美國真的那麼危險嗎？

Abstract

這是一份調查美國2012年,51州的犯罪率與10個因子之間的關係的報告。10個因子包括，失業率，人口數，人均GDP，人均教育花費，綜合稅，人均社會福利，人均醫療花費，消費者物價指數(CPI)，人均年消費，警力比例。我們使用了複迴歸分析，選擇變數，多重共線性，殘差分析等方法去分析資料。資料分析發現：人均GDP，人均社會福利以及CPI與犯罪率之間呈負相關的關係；而失業率，人口數，人均教育花費，綜合稅，人均醫療花費，人均年花費以及警力比例這7個變數與犯罪率呈正相關的關係。換句話說，當人均GDP或人均社會福利或CPI上升的時候，犯罪率會隨著下降，反之亦然。

This report examines the relationship between crime rate in 51 states of the USA in 2012 and 10 other factors including: GDP per capita, social welfare per capita, the consumer price index (CPI), unemployment rate, population, education spending per capita, income tax, health spending per capita, overall spending per capita, and the proportion of police. Using the multiple regression analysis, we work on model selection and residual analysis. Our findings suggest: (1) GDP per capita, social welfare per capita and CPI have negative correlation with the crime rate; (2) the unemployment rate, population, education spending per capita, income tax, health spending per capita, overall spending per capita, and the proportion of police have positive correlation with the crime rate. In other words, high GDP per capita, social welfare per capita and CPI lead to the decrease in crime rate while the other factors increase it.