Note that the test placed in Cell F30 is oversimplified. An error of 1.17% would indeed be "bad" given our ideal "Class A" error of 0.2%. Should your error be 0.3%, it is not suddeny "bad," it just means it could have been improved a bit.
Common errors to look for when computing your final error value:
- The error looks really small, e.g. 0.0053 when the reported mean value is 53.95 %Cl. This is probably the relative error, not the absolute error you get when multiplying 0.0053*53.95 %Cl, for example.
- The error is looks really small, and you are sure you didn't make the "relative error" mistake above. Students commonly forget to take the square root after adding the sums of the squares of the relative errors.
|