or nonlinear regression? I'm doing some research in the subject for career purposes and it would be great to bounce off ideas.
What career path are you considering? I assume it has something to do with economics?
Without getting into detail, econometrics is VERY theoretical, and has VERY limited practical use. You can get data, and plot it, and make a nice curve, and then come up with a formula, (They have software for this), but you never really know if one parameter is REALLY corolated with another, and whether or not, your formula REALLY explains this corolation, and whether or not this really works in the real world.
I honestly didnt enjoy my stat classes where we focused on regression models. It was very theoretical, and subjective. I felt like I was in a Harry Potter class, trying to read the future.
It honestly reminds me of Al Gore. You can take one variable... say... carbon output, and another... global temperature, and if you plot them, you may get a curve, and if you add variable after variable, and twaek those variables, you can eventually get something that you like, or that can explain the temperature trends, but then when you use the model to predict whats gonna happen in the future, it falls flat.
If you want proof of this, look at many of today's economists.... they are all over the place as far as their predictions . I heard several last year saying that it was impssible for California's unemployment to go above 10%. (Boy were they off). They always carry their own biases, and they think formulas can solve everything.
Look at Keynsian 'economists'. They came up with all sorts of strange and counter-intuitive formulas, and they all fall flat when they try and use it in the real world. (You can see my bias here).
Personally, if you want to undertsand economics, take an economic history class. and dont forget to take a little Micro/Macro/Money and Banking, and Finance.
Topic: Anyone on the forum know anything about econometrics or time series analysis? (Read 591 times)