Page 1 of 1
Accuracy and Precision
Posted: Mon Nov 17, 2008 3:01 am
by thohry
Hi all,
I can not distinguish the 2 terms clearly. In which context accuracy is used and what situation precision is used puzzles me a lot. Please help
Thank you.
Posted: Mon Nov 17, 2008 5:13 am
by tom jupille
Accuracy is getting the right answer.
Precision is getting the same answer every time.
This diagram taken from our
"Getting Started in HPLC" course may help:

Posted: Tue Nov 18, 2008 3:11 am
by thohry
Thank you tom jupille for the great help.
Now i have one more quest : precision is somewhat similar to reproducibility, is'nt it?
Posted: Tue Nov 18, 2008 3:20 am
by Consumer Products Guy
precision is somewhat similar to reproducibility, is'nt it?
You're concerned about precision, reproducibility, and repeatability, and I can't figure out how Tom posted his picture !!!
Posted: Tue Nov 18, 2008 2:20 pm
by DR
Thank you tom jupille for the great help.
Now i have one more quest : precision is somewhat similar to reproducibility, is'nt it?
Yes, very much so.
Posted: Wed Nov 19, 2008 5:06 am
by tom jupille
I can't figure out how Tom posted his picture
Basically the way I described in the "sticky" message at the top of the LC board. You have to have the graphic stored somewhere accessible via the web. In this case, the figure is actually in the on-line course, so I just used that link.
Posted: Wed Nov 19, 2008 9:57 am
by Victor
Tom, your pictures are very pretty and I use the same analogy myself in teaching.
However, I always wondered how I would answer the question from a student with regard to the "accurate but imprecise" picture.
"Hey, I can see that an analytical result could be too high or too low...but how can it be too far to the left or too far to the right?"
No-one ever asked this question though......
Posted: Wed Nov 19, 2008 10:59 am
by lmh
Victor,
It's much worse than that. If you are measuring 18 analytes, and 17 are spot-on, but one is too high, then you need a word for "up-a-bit" in 18-dimensional space. "too far to the left" only scratches the surface, being quite nice in 2-analyte, 2-dimensional space!
The rest,
Can I disagree with the idea of reproducibility being nearly a synonym of precision? It's helpful if different words have different meanings, so I'd prefer to reserve precision for the agreement of measurements within one batch of samples, while using reproducibility to indicate whether I can get the same result again tomorrow, or on a different machine, or in a different lab. Any sense in this, or is there another word I ought to be using?
Tom,
Really like the diagrams!
Posted: Wed Nov 19, 2008 12:14 pm
by krickos
lmh,
I agree with you regarding precision, at least when using ICH guidelines one need to keep the different terms separated to avoid confusion.
4.1. Repeatability
Repeatability expresses the precision under the same operating conditions over a short interval of time. Repeatability is also termed intra-assay precision .
4.2. Intermediate precision
Intermediate precision expresses within-laboratories variations: different days, different analysts, different equipment, etc.
4.3. Reproducibility
Reproducibility expresses the precision between laboratories (collaborative studies, usually applied to standardization of methodology).
Posted: Wed Nov 19, 2008 6:02 pm
by lmh
Krikos, thanks for that set of definitions! Much appreciated...