Stat Regression Project

1

STAT 378 – 01/02 Spring 2018 Project dataset ideas and proposal example Potential places to find datasets:

Website Type of data http://www.sports-reference.com Sports https://www.data.gov Climate, energy, education. finance, health,

marketing, etc. https://www.census.gov/data.html U.S. census information https://fred.stlouisfed.org/ Federal Reserve Bank economic data https://bea.gov/ Bureau of Economic Analysis datasets,

organized by territorial units https://www.healthdata.gov U.S. healthcare data https://aws.amazon.com/public-datasets/ Amazon web services public datasets http://www.pewinternet.org/datasets/ Internet, science, technology https://www.icpsr.umich.edu/icpsrweb/ Social science, health, etc. http://www.stat.ufl.edu/~winner/datasets.html Multitude of topics, organized by statistical

analysis type https://www.umass.edu/statdata/statdata/ Multitude of topics, organized by statistical

analysis type

2

Sample project proposal: Excess greenhouse gas emissions can create climate change by over-heating Earth’s atmosphere. If too much gas is present, not enough infrared light leaves the atmosphere. An unnatural increase can be detrimental to the ecosystem. Carbon dioxide is the primary greenhouse gas emitted by humans, which over-saturates the carbon cycle. Many different factors spark the growth of carbon dioxide emissions, and the global population should be aware of ways that humans alter natural carbon recycling. We are analyzing this data in order to understand the importance of economic globalization upon the environment, namely CO2 Our project will examine the relationship between economic globalization and carbon dioxide emissions. We have data on CO2 emissions, trade openness, urbanization, and energy consumption for a random sample of 50 countries in 2010. The data for carbon dioxide emissions, trade openness, and urbanization came from the World Bank Development Indicators, while the data for energy consumption came from the United States Energy Information Administration. The response variable of interest is carbon dioxide emissions, measured in metric tons per capita. The potential explanatory variables are trade openness (measured as the sum of imports and exports as a percentage of GDP with higher percentage indicating more openness to trade), urbanization (measured as the percent of the population that lives in an urban setting), and energy consumption (measured as total primary energy consumption in million btu per capita). In addition to creating a model to understand the relationship between economic globalization variables and CO2 emissions we are also interested in examining influential observations to see which countries have smaller CO2 emissions than predicted and which countries have larger CO2 emissions than predicted.