C RIME I N A MERICA : L ET ’ S C LEAN U P T HE S TREETS ! Chris Blees Conner Faught Ryan Shaffer
V IOLENT C RIME IN A MERICA Question: What factors lead to violent crime and do those factors vary by the size of the city? –Factors Examined: Percentage of Population Under the Poverty Line People per Family Distance from State Capitol Size of Police Force High School Graduation Rate Annual Precipitation Divorce Rate Average Travel Time to Work Percentage of Population that are Civilian Veterans Median Age
C ITIES U SED IN A NALYSIS 250,000+ Atlanta, GA Aurora, CO Buffalo, NY Phoenix, AZ Pittsburgh, PA Tampa, FA Toledo, OH Santa Ana, CA Oakland, CA Arlington, TX 100, ,000 Richmond, VA San Bernardino, CA St. Petersburg, FA Portsmouth, VA Charleston, SC Paterson, NJ Westminster, CO Orange, CA Birmingham, AL Syracuse, NY 60, ,000 Upper Darby Township, PA Tustin, CA Tuscaloosa, AL Tracy, CA Temecula, CA Sugar Land, TX Beaverton, OR Bellflower, CA Clifton, NJ Columbia, MO 20,000-60,000 Ames, IA Anderson, IN Binghamton, NY Bloomfield, NJ Bowling Green, KY Bullhead City, AZ Greenburgh Town, NY Hendersonville, TN Kirkland, WA Mansfield, TX
D ESCRIPTIVE S TATISTICS MeanMedianStd DeviationMinimumMaximumCount Under Poverty Line People/Family Distance from State Capitol Police Officers HS Graduation % Precipitation % Divorced Time to Work % Civilian Veterans Median Age
E XPECTED R ESULTS Independent VariableExpected Relationship Percentage of Population Under the Poverty Line Positive People Per FamilyPositive Distance From State CapitolPositive Size of Police ForceNegative High School Graduation RateNegative Annual PrecipitationPositive Divorce RatePositive Average Travel Time to WorkNegative Percentage of Population that are Civilian Veterans Negative Median AgeNegative
R EGRESSION A NALYSIS Regression Statistics Multiple R R Square Adjusted R Square Standard Error Observations40 ANOVA dfSSMSFSignificance F Regression E-07 Residual Total
R EGRESSION A NALYSIS, CONT. CoefficientsStandard Errort StatP-value Intercept % under Poverty Line Average Number of People/Family Distance From State Capital(Miles) Police Officers (2008) High School Graduates (Percentage) Average Precipitation (inches) % population divorced mean travel time to work % of population that are civilian veterans median age
L INEAR R ESULTS Independent VariableExpected Relationship Percentage of Population Under the Poverty Line Positive People Per FamilyNegative Distance From State CapitolPositive Size of Police ForcePositive High School Graduation RateNegative Annual PrecipitationPositive Divorce RatePositive Average Travel Time to WorkPositive Percentage of Population that are Civilian Veterans Negative Median AgeNegative
V ALIDITY Because the test for homoscedasticity for the predicted values versus the residuals produces a notable ovoid shape, and because the histogram of the residuals produced a curve with a normal, bell-curve shape, the model appears to be valid.
C ONCLUSIONS Of the factors we chose, only four had a significant impact on violent crime rates nationwide. –Percentage of the population under the poverty line. –High school graduation rate. –Mean travel time to work. –Percentage of the population that are civilian veterans. Possible limitations of our model could include: –Small sample size (40 cities). –Data taken from –Only examines violent crimes.
S OURCES U.S. Census Bureau of Justice Statistics