two independent variables, i.e. education and poverty cause fluctuations upon the dependent variable, in what proportion (direct or inverse) and of the two independent variables, which is a better predictor for determining crime rate in USA. Data description

[The states selected for this study are highlighted with yellow in the above map]

The Data that is used to define our dependent variable include both, violent crime (murder and non- negligent manslaughter, forcible rape, robbery, and aggravated assault) as well as property crime (burglary, larceny-theft, motor vehicle theft, and arson). Crime statistics used in this study are published by FBI (Federal Bureau of Intelligence) serving as a governmental agency to the United States Department of Justice. The independent variable that comments upon the education levels in the United States of America is carried out by analysing the total number of public high school graduates per state. This data includes students of all the ethnicities for the school year 2008-2009. The education universe in this study is equivalent to the total population of the state. This data has been collected by National Centre for Education Statistics (NCES), which is the primary federal entity that collects education related data in the U.S. and other countries and analyses it.

The poverty status for an individual is measured by comparing his/her income to a preset amount of dollars known as the threshold value. The poverty universe excludes children below the age of 15, people living in military barracks, institutional group quarters and college dormitories. This data is collected by the U.S. Census Bureau, serving as the most reliable source about Americas people and economy. All the data collected is cross-sectional, since it was taken during the same time period (year 2009) across different parameters. Also, the scale of measurement for these variables is the ratio scale, since the ratio between two values is meaningful and the observations are comparable to a zero value. Analysis

Mean: It is the representative of a central value for a given data set, i.e. average.

The mean value for crime variable suggests that in the year 2009, the percentage of crimes being reported in any state of USA was 3.26%. The mean value for education variable suggests that the percentage of public high school graduates being reported in any state of USA was 1% for the same time period. Similarly, the mean value for the poverty variable suggests that the percentage of individuals living below the poverty line being reported in any state of USA was 13.54%. Standard deviation & Variance: The higher the value of the standard deviation, greater is the dispersion of the data set. Out of the three variables, poverty has the highest standard deviation value of 2.98. Therefore, the percentage of individuals below poverty level is more widely dispersed over the states as compared to the other two variables. Variance is the average of the sum of squared deviation scores. It is used to compute the standard variation since its a better means for determining the dispersion of data. It is measured as the square of standard deviation for any data set.

Skewness: The symmetry of the variable distribution is measured by the help of this statistic. Crime rate has a skewness of 0.083, making it a symmetrical distributed variable since the value is closer to zero.

The education variable is skewed negatively at -.367 since the variable has lower values, indicating a left skewed histogram.

Whereas, poverty shows a positive skewness value of .670 since its variables have numerous high values, which justifies the right skewness of the histogram.

Simple linear regression model: a. Crime and Education Y = Dependent variable, Crime X = Independent variable, Education. The regression model is the equation that describes how y is related to x. This regression equation is:

From Table 2.4 in appendix, the regression equation is, Crime = 6.17 - 2.9 (Education) This regression equation can be graphed as follows assuming 0 as the intercept and 1 as the slope:

Here the slope 1 is negative.

Interpretation of the slope: For every 1% increase in the number of students being graduated from high school, there is a decrease of 2.9% in crime activities in the USA. Interpretation of the intercept: Even if there is no variation in the education level, the estimated crime rate would be 6.17%. The coefficient of determination or r2: It determines the proportion of variation in the dependent variable by the independent variable. From Table 2.2, r2 = .181 This states that 18.1% of the variation in crime rate is explained by regression of education on crime. Since this value is not close to 1, it doesnt seem to be a appropriate predictor to determine the crime rate in USA.

