> this functionality, e.g. gnuplot, scipy or ….

That’s what inspired the “can of worms” comment…. B)

]]>Behind the curtain the the calculation is indeed a bit complicated and involves the (estimated) Jacobian which then gets turned into a variance-covariance-matrix of the parameters. Then you use the diagonal of that matrix as estimates for the uncertainty of the parameters.

If Excel does not offer a way to get these numbers in an easy way, I suggest you switch to another program for your data analysis. 😉

And another tip: If you want to know how this stuff works, you should take a look into the documentations of programs which have realised this functionality, e.g. gnuplot, scipy or any other open source scientific library. Usually there is at least a bit of theoretical background explained.

]]>The answer to that question appears appropriate for a weeks worth of university statistics classes. ]]>

Q) My initial data plot and linear trendline exactly match yours. My trendline equation provides m=28.984 and n=-78.300 (same as you within the precision you provided). It doesn’t provide the +/-error for m or n. How was the +/- error calculated?

]]>But the mounting of the main board is not as sturdy as it could be. It’s enough for this application, though. ^^

]]>Unfortunately it gets quite expensive really fast if you want to cover areas with this. So I just had them laser the contours and filles the areas with a pen myself.

]]>