What I mean is, for the same values of X give the same values of Y, where both parts of the formula B are non trivial? I guess it is formerly impossible for them to be identical, but can they be similar, say within 1% over what range of values? I have two regression curves, one from A and one from B. They both appear to fit the data equally well.
2006-08-03 05:20:03 · 6 answers · asked by faceface 1 in Science & Mathematics ➔ Mathematics
To the liberal economist, theory tells me formula B, but simplicity asks for formula A. And the theory isn't necessarily the correct interpretation of the data. Seems reasonable but not certain.
I don't know what you mean about 'u' ... p, q, r, s, t and u are parameters.
To annonymouse, looks good, but I figure I will just end up squashing the "r * x + s" part of the equation ( r = 0 ) with t = -p and q = u... or something. That makes the two identical, but I need a 'non trivial' solution... perhaps your answer points me at a non-trivial solution... but I don't know...
2006-08-03 07:29:35 · update #1
I think so. The derivative of A with respect to x is p*q*x^(q-1). The derivative of B with respect to x is r - t*u*x^(u-1). With the right values of p, q, r, t, and u, I think you could make the derivative functions the same. That would determine the shape of the curve to be the same. You can make identical curves line up by changing the value of a constant, which is in this case s. (That shifts the curve verically.) Good luck!
2006-08-03 05:53:36 · answer #1 · answered by anonymous 7 · 0⤊ 1⤋
Well, the real question is: what does theory tell you the relationship between your dependent variable and your regressor should be?
Just out of curiosity, why do you have 'u' as a parameter? Did you mean that p, r, and t are parameters instead?
2006-08-03 06:31:40 · answer #2 · answered by a_liberal_economist 3 · 0⤊ 0⤋
Dude I'm pretty sure your right. Good job, did you do this just cause? E-mail me we should talk sometime
2006-08-03 06:10:45 · answer #3 · answered by Anonymous · 0⤊ 0⤋
oww my head hurts :(
this site has some different typrs of calculators, etc for math problems
http://www.math.com/
http://www.webmath.com/
2006-08-03 05:24:06 · answer #4 · answered by shane 2 · 0⤊ 0⤋
I'm suing. You broke my brain.
2006-08-03 05:24:49 · answer #5 · answered by Anonymous · 0⤊ 0⤋
sorry dont know
2006-08-03 05:23:34 · answer #6 · answered by ♥I know these things♥ 4 · 0⤊ 0⤋