Sunday, March 02, 2008

Missing Methodology...

Modesty aside, when I conceptualized this study on decoupling in August of 2007, I believe I was demonstrating an incredible amount of foresight. I wanted to treat the GDP/GNP and currency as products of a complex system and use them as indicators of economic coupling or decoupling. Basically, I wanted to find out if a model Asian economy (probably Thailand or Indonesia, not the Philippines) would decouple from the US economy and, if so, what next? Would it couple with China? Would some form of independence be achieved? The big BUT of this study, though, is the methodology.

What would be the best way to go about the study?

At that time, I tried to devise a risk, compensation, and response model, but it's a bit complicated (needs to be tested on simpler systems), is computationally expensive (but not as much as MD), and would not be able to work for longtime simulations (try simulating several years). I'm also considering projection methods, which would use previous data to predict the future trends. I'm just not sure what would work well on coupled data like the ones that I would have to use. And then, of course, I could just use differential equations.

Hmmm... what to do, what to do? I want to finish a concept paper on this before June so I can formally propose it to Dr. Banzon and my prospective research partner.

No comments: