Project Proposal: Alessandra, Medhi, Sonja

Household energy consumption is an important element in household expenditures and environmental impact. Understanding the factors that drive energy consumption is important to determine what affects it the most and how to improve efficiency. Energy consumption is a crucial issue worldwide because of the limited availability of energy resources. Because the United States is a world leader and among the top two countries of mass energy consumers worldwide[1], knowing the patterns and distributions of energy consumption in the U.S. could help affect energy worldwide.

The United States Energy Administration conducts period surveys of household energy consumption and publishes it in a document “Home Energy Use and Costs”. This data describes the many different factors that drive energy consumption in households including geographic location, demographics, etc. It would be interesting to hypothesize why households in different geographic locations use more or less energy, as well as which demographics affect household energy consumption. From the data provided, energy consumption varies dramatically by both geographic location and race and income.

First, we will analyze the data from 2005[2] to observe geographic variance, determining which location uses the most energy per household and the size of the standard deviation between data values. The same will be done for the data for race and income. To determine which factor drives the household energy consumption the most, we will compare the variability in each factor – the higher the variability, the more dependent the consumption on that factor.

According the data, the northeast region shows the highest energy consumption values per household at 122.2 million Btu’s per household. In order to verify this, we will conduct a statistical analysis that will yield the percentile value of energy consumption associated with the northeastern region. This would give us an idea of how much greater northeast energy consumption is compared to the rest of the regions in the United States. For household consumption by income, households whose incomes are $100,000 or more consume the most energy compared to lower income levels at 130.5 million Btu’s per household. The trend appears to be that the greater the income the more energy consumed. We can test the linearity of this trend. From the data for household consumption by race, non-hispanic white households consume the most energy at 99.9 million Btu’s per household. However, it seems as though the energy consumption does not vary as much as it does for geographic location and income.

For each factor, we will attempt to determine whether the data fits to a normal distribution or if it is skewed. Then, we will find the mean and standard deviations for the data and if the data is normal, calculate the z-scores based on these values. Next, for each factor, the percentage will be calculated to determine what the probability is that energy consumption will be a certain value based on that factor.

After the locations, races and incomes with the highest and lowest household energy consumptions are determined, it can be hypothesized why this is the case. For example, certain locations have higher temperatures that may result in higher energy consumption for cooling. For race, it may be generalized that some races are not as concerned with sustainability or educated on the use of energy. For households with higher income, they may use more energy due to more appliances and other commodities that result from a more expendable income. Lastly, these factors may be interdependent so a more in depth study would look at the ways they affect one another.

References

Energy Information Administration. (2005). 2005 residential energy consumption survey: energy consumption and expenditures tables. Retrieved from http://www.eia.gov/consumption/residential/data/2005/c&e/summary/pdf/alltables1-15.pdf

Swartz, S., & Oster, S. (2010, July 18). China tops U.S. in energy use. Retrieved from http://online.wsj.com/article/SB10001424052748703720504575376712353150310.html


[1]http://online.wsj.com/article/SB10001424052748703720504575376712353150310.html

[2]www.eia.gov/consumption/residential/data/2005/c&e/summary/pdf/alltables1-15.pdf