Search the Elsmar Cove!
**Search ALL of Elsmar.com** with DuckDuckGo including content not in the forum - Search results with No ads.

Determining Sample Sizes for Testing (Chemistry or Microbiology)

R

regina_q

#1
I am looking for what people are using to determine the sample size for testing (chemistry or microbiology) when the method or standard does not list sampling calcs.

I have previously used a simple process (FDA Guide to Inspections for Quality Systems, August 1999) by which confidence level is (95 or 99%), confidence interval (at 0.3) and population are the inputs and the output is the sample size. This calculation is for sampling of records so I'm looking for another easy calc for determining sample size for testing. BET (LAL or endotoxin) testing is one example. So if my lot size is 500, how many sample would I need to test to maintain a certain confidence?

Anyone know of an "industry standard" calculator to determine sample size?
 

Ajit Basrur

Staff member
Admin
#2
I am looking for what people are using to determine the sample size for testing (chemistry or microbiology) when the method or standard does not list sampling calcs.

I have previously used a simple process (FDA Guide to Inspections for Quality Systems, August 1999) by which confidence level is (95 or 99%), confidence interval (at 0.3) and population are the inputs and the output is the sample size. This calculation is for sampling of records so I'm looking for another easy calc for determining sample size for testing. BET (LAL or endotoxin) testing is one example. So if my lot size is 500, how many sample would I need to test to maintain a certain confidence?

Anyone know of an "industry standard" calculator to determine sample size?
Did you refer this page - Sampling Plans: Instructions & Tables ?

Yes, it does refer to 95 and 99 % confidence level.

Lets wait for more responses.
 
R

regina_q

#3
Thanks for the link...and yes, this is the sampling plan determination we use for records audits, etc. I really just need to know how others are determining sampling sizes for chemistry or microbiological testing when the ISO or method doesn’t dictate it. Like you said, we'll wit to see if we get further feedback.
 

Bev D

Heretical Statistician
Staff member
Super Moderator
#4
The industry "Standard" would be either ANSI Z1.4 for categorical (attributes data) or ANSI Z1.9 for continuous (Variables) data.

You can also use the straightforward calculation for categorical data sample sizes for a c=0 plan using the Binomial. There is also calcualtor here on the site that eanbles you to optimize for acceptable and rejectable level as well as confidence.

If you have continuous data the formula choices are a bit more complex as not all distributions are naturally Normal...

If you search for sample size on this site you will find many helpful posts.
 
R

regina_q

#5
Thanks so much Bev. I appreciate your input and will look for the calculator as you suggested.
 
Top Bottom