I check my training powerpoint file I was trained on year 2007. When we want to make control chart, we have to check normality of data. And also autocorrelation by JMP.
I try to search autocorrelation check on JMP14 but I cannot find it.
Any function in JMP14 to use instead of autocorrelation.
In the previous version of JMP we need to go by
Autocorrelation Check – In JMP:
Analyze > Modeling > Time Series
If take multiple measurements per lot,
conduct analysis on lot averages
Check to see if all bars fall within blue bars.
No bars outside limits = no autocorrelation
Yes, the Time Series platform provides the auto-correlation coefficient for different lags.
You can also get information about the first-order auto-correlation in the Fit Least Squares platform. Select Analyze > Fit Model. Cast the response in the Y role and add the Time variable to the Effects. Click OK. Click the red triangle at the top and select Row Diagnostics > Plot Residual by Row and Durbin-Watson Test.
@markbailey already answered your auto correlation question, but I have to ask why do you check the data for normality before using a control chart. You realize there is NO assumption of normality to use a control chart. Shewhart’s argument on pages 275