(((Note: The dataset I have chosen to use for the rest of the semester does not conform perfectly in relation to binary statistical testing.)))

In continuation with last weeks assignment, after downgrading Zelig to version 3.5, I was able to finish the last part of last weeks homework assignment:

How to cite this model in Zelig: Kosuke Imai, Gary King, and Oliva Lau. 2008. “logit: Logistic Regression for Dichotomous Dependent Variables” in Kosuke Imai, Gary King, and Olivia Lau, “Zelig: Everyone’s Statistical Software,” http://gking.harvard.edu/zelig 

##       2.5%      97.5% 
## -0.1846736  0.1940382

Based on these results, we have failed to reject the null hypothesis as our confidence interval includes the number 0. The results of this analysis are not statistically significant.

How to cite this model in Zelig: Kosuke Imai, Gary King, and Oliva Lau. 2007. “poisson: Poisson Regression for Event Count Dependent Variables” in Kosuke Imai, Gary King, and Olivia Lau, “Zelig: Everyone’s Statistical Software,” http://gking.harvard.edu/zelig

stargazer(count2, type = "html")
Dependent variable:
pop
life 0.111***
(0.00000)
health -0.141***
(0.00001)
water 0.055***
(0.00000)
san -0.049***
(0.00000)
Constant 9.236***
(0.0002)
Observations 179
Log Likelihood -9,426,191,937.000
Akaike Inf. Crit. 18,852,383,884.000
Note: p<0.1; p<0.05; p<0.01

I have created a new variable for population which will be used for the rest of this analysis. I have gathered the most updated numbers on population for the 173 countries I am using for this observation. The table above shows some interested results. First off, as life expectancy increases, the overall population increases which is intuitive. However, something I have been proving over the past few weeks when it comes to the health variable still holds true which is very interesting. As countries spend more money on health expenditure as a percentage of GDP, the population actually DECREASES. In other words health expenditure does not carry as much weight as is often percieved. Next, as access to improved water increases, so does the population naturally since people are not dying due to lack of water. Lastly, fo each increase in access to improved sanitation conditions, life expectancy actually decreases. After further research into this, it seems that this variable might introduce the most noise in the data as “improved sanitation” can mean something very different when compared to first and third world countries, hence I speculate this might be a confounding variable. However, despite this, my entire analysis was statiscally significant.

How to cite this model in Zelig: Kosuke Imai, Gary King, and Oliva Lau. 2007. “poisson: Poisson Regression for Event Count Dependent Variables” in Kosuke Imai, Gary King, and Olivia Lau, “Zelig: Everyone’s Statistical Software,” http://gking.harvard.edu/zelig

The graph above further shows the interaction with the sanitation variable. (Note: Data points are difficult to see due to the formatting of the variables in this dataset.)

## How to cite this model in Zelig:
## Kosuke Imai, Gary King, and Oliva Lau. 2007. "poisson: Poisson Regression for Event Count Dependent Variables" in Kosuke Imai, Gary King, and Olivia Lau, "Zelig: Everyone's Statistical Software," http://gking.harvard.edu/zelig

##      2.5%     97.5% 
## -1622.365  1757.131

The above is a histogram of population on life average and health expenditure. Based on these results, we have failed to reject the null hypothesis as our confidence interval includes the number 0. The results of this analysis are not statistically significant.