I measured some voltage values ($x$) and some corresponding frequencies ($y$). I know that the voltage can only be measured within an uncertainty $\Delta x$. And the uncertainty of $y$ shall be $\Delta y$.

Now I'm interested in the slope of the regression line and analyze it using root (the statistics tool). I used $x$- and $y$ error bars. Root returns an Error on the slope ($\chi^2$-Method).

Is this the final error on the slope or do I have to make some further propagation of errors?

  • 2
    $\begingroup$ Perhaps stats.stackexchange.com would be a better home? $\endgroup$
    – Qmechanic
    Feb 2, 2013 at 21:18
  • $\begingroup$ did not know about this one - good idea, if a mod could move it… :-) $\endgroup$ Feb 2, 2013 at 21:22
  • $\begingroup$ paging the Reverend Bayes ... call for you on line 1 $\endgroup$
    – 410 gone
    Feb 2, 2013 at 21:24
  • $\begingroup$ somehow I'm not able to directly apply the Bayes' Theorem in this case, perhaps you might give me a hint :-) $\endgroup$ Feb 2, 2013 at 21:37
  • $\begingroup$ This would be better on Stats.SE if you are primarily asking about the method or on SciComp.SE if you are primarily asking about the tool. Let me know which you prefer and I'll ask the mods if they want it (Stats in particular may already have a version...). $\endgroup$ Feb 2, 2013 at 22:11

1 Answer 1


Assuming that you mean ROOT,1 and that you fit to a TGraphErrors, then you are done.

1 First time I've heard it described as "a statistics tool" though it will do a passable job in that context.

  • $\begingroup$ yep, I meant that root and that TGraphErrors-Class :-) $\endgroup$ Feb 2, 2013 at 22:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.