Comparison Between Two Results

[sams]

Hello Everyone,

I am performing comparisons between our obtained values with the values presented in the literature. These comparisons will be presented in our future paper. Usually, every relative error or a discrepancy below 10% is considered to be a "good" agreement in the Molecular Physics framework. I was wondering whether I can use "excellent" and "very good" in addition to "good" agreement in my comparisons that show relative errors below 10%.

Any advice is greatly appreciated.

Kind regards,
Sams

 

[BvU]

Hi Sams,

You mention two results. And a relative discrepancy. But nothing about the errors in the results.

E.g. 500 +/- 120 agrees very well with 600 +/- 120, even though the discrepancy is 20 %.

491 +/-2 differs significantly from 500 +/- 3, even though the discrepancy is only 2%.

I suppose you want to read up a bit on hypothesis testing and such.

 

[sams]

Hi BvU,

Thank you so much for your clear explanation...

Regards