Large RSD

GOOD VALIDATION PRACTICE (cGVP) EQUIPMENT VALIDATION
1

values 1:
1
1
1
1
1.0000001
RSD=4.472

values 2:
0
0
0.0000001
0
RSD=223.6067% !!!

What’s wrong with 0 ???
why is the RSD too large ???

Why is the RSD too large ?

2

[quote=strong46202]values 1:
1
1
1
1
1.0000001
RSD=4.472

values 2:
0
0
0.0000001
0
RSD=223.6067% !!!

What’s wrong with 0 ???
why is the RSD too large ???

Why is the RSD too large ?[/quote]

Is there anything wrong with my question!!!

3

RSD (Relative Standard Deviation) is the Standard Deviation divided by the Average and then multiplied by 100 (to make it a percentage).

		0.0000000000		1.0000000000
		0.0000000000		1.0000000000
		0.0000000000		1.0000000000
		0.0000000000		1.0000000000
		0.0000000000		1.0000000000
		0.0000001000		1.0000001000

SD 0.0000000408 0.0000000408
Average 0.0000000167 1.0000000167
%RSD 244.9489742783 0.0000040825

The SD is the same in both cases. Dividing by a smaller number gives a larger result (or the other way around, dividing by a larger number gives a smaller result). It’s just normal math. The SD of 0.0000000408 is a much smaller percentage of nearly 1 and a much larger part of a tiny number nearing 0.

The take away message: though the SD is the same, it’s huge in comparison (relative) to the original data.

4

[quote=neerav]RSD (Relative Standard Deviation) is the Standard Deviation divided by the Average and then multiplied by 100 (to make it a percentage).

		0.0000000000		1.0000000000
		0.0000000000		1.0000000000
		0.0000000000		1.0000000000
		0.0000000000		1.0000000000
		0.0000000000		1.0000000000
		0.0000001000		1.0000001000

SD 0.0000000408 0.0000000408
Average 0.0000000167 1.0000000167
%RSD 244.9489742783 0.0000040825

The SD is the same in both cases. Dividing by a smaller number gives a larger result (or the other way around, dividing by a larger number gives a smaller result). It’s just normal math. The SD of 0.0000000408 is a much smaller percentage of nearly 1 and a much larger part of a tiny number nearing 0.

The take away message: though the SD is the same, it’s huge in comparison (relative) to the original data.[/quote]

So What should i do for smaller numbers?? I shouldn’t calculate RSD? How can i obtain Precision ?

5

You are comparing apple and oranges. If you are calculating RSD at the accuracy that is being presented in this thread that means that you have a very accurate instrument and the results presented should fail and an investigation triggered. RSD should be based on the accuracy and resolution of the instrument and adjusted accordingly. The lowest the LOD the highest the allowed RSD because of accuracy and resolution.

6

Thanks for reply
Is there any source for your information ?

7

[quote=strong46202]Thanks for reply
Is there any source for your information ?[/quote]

Development and Validation of Analythical Methods
Christopher M Reily and Thomas W. Rosanske

What are the specifications for the isntrument you are using?