How do I develop OC Curves from data collected?

I

irzs_2ue

hi..
ive got problem in developing the OC curve
right now i have 6 months data with various number of batches and sample sizes..i substitute all the data so then it given all the number of total incoming,total sample size and total reject

for example:

MONTH 1

Total Incoming : 7000
Total Sample Size : 500
Total Reject : 99
Target: 100%
OK : 98.59%
Reject : 1.41%

ive also develop percentage graph for that data..so is that possible to develop OC curve from the data that i got?or ive to calculate every batches which means ive to develop alot of OC curves for 6 months?how??

i really need help for this.
thanks.
 

Bev D

Heretical Statistician
Leader
Super Moderator
what exactly are you trying to do?
and what is your sample plan? how many parts inspected and how many rejects required to reject the lot? do you 100% inspect the lot after inspection? how many parts are found during 100% inspection and how many found during the sample inspection....

OC curves are not generated from actual data. they are generated from the sample size and reject on number vs various possible defect rates.

I would suggest that you search this site and others for OC curve and AOQ curve and then come back with specific clarifying questions.
 

Jen Kirley

Quality and Auditing Expert
Leader
Admin
Based on the information provided, I was not able to give an answer to this question. But I have received some more information in a message:
hi there^^
thanks for your concern
actually it is not a class assignment..
it is more to case study since the data that ive used were came from one of company here..uh~how do i say?a thesis??

after searching and reading alot of sampling plan info...correct me if im wrong
-sample size = n
-lot size = N
-acceptance number = c
-true lot fraction defective = p

but im still confused on how to determine c and p..

if ive N=1000,n=20
and from the company table,for n = 20,
-if one defect found = accept
-if 2 and more defect found = reject
so how to determine p?from p,then i can get Pa and draw the OC curve right?

another question,
since for 1 month (example;January) the company have a lot of incoming for this part(eg;XX),if that any possibilities that i can combine all the curves in one graph paper?

im really confused with this matter.
thanks for your concern..
 

Steve Prevette

Deming Disciple
Leader
Super Moderator
If you are referring to "Operating Characteristic Curves", these are determined from the parameters of the test you use, not from the data.

What an OC Curve represents is - if the "true" value of the characteristic being measured were a certain quantity, what would be the percent chance of the test rejecting a sample from that population. Thus you can assess the probability of a "false alarm" with a relatively good population, and the probability of a "failure to detect" from a relatively bad population.
 
I

irzs_2ue

If you are referring to "Operating Characteristic Curves", these are determined from the parameters of the test you use, not from the data.

What an OC Curve represents is - if the "true" value of the characteristic being measured were a certain quantity, what would be the percent chance of the test rejecting a sample from that population. Thus you can assess the probability of a "false alarm" with a relatively good population, and the probability of a "failure to detect" from a relatively bad population.

thanks for your concern..
yes,AQL that im concerned is 'Operating Characteristic Curve'..
can you tell me more in d most simple way so that i can fully understand?
im still new in sampling plan and AQL..and more of d info in internet not showing the basic explanation for me.
 

Marc

Fully vaccinated are you?
Leader
Here are 2 quick searches which might help:

OC (Operating Characteristics) Curves

Using OC Curves

The OC curve has the probability of accepting the lot on the y-axis and the actual proportion defective on the x-axis (1% defective would be 0.01). This Excel formula will generate the probability of acceptance: BINOMDIST(c, n, p, TRUE) where c = accept number, n = sample size, p = actual proportion defective, and TRUE is a cumulative flag. To draw an OC curve (or even to look one up in a book of curves), you need to know c and n. Put values from 0 to 1 in a column to use as p, and put the formula in a second column. Have Excel draw a curve with p as the x-axis and the formula results as the y-axis.
 
I

irzs_2ue

The OC curve has the probability of accepting the lot on the y-axis and the actual proportion defective on the x-axis (1% defective would be 0.01). This Excel formula will generate the probability of acceptance: BINOMDIST(c, n, p, TRUE) where c = accept number, n = sample size, p = actual proportion defective, and TRUE is a cumulative flag. To draw an OC curve (or even to look one up in a book of curves), you need to know c and n. Put values from 0 to 1 in a column to use as p, and put the formula in a second column. Have Excel draw a curve with p as the x-axis and the formula results as the y-axis.
thanks alot^^
is that c is how many accept parts within the sample size inspect?
if we have N=1000,n=20,and we reject if found 3 defects..so c = 17?
 

Miner

Forum Moderator
Leader
Admin
No. C is the number of nonconforming parts allowed in the sample.

Many sampling plans used today are C = 0 plans, which means the lot is rejected if one nonconforming part is found in the sample.
 
Top Bottom