How do you find the best fit equation of a curve you found on a scatter plot???

Question

amistre64 · Answer

least squares regression is touted as the best

amistre64 · Answer

personally; id find the average of the xs and the ys and the inerquartile range of the slopes

anonymous · Answer

...inerquartile??? range of slopes??? egads...

amistre64 · Answer

my way gets you close to the least squares; but thats just me

amistre64 · Answer

the least squares tho, thats the way you should prolly go; and its just as complicated as it sounds

anonymous · Answer

kk i will look that up. thanks.

amistre64 · Answer

good luck with it :)

anonymous · Answer

uh, Excel works too

amistre64 · Answer

excel is helpful, but im not so sure about how it works :)  never tried it on there

anonymous · Answer

and how would one find it on excel???

anonymous · Answer

Make the scatter plot, right click on the data points and select add trendline, then choose between linear, power, exponential, etc... and check the box next to "display equation on chart"
Excel uses least squares regression as far as I know.

anonymous · Answer

THANK YOU YES THAT ROCKS!!!!

anonymous · Answer

I'd hate to have to do my own linear regression every time ;)

amistre64 · Answer

i still like my idea of finding the average slopes ... lol

anonymous · Answer

It's not totally unreasonable, but the tremendous variation you could have between data points with statistical fluctuations and all would leave an enormous margin of error, and linear regression is a pretty tried-and-true first attempt.

anonymous · Answer

ok so lets say you wanted to find like the r^2 values like you would on a graphing calculator...can you do that on excel???