Predictors of Violent Crime Rate in the United States

This regression analysis was prepared as part of a final project for Statistics 104: Introduction to Quantitative Methods for Economics.


The following report seeks to provide explanations of the variation in violent crime rates throughout the United States by analyzing the statistical signi cance of relationships between the violent crime rate and seven potential predictor variables. A multiple linear regression showed that two of the seven potential predictor variables, teen birth rates and the percentage of the population living in metropolitan areas, proved to be strongly signi cant (p < 0:001). As each of these variables increase, the violent crime rate rises. This fi nding suggests that states with greater teenage promiscuity and larger cities exhibit higher violent crime rates. Despite demonstrating a strong statistical correlation, these results cannot be used to support a causal relationship. Still, though, the results of this report provide important insights into violent crime rates and potential ways to address them.

Read more…