Hypothesis testing: Ho: 1 = 0 (education is not a useful predictor of crime) Ha: 1 0 (education is a useful predictor of crime) Significance level: = 0.05 According to the rejection rule, the null hypothesis will be rejected if p-value . From table 2.4, p-value = 0.019 Since 0.019 0.05, we reject the null hypothesis. At 95% confidence level, there is enough evidence to conclude that education is a useful predictor for crime in USA since the slope of the regression line is not zero.

b. Crime and Poverty: Y = Dependent variable, Crime X = Independent variable, Poverty. The regression equation is as follows:

Plugging in the values to from Table 3.4, get: Crime = 1.819 + 0.107 (Poverty) This regression equation can be graphed as follows assuming 0 as the intercept and 1 as the slope:

Here the slope 1 is positive.

Interpretation of the slope: For every 1% increase in the individuals below poverty line, there is an increase of .11% in crime activities in the USA. Interpretation of the intercept: With the poverty level remaining constant, the estimated crime rate would be 1.82%. The coefficient of determination or r2 From Table 3.2, r2 = .191 This states that 19.1% of the variation in crime rate is explained by regression of poverty on crime. Hypothesis testing: Ho: 1 = 0 (poverty is not a useful predictor of crime) Ha: 1 0 (poverty is a useful predictor of crime) Significance level: = 0.05 According to the rejection rule, the null hypothesis will be rejected if p-value . From table 3.4, p-value = 0.016 Since 0.016 0.05, we reject the null hypothesis. At 95% confidence level, there is enough evidence to conclude that poverty is a useful predictor for crime in USA since the slope of the regression line is not zero.

Conclusion and recommendations From this study conducted, it is assured that the crime rate in USA is directly proportionate to the people below the poverty line and inversely proportionate to the number of high school students graduating in the year 2009. When simple linear regression was performed to both the independent variables separately, the coefficient of determination (r2) and the p-value aided our study to select the variable that was a better predictor for determining the crime rate in America. Poverty, with the significance level of 19.1% is known to be a better predictor in this case as compared to the 18.1% significance level shown by the independent variable, education. This fact was further proved when the p-value for poverty stood at a lower amount as compared to its counterpart. Even though it can be concluded that poverty is a better predictor for crime rate in USA, the level of significance still stands at a diminutive 19.1%. Much stronger predictors could be used for the above study. GDP, income level, provision of federal aid or employment rate could be a few options to choose amongst.

Appendix Table 1.1 Statistics for crimes reported in 30 states of USA. State Population Violent Crime 21,179 4,421 26,929 1,74,459 16,976 10,508 1,13,541 3,559 8,397 11,278 49,547 12,842 8,304 29,444 2,473 5,059 18,559 27,121 12,440 75,176 37,929 1,298 9,744 47,965 1,508 1,21,668 17,879 22,056 14,533 1,242 Property Crime 1,77,629 20,577 2,34,582 10,09,614 1,33,968 82,181 7,12,010 47,419 69,441 90,420 2,82,918 1,39,083 87,181 2,02,698 24,024 49,614 80,763 1,81,097 75,078 3,78,315 3,44,098 12,502 1,13,511 2,77,512 13,968 9,95,145 1,91,453 2,44,368 1,47,486 14,354 Total Crime
Percentage of Total Crime

Alabama Alaska Arizona California Colorado Connecticut Florida Hawaii Iowa Kansas Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Jersey New Mexico New York North Carolina North Dakota Oregon Pennsylvania South Dakota Texas Virginia Washington Wisconsin Wyoming

47,08,708 6,98,473 65,95,778 3,69,61,664 50,24,748 35,18,288 1,85,37,969 12,95,178 30,07,856 28,18,747 99,69,727 52,66,214 29,51,996 59,87,580 9,74,989 17,96,619 26,43,085 87,07,739 20,09,671 1,95,41,453 93,80,884 6,46,844 38,25,657 1,26,04,767 8,12,383 2,47,82,302 78,82,590 66,64,195 56,54,774 5,44,270

