16-05-2021, 03:28 PM
(This post was last modified: 16-05-2021, 03:30 PM by farmerjohn.)
The story with the report (sorry, "internal report") might be incomplete without this You are not allowed to view links. Register or Login to view. in Koen G's blog.
bi3mw, the tolerance should (at least partly) depend on the volume of material burned. General statistics rule says that in order to increase tolerance by N times one should make N*N times more experiments.
As an example of this rule take the results from the report. 4 out of 5 (for whatever reason) were use to obtain overall average:
550 +- 35
506 +- 35
514 +- 35
490 +- 37
Average result is clearly 515 (=(550+506+514+490)/4)
Since number of experiments is 4, the overall tolerance increased 2 times and is roughly 18 (=(35+35+35+37)/4/2)
So final result is 515 +- 18 (as is stated in "Averaging the Results..." section, rounding errors may apply)
(of course this can only be done if we are repeating the same experiment
)
bi3mw, the tolerance should (at least partly) depend on the volume of material burned. General statistics rule says that in order to increase tolerance by N times one should make N*N times more experiments.
As an example of this rule take the results from the report. 4 out of 5 (for whatever reason) were use to obtain overall average:
550 +- 35
506 +- 35
514 +- 35
490 +- 37
Average result is clearly 515 (=(550+506+514+490)/4)
Since number of experiments is 4, the overall tolerance increased 2 times and is roughly 18 (=(35+35+35+37)/4/2)
So final result is 515 +- 18 (as is stated in "Averaging the Results..." section, rounding errors may apply)
(of course this can only be done if we are repeating the same experiment
)