Rocky

Statman,
Are you still in vacation/holiday ?? We haven’t hear anything from you for quite sometimes now. We are wondering when will you surface… we are eager to know update for this debate..
 
Reigle,Statman Too, Dixon…etc…
IS there a predetermined date already for this debate?? Please update us for development….this is turning…[Read more]

Cantula,Go to:http://www.lean.org/Community/Resources/PresentationLibrary.cfm you can find plenty of idea there not here. Your in a wrong channel.rocky

Please include:NokiaElcoteqMagnecompSAEGKI 

try :
http://www.lean.org
rocky

Jean,
In layman’s explanation:
Correlation coefficient measure the linear relationship of a two measureable variable.
If I have an apple compare to another apple, a linear relationship (Correllation coefficient of 1) is that they are exactly the same. If I bite on one of the apple, correllation coefficient changes to let say -0.90 (depends on…[Read more]

Tony,
Have you had DOE classroom session ? Step could vary depends on skills and individual.
Anyway, check for the main effect, interactions and cube plots. Analyze. Draw conclusion base on your plots at your desired response. Determine your regression equations. Know your variance of error. Determine your factor settings that will provide you…[Read more]

Tony,
It is surprising for me to know that you use blocking in your DOE without knowing what it is?? Hmmmm.  Strange…
Anyway, a pvalue of less than 0.05 simply means that your blocks effect are statistically significant. It means, that data collected in your blocks has statitical effect with your response whatever it is. Block by the way, are…[Read more]

Bayanbaru,
Sampling inspection are created for rational inspection in consideration of inspection cost and cost due to poor quality of your end product. In selecting AQL, you need to review your end product history and if there are any costumer requirements. The AQL should be on the break-even point quality and can be rationally calculated…[Read more]

Dear Manager,
It is nice to know that your interested for this knowledge. You can help for better deployment of increasing number of stagnant BB.
DPMO is a management metric to determine/compare the performance of plant ranges product. Your right, to compare apple to orange. If your company engage in manufacturing television and washing m…[Read more]

You can’t find apllicable table for DPU convertion but you can calculate.
DPU = DPMO * (No. of oppurtunities per unit) /(1,000,000)  
DPMO can be find in applicable Z-table
 
 

TM Aw,
SPC charting are used to verify the process are behaving the way you expected it to be. Having SPC around will  not solve the problem but will tell you the problem exists thus you can react swiftly on the matter. Are you getting tired of analyzing the data for your HMLV production ?? Hope not !
I am in the same manufacturing industry…[Read more]

Mike,
Every game set in casino has very high probability for  a bet to loss than win. Thus, most casino prosper than any other else bussiness. Most of the game set in the casino have a worst case of winning compare to tossing a coin. Considering a coin was symetrically manufactured with no-bias, the probability of winning to lossing was 50/50…[Read more]

Tejas Dalal,
Your at too late for the discussion and not keen enough to search this site. You could go to the link below and follow the long trend of discussion about this matter….peace…
rocky
http://www.isixsigma.com/forum/showmessage.asp?messageID=25322

Brad,
Sorry if I confuse you. The number of oppurtunity increases since you are now including the component body (1) due to chip off or damage criteria plus the number of pins (8) for soldering criteria. Hope it will now be very clear.
rocky 

Brad,
Let say you have an IC (Integrated Circuits) with 8 pins and your inspection are focus on solder joints, then you have 8 numbers of opportunities. But if a smallest component chip off or damage will also included in the criteria, then you have 9 number of opportunities. Probably, the 150 opportunity in the sample you have are the number o…[Read more]

Matt,
Don’t confuse Love…..It’s just a simple Run Chart…
 
rocky 

Love,
You can send message to me at:
[email protected]
rocky

Love,
If you just read my previous e-mail thoroughly. Try to get more data.
I stated that even you have 5 or 3 sample you can get normal distribution via software. To eliminate you confusion, get more data. If your bimodal distribution still exist, then you can analyze the rootcause.
rocky  

Puppy Love,
Have you get more samples? or just running your previous data to a software?
Are you using Minitab? What is your alpha level? What is the your p-value?
If you have a small sample size, the normality test will focus only on your inputed data. Software will calculate everything based on this. It will layout the data and try to fit a…[Read more]

Let me correct myself….
its………..statistical abyss…