A statistical/probablity question from a dummy, for work.

W

waybomb

New member
Joined
Oct 9, 2012
Messages
3
We are a food plant, and are looking for a probability or a confidence level that our sanitation is working as planned. We clean and sanitize the factory on an ongoing basis, every day. We then test food contact surfaces and non-food contact surfaces for cleanliness and freedom of pathogens. We have increased this testing in an effort to make sure we have found, cleaned, and sanitized every nook and cranny where the pathogens could grow. In the past two months, we have gathered 195 tests, and none have shown to be positive for any pathogens. Could we develop a confidence level or a some other measure from this information that would infer our sanitation program is adequate? I'm an old guy and took statistics in college way too many years ago to remember any of this. I did find an old college stats book, and my eyes became glossed over and my head began to hurt while attempting to figure out what technique to apply. Your help is greatly appreciated!!! Thank you for your time and have a great day! Fred
 
Last edited:
Fred, you are doing good work. Who cares what test you apply if you never get a negative result. The only thing you might manage to prove is either that you could clean less to acomplish the same thing or that you could establish an acceptable number of pathogens. Why would you want to do either of those?

Based on your results, I might suggest randomizing your test procedure just to sure you test multiple locations. If all your samples are from the same spot, that would be no good.
 
tkhunny said:
Fred, you are doing good work. Who cares what test you apply if you never get a negative result. The only thing you might manage to prove is either that you could clean less to acomplish the same thing or that you could establish an acceptable number of pathogens. Why would you want to do either of those?

Based on your results, I might suggest randomizing your test procedure just to sure you test multiple locations. If all your samples are from the same spot, that would be no good.
Click to expand...

It is randomized. Nobody knows what we will test, when, until the sampler is told where to gather them and how many.
 
waybomb said:
Am 100% confident.
Click to expand...

There's your confidence level!

Seriously, I agree with TK. As you have total faith that every base is well-covered, and your confidence in those administering and analyzing the actual tests is also high, then you may actually have little need for some "fancy", statistical statement.

I don't know if you were planning on presenting such a statistical statement to anybody in particular, but (if so) would they accept such math at face value? Maybe a focus on demonstrating or explaining to others all of the reasons why your personal confidence is at 100% would be a better presentation.

Of course, if you're dealing with government inspectors, you'll need to get their playbook and focus on their ways, instead. (I spent five months on the Bering Sea one winter, working on a crab processor. I could have scrubbed a wall with iodine until the paint came off, and those inspectors would still say, "Something smells fishy here". LOL)
 
You must log in or register to reply here.
Top