Data analysis of sales of natural gas to households in the United States

Regression diagnostics and a time series analysis of residuals are used to help define regression equations and to identify, the weaknesses of these equations for explaining monthly deliveries of natural gas to residential customers in the United States for the time period April 1979 through March 1983. More than 99% of the monthly variation in deliveries is explained by a linear regression equation which includes heating degree days, cooling degree days, and the price of natural gas as independent variables. Final estimated relationships yield useful monthly and annual estimates of natural gas deliveries to residential customers in the United States for the time period April 1983 through March 1984. Most importantly, the estimated results when used in conjunction with the diagnostics and the time series analysis of the residuals indicate the possible strengths, weaknesses, and applicability of the estimated relationships.