[angle987654321]

I hae youre' stacken gusse

[sailormars84]

skaar wrote: yes, not a math problem, a calculator use problem. i have a calculus book, i haven't been able to get past the first two chapters without grinding to a good firm halt, i halfarse my way through trig and algebra, i had an idea what you were asking for, but was far from sure.

Yes and no. I explained it in calculator-problem format because that's what's taught in schools now. If you can do linear regression without a calculator, (which if I remember right is SUCH a pain unless you estimate) the problem is still the same. You come up with a problem in the form y = a*x + b, but the answer choices you have are ln y = a*x + b. Most people would think, "I'll just take the natural log of the a & b in the answer that I got," but that wouldn't work. You have to redo the problem using the natural log of all your y data. Then you get the right answer.

Wow, is it any wonder I'm a mod for the Geek forum?