1,98,808 24,998 2,61,511 11,84,073 1,50,944 92,689 8,25,551 50,978 77,838 1,01,698 3,32,465 1,51,925 95,485 2,32,142 26,497 54,673 99,322 2,08,218 87,518 4,53,491 3,82,027 13,800 1,23,255 3,25,477 15,476 11,16,813 2,09,332 2,66,424 1,62,019 15,596

4.22 3.58 3.96 3.20 3.00 2.63 4.45 3.94 2.59 3.61 3.33 2.88 3.23 3.88 2.72 3.04 3.76 2.39 4.35 2.32 4.07 2.13 3.22 2.58 1.91 4.51 2.66 4.00 2.87 2.87

Source: http://www.fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2011/crime-in-the-u.s.-2011/tables/table-5

Table 1.2 Statistics for public high school graduates in 30 states of USA. State Population Total Public High School Graduates 42,082 8,008 62,374 3,72,310 47,459 34,968 1,53,461 11,508 33,926 30,368 1,12,742 59,729 24,505 62,969 10,077 19,501 19,904 95,085 17,931 1,80,917 86,712 7,232 35,138 1,30,658 8,123 2,64,275 79,651 62,764 65,410 5,493 Percentage of High School Graduates 0.89 1.15 0.95 1.01 0.94 0.99 0.83 0.89 1.13 1.08 1.13 1.13 0.83 1.05 1.03 1.09 0.75 1.09 0.89 0.93 0.92 1.12 0.92 1.04 1.00 1.07 1.01 0.94 1.16 1.01

Alabama Alaska Arizona California Colorado Connecticut Florida Hawaii Iowa Kansas Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Jersey New Mexico New York North Carolina North Dakota Oregon Pennsylvania South Dakota Texas Virginia Washington Wisconsin Wyoming

47,08,708 6,98,473 65,95,778 3,69,61,664 50,24,748 35,18,288 1,85,37,969 12,95,178 30,07,856 28,18,747 99,69,727 52,66,214 29,51,996 59,87,580 9,74,989 17,96,619 26,43,085 87,07,739 20,09,671 1,95,41,453 93,80,884 6,46,844 38,25,657 1,26,04,767 8,12,383 2,47,82,302 78,82,590 66,64,195 56,54,774 5,44,270

Source: http://nces.ed.gov/CCD/tables/ESSIN_Task5_f2.asp

Table 1.3 Statistics for individuals below Poverty line in 30 states of USA. State Population for whom poverty status is determined 45,88,899 6,82,412 64,75,485 3,62,02,780 49,17,061 34,09,901 1,81,24,789 12,64,202 29,05,436 27,32,685 97,35,741 51,33,038 28,48,335 58,18,541 9,46,333 17,39,311 26,06,479 85,31,160 19,68,078 1,90,14,215 90,95,948 6,20,821 37,48,545 1,21,65,877 7,82,725 2,41,76,222 76,23,736 65,30,664 54,95,845 5,29,982 Individuals Percent in poverty below poverty 8,04,683 61,653 10,69,897 51,28,708 6,34,387 3,20,554 27,07,925 1,31,007 3,42,934 3,65,033 15,76,704 5,63,006 6,24,360 8,49,009 1,43,028 2,14,765 3,21,940 7,99,099 3,53,594 26,91,757 14,78,214 72,342 5,34,594 15,16,705 1,11,305 41,50,242 8,02,578 8,04,237 6,83,408 52,144 17.54 9.03 16.52 14.17 12.90 9.40 14.94 10.36 11.80 13.36 16.20 10.97 21.92 14.59 15.11 12.35 12.35 9.37 17.97 14.16 16.25 11.65 14.26 12.47 14.22 17.17 10.53 12.31 12.43 9.84

Alabama Alaska Arizona California Colorado Connecticut Florida Hawaii Iowa Kansas Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Jersey New Mexico New York North Carolina North Dakota Oregon Pennsylvania South Dakota Texas Virginia Washington Wisconsin Wyoming

