[ | Next | Previous | Up ]

Re: Determination of sample size Pharmaceutical  Industry

From: Stan Hilliard
Date: 9/21/2004
Time: 3:32:07 PM

Comments

Alexandru, the method that you would use to calculate n depends on what kind of quality characteristic you have.

VARIABLES PLANS FOR THE MEAN:

For example, if the raw substance is a liquid, powder or a batch of something that will be blended later, and you want to base the acceptance decision on the average and standard deviation of a measured variable, use a sampling plan for the mean:

www.samplingplans.com/outputvariablesmean.htm

VARIABLES PLANS FOR INDIVIDUALS:

Alternatively, if each measurement corresponds to the quality of an individual item that will be used as such, and you want to base the sampling plan on percent of items that conform to specification limits for individual items, use a sampling plan for variables based on "Individual Specification Limits" (ISL):

http://www.samplingplans.com/outputvariablesisl.htm

VARIABLES PLANS FOR ATTRIBUTES:

Alternatively, if each item can only be classified as defective or non-defective, use an attribute sampling plan:

http://www.samplingplans.com/outputbinomial.htm

Our website -- www.samplingplans.com - describes software and provides tutorials to calculate sample size and decision limits for these different kinds of quality characteristics.


Last changed: November 20, 2007