Project 35
Project description

Transformations are not a cure-all for everything that ails a linear regression model. Analyze the following simple linear regression models relating the population of Tacoma, WA to the year:

  1. the population modeled on the year
  2. the logarithm of the population modeled on the year
  3. any other transformed version of the population modeled on the year
Explain the problem(s) that all of these models have in common and why this data set is so difficult to model with simple linear regression, no matter what transformation(s) might be applied.

Background on the data set

This data set gives the population of Tacoma, Washington annually from 1968 through 2009. The data comes from the Office of Financial Management for the state of Washington, specifically from the spreadsheet Total Resident Population by Year for the State and Counties: 1960 to 2009. (accessed 2/20/2010)

Variables in the data set
The variables in the data set are as follows:
NameUnitsDescription
year(years)year of population observation
populationpeoplepopulation of Tacoma, Washington
Link to the data set
The full data set in csv format is at:
http://hoard.projectivespace.com/datasets/tacomaPopulation1968-2009.csv