Question

A. 5.1(plus/minus 0.3) - 3.7(plus/minus 0.2) I get 36% for relative error, why is this wrong?

Answer 1

.36 is your propagated uncertainty, not your relative error. Relative error is a statistical tool and this is just subtraction.

Answer 2

@Sinbearciante I got the same thing.. So it is propagated uncertainty? Is relative error not involved in this, then?
\( 5.1(\pm 0.3) - 3.7(\pm 0.2)\)
Absolute error adds up for addition and subtraction, so
\( 5.1(\pm 0.3) - 3.7(\pm 0.2)\equiv 1.4\pm0.5\)

Then the relative error, or fractional error is \(\dfrac{0.5}{1.4}\approx 0.3571\rightarrow 0.36\) with significant figures.

Answer 3

Which is a relative error of \(36\%\), I mean. :)

Answer 4

So, I don't know why it's wrong either.

Answer 5

well, I guess it would be \(.4\) or \(40\%\) with significant figures. Only 1 sig. fig. each error.

Answer 6

Hmmm I have always used absolute and relative error for analyzing an experimental value against a true value and this seems to be simple arithmetic. And I misspoke when I said that relative error was only for statistical analysis! When I saw the % I thought they were referring to another application. Relative error, in the way described here, is a decimal value (percent error is when it is then multi by 100).

Answer 7

And yes I do think sig figs may be the problem!

Answer 8

Thank you! :) I think I know what you are talking about! And thank you correcting my terminology! I think that relative error is that \(\dfrac{\text{absolute error}}{\text{amount}}\), and can be expressed as fractional or a percentage. Here is a use of it being a percentage from a ".edu" site: http://www.phy.ilstu.edu/slh/absolute%20relative%20error.pdf