Control Variables and Regression Analysis (-_-)’

There has already been a post about the first week of my research internship and now I will continue with describing my work for the second and third week of my research internship (the second and third week of research respectively fall in the fourth and fifth week of the entire UROP International program – I hope that wasn’t confusing – the first two weeks of the UROP program was strictly language class and workshops).

I mentioned that during the first week, I did calculations for the “variables of interests” for the years 2003 to 2007 (inclusive). For the second week, I was introduced to the control variables of this Cultural Finance project. The control variables are: GDP per head, gross personal income, net financial worth of households, median household income, nominal GDP, inflation, private sector credit divided by GDP, local stock market capitalization divided by GDP, shareholder rights, contract enforcement, and dependency ratio. I looked over the control variables for about 46 countries and also performed calculations for the years 2003 to 2007 (inclusive) for each of these countries. (I’m sorry my explanation of my work is rather vague, but I’m not sure how much people really want to hear about this stuff. Also, to be honest, I’m starting to get a tad confused as to what is going on because way too many things are being assigned numbers… hmm, this kind of sounds like MIT since everyone at MIT speaks with numbers.)

For the third week of work, I was told to do regression analysis in MATLAB. Unlike the majority of MIT students, I have rarely programmed in MATLAB. It was a bit of a challenge for me to figure out how to program regression analyses in MATLAB. My supervisor was very understanding of that and I ended up reading some tutorials and manuals as well as working very closely with Pascal, the RWTH student who is also working on the cultural finance project. Pascal helped me a lot with getting acquainted with MATLAB and writing a program for the regession analysis I needed to do.

Regression analysis is the use of regression to make quantitative predictions of one variable from the values of another. So the regression analysis was done by using the variables of interests, the control variables, and the following beta values: autonomy, embeddedness, harmony, mastery, egalitarianism, and hierarchy. Again, everything keeps being assigned a number value and I’m starting to get a bit confused as to what exactly is going on. I hope things get clarified soon for me. For those of you who understand Cultural Finance and regression analysis, I hope I didn’t just make a complete fool out of myself in this blog post.