Source:

http://www.census.gov/compendia/statab/cats/income_expenditures_poverty_wealth/income_and _poverty--state_and_local_data.html

Regression (Independent variable: Education)

Table 2.1
Variables Entered/Removed Variables Model 1 Entered Education
a b

Variables Removed Method . Enter

a. All requested variables entered. b. Dependent Variable: Crime

Table 2.2
Model Summary Adjusted R Model 1 R .425
a

Std. Error of the Estimate

R Square .181

Square .152

.67068

a. Predictors: (Constant), Education

Table 2.3
ANOVA Model 1 Regression Residual Total Sum of Squares 2.784 12.595 15.379 df 1 28 29
b

Mean Square 2.784 .450

F 6.189

Sig. .019
a

a. Predictors: (Constant), Education b. Dependent Variable: Crime

Table 2.4
Coefficients
a

Standardized Unstandardized Coefficients Model 1 (Constant) Education B 6.165 -2.904 Std. Error 1.173 1.167 -.425 Coefficients Beta t 5.257 -2.488 Sig. .000 .019

Regression (Independent variable: Poverty)

Table 3.1
Variables Entered/Removed Variables Model 1 Entered Poverty
a b

Variables Removed Method . Enter

a. All requested variables entered. b. Dependent Variable: Crime

Table 3.2
Model Summary Adjusted R Model 1 a. R .437
a

Std. Error of the Estimate

R Square .191

Square .162

.66665

Predictors: (Constant), Poverty

Table 3.3
ANOVA Model 1 Regression Residual Total a. Predictors: (Constant), Poverty b. Dependent Variable: Crime Sum of Squares 2.935 12.444 15.379 df 1 28 29
b

Mean Square 2.935 .444

F 6.604

Sig. .016
a

Table 3.4
Coefficients
a

Standardized Unstandardized Coefficients Model 1 (Constant) Poverty a. Dependent Variable: Crime B 1.819 .107 Std. Error .575 .042 .437 Coefficients Beta t 3.162 2.570 Sig. .004 .016

Bibliography
1. FBI Table 5. 2012. FBI Table 5. [ONLINE] Available at:http://www.fbi.gov/aboutus/cjis/ucr/crime-in-the-u.s/2011/crime-in-the-u.s.-2011/tables/table-5. [Accessed 28 November 2012]. 2. Income and Poverty--State and Local Data - The 2012 Statistical Abstract - U.S. Census Bureau. 2012. Income and Poverty--State and Local Data - The 2012 Statistical Abstract - U.S. Census Bureau. [ONLINE] Available at:http://www.census.gov/compendia/statab/cats/income_expenditures_poverty_wealth/in come_and_poverty--state_and_local_data.html. [Accessed 28 November 2012]. 3. Table 2.Number of public high school graduates, by race/ethnicity, gender, and state: School years 199293 through 200809. 2012. Table 2.Number of public high school graduates, by race/ethnicity, gender, and state: School years 199293 through 200809. [ONLINE] Available at:http://nces.ed.gov/CCD/tables/ESSIN_Task5_f2.asp. [Accessed 28 November 2012]. 4. World's best economies - United States: Largest economy (3) - CNNMoney. 2012. World's best economies - United States: Largest economy (3) - CNNMoney. [ONLINE] Available at:http://money.cnn.com/gallery/news/economy/2012/08/13/worldsbest-economies/3.html. [Accessed 28 November 2012]. 5. Total crimes statistics - countries compared - NationMaster Crime. 2012.Total crimes statistics - countries compared - NationMaster Crime. [ONLINE] Available at: http://www.nationmaster.com/graph/cri_tot_cri-crime-total-crimes. [Accessed 28 November 2012]. 6. Interactivate: Histogram. 2012. Interactivate: Histogram. [ONLINE] Available at: http://www.shodor.org/interactivate/activities/Histogram/. [Accessed 28 November 2012].

S-ar putea să vă placă și