cd "C:\<ENTER THE APPROPRIATE DIRECTORY HERE>"
*1.Aggregate and merge the county-level data with the state-level data. 
use "http://www.unm.edu/~ckbutler/stats/classData/CountyCensusData.dta", clear
rename state stnumber
collapse (mean) stnumber (sum)  census2000pop popestimate2000 popestimate2001 popestimate2002 popestimate2003 popestimate2004 popestimate2005 popestimate2006 popestimate2007, by(stname)
sort stnumber
save "AggregatedCountyCensusData.dta", replace
use "http://www.unm.edu/~ckbutler/stats/classData/StateDemographics2008.dta", clear
gen stnumber=.
replace stnumber=1 if state=="Alabama"
replace stnumber=2 if state=="Alaska"
replace stnumber=4 if state=="Arizona"
replace stnumber=5 if state=="Arkansas"
replace stnumber=6 if state=="California"
replace stnumber=8 if state=="Colorado"
replace stnumber=9 if state=="Connecticut"
replace stnumber=10 if state=="Delaware"
replace stnumber=11 if state=="District of Columbia"
replace stnumber=12 if state=="Florida"
replace stnumber=13 if state=="Georgia"
replace stnumber=15 if state=="Hawaii"
replace stnumber=16 if state=="Idaho"
replace stnumber=17 if state=="Illinois"
replace stnumber=18 if state=="Indiana"
replace stnumber=19 if state=="Iowa"
replace stnumber=20 if state=="Kansas"
replace stnumber=21 if state=="Kentucky"
replace stnumber=22 if state=="Louisiana"
replace stnumber=23 if state=="Maine"
replace stnumber=24 if state=="Maryland"
replace stnumber=25 if state=="Massachusetts"
replace stnumber=26 if state=="Michigan"
replace stnumber=27 if state=="Minnesota"
replace stnumber=28 if state=="Mississippi"
replace stnumber=29 if state=="Missouri"
replace stnumber=30 if state=="Montana"
replace stnumber=31 if state=="Nebraska"
replace stnumber=32 if state=="Nevada"
replace stnumber=33 if state=="New Hampshire"
replace stnumber=34 if state=="New Jersey"
replace stnumber=35 if state=="New Mexico"
replace stnumber=36 if state=="New York"
replace stnumber=37 if state=="North Carolina"
replace stnumber=38 if state=="North Dakota"
replace stnumber=39 if state=="Ohio"
replace stnumber=40 if state=="Oklahoma"
replace stnumber=41 if state=="Oregon"
replace stnumber=42 if state=="Pennsylvania"
replace stnumber=44 if state=="Rhode Island"
replace stnumber=45 if state=="South Carolina"
replace stnumber=46 if state=="South Dakota"
replace stnumber=47 if state=="Tennessee"
replace stnumber=48 if state=="Texas"
replace stnumber=49 if state=="Utah"
replace stnumber=50 if state=="Vermont"
replace stnumber=51 if state=="Virginia"
replace stnumber=53 if state=="Washington"
replace stnumber=54 if state=="West Virginia"
replace stnumber=55 if state=="Wisconsin"
replace stnumber=56 if state=="Wyoming"
sort stnumber
save "StateDemographics2008_newMaster.dta", replace
merge stnumber using "AggregatedCountyCensusData.dta"

*2.Check whether population2004 and population2007 from StateDemographics2008.dta popestimate2004 and popestimate2007 from the aggregated CountyCensusData.dta, respectively, are the same.
corr  population2004 popestimate2004
corr  population2007 popestimate2007

*3.What is the correlation between a state's electoral votes and its population in the year 2000?
corr electoralvotes census2000pop

*4.What is the correlation between a state's electoral votes and its estimated population in the year 2006?
corr electoralvotes popestimate2006

*5.Generate a new variable that is a state's estimated population in 2004 divided by its electoral votes. Summarize the new variable and the two variables that went into its construction.
gen  ratio2004=popestimate2004/electoralvotes
sum ratio2004 popestimate2004 electoralvotes

*6.Calculate the national ratio of population to each electoral vote for the year 2004. (You can use the “display” command.)
*Here are two ways to do this.
*The first recognizes that national totals = number of observations * the mean. (Yes, the 51 cancels itself out, but you can check the subvalues.)
display (51*5748853)/(51*10.54902)
*The second method uses 'extensions to generate' and then a simple generate command to arrive at the same number.
egen totalElectoralVotes=sum(electoralvotes)
egen totalNationalPopulation2004=sum(popestimate2004)
gen nationalRatio2004=totalNationalPopulation2004/totalElectoralVotes
display nationalRatio2004

*7.Do a t-test of whether the variable in question 5 equals the value calculated in question 6. (“ttest var=value”)
*Depending on which method you used in question 6:
ttest ratio2004=544965.6
ttest ratio2004=nationalRatio2004
*The tests are somewhat different, but the interpretation of the p-values is the same: The average state ratio is different from the national one.

*8.Produce a scatterplot of  a state's electoral votes (on the y-axis) and its estimated population in the year 2004.
scatter electoralvotes popestimate2004, name(graph1)
*This graph is a no-brainer: bigger states (by population) have more electoral votes.
*A more telling graph is the following in which the state ratio is plotted against the national ratio of population to electoral votes.
*It shows clearly that larger states have more population per electoral vote compared to smaller states.
scatter ratio2004 popestimate2004, yline(544965.6) ylabel(0 167753 544966 660436) mlabel(state) name(graph2)