6.085 Assignment Three
Analysis with Multiple Variables

We'll return to the cellphones data set from problem set one.

1. In assignment one, we were limited to comparing two regions at a
   time.  Now, we can ask the following questions: Is the average
   cellphone ownership different by region?  By HDI?  Are there any
   interactions between regions and HDI?  Describe your analysis and
   results.

Optional: Fit a model to predict 2007 cellphone ownership based on 1990
   ownership.  Does adding 1999 data significantly improve results?
   How about HDI?  Describe your analysis and results.

2. Create a table of regions and HDI.  First compute a chi-squared
   test on the counts of high HDI countries by region.  What
   hypothesis is this testing?  Next do the same test on the entire
   table (both high HDI and non-high HDI countries).  What hypothesis
   is this testing?  Explain the warning that occurs and correct it.