Error Propagation: Calculating Mean of Error of Measurement

AI Thread Summary
The discussion centers on confusion regarding the 'error propagation' section of a university assignment, specifically how to calculate the mean of measurement errors. Participants suggest that the assignment likely requires the mean of the errors rather than just duplicating the measurement means. There's a focus on calculating the absolute uncertainty for the mean using quadrature and scaling uncertainties appropriately. The complexity of the Excel sheet and the lack of clear instructions contribute to the frustration expressed by the users. Clarification on the purpose of the green table labeled 'Error Propagation' and specific calculations needed is sought.
sltungle
Messages
26
Reaction score
0
Okay, so I have an assignment for uni and my friends and I need to work out some info to fill out an excel document, however we're not sure exactly what it is that we're looking for.

The section we're stuck on, as the title suggests, is the 'error propagation' section. Are we looking for the standard deviation OF the mean? I don't even think we've filled the 'mean' column in correctly. I'm fairly sure it wants the mean OF the error of measurement (because if it just wants the mean of the measurements we're just duplicating the results in the Calculation of Basic Statistical Quantities section), but I'm not entirely sure to be honest.

Advice would be greatly appreciated, even if it's just a nudge in the right direction. I've attached the .xls in question so all of the relevant data is included.

Thanks in advance.
 

Attachments

Physics news on Phys.org
Since they have STDEV as a separate row, I suspect they want the absolute uncertainty for the mean:

x(avg) = sum(x)/N...so if you add up all the uncertainties in quadrature, and then scale it by N, you have

uncert(avg)=(1/N)*sqrt(uncert(x1)^2+uncert(x2)^2+uncert(x3)^2...)

uncert(avg)=(1/N)*sqrt(N*uncert(x)^2)

uncert(avg)=uncert(x)/sqrt(N)

Don't forget that when you scale your numbers (as in a conversion), you scale the uncertainity as well. (y=2*x, where x=3+/-0.5...then y=6+/-1)
 
I've got about 10 hours until this is due in and I'm still having difficulty with it (not touched it in about a week due to it having been our mid-semester break and I didn't manage to get in contact with anyone in my group from uni).

Can somebody explain what the green table labelled 'Error Propagation' is even about (and don't answer error propagation!).

For example, what am I looking for in the first box? The mean error of the flow rate? If all of the measurements are uncertain in 0.1 L/min then how can I have a mean error? It'd just be 0.1 L/min again, wouldn't it?

Maybe it's just me, but this excel sheet seems horribly confusing. I'm sure they could have put an extra sentence or two in there that would have made the whole thing clearer to me.
 
Last edited:
Back
Top