Ruchika,

This would be expressed as the RPD relative percent difference.

RPD = (X1-X2) / ((X1+X2)/2) * 100

where: X1 and X2 are measurements of independently prepared aliquots of the same sample or replicate samples.

It's expressed as the difference between two values (X1-X2), divided by the average of those two values times 100

example: RPD of these 2 values X1=121 , X2=115

X1 - X2 = 6

(X1 + X2)/2 = 118

RPD = 5.08%

See page 22 of the following for more:

https://www.epa.gov/sites/production/files/2015-10/documents/chap1_1.pdf

Ruchika,

This would be expressed as the RPD relative percent difference.

RPD = (X1-X2) / ((X1+X2)/2) * 100

where: X1 and X2 are measurements of independently prepared aliquots of the same sample or replicate samples.

It's expressed as the difference between two values (X1-X2), divided by the average of those two values times 100

example: RPD of these 2 values X1=121 , X2=115

X1 - X2 = 6

(X1 + X2)/2 = 118

RPD = 5.08%

See page 22 of the following for more:

https://www.epa.gov/sites/production/files/2015-10/documents/chap1_1.pdf

You can use the RPD as Peter stated but if there is more than one pair of duplicate analyses available, there are better approaches to evaluate the standard deviation of Your method. There is a compilation of several such approaches that is available here at researchgate. Please see Article Evaluation of the standard deviation from duplicate results

Best regards

Ludwig

It is not possible to do statistical analysis with two data. Either both are correct or incorrect or one is Ok the other is not. No way to decide.

Peter Philbrook : Thank you so much for detailed explanation. This is indeed very helpful.

Ludwig Gruber : Thank you for valuable inputs and sharing the document. this is indeed very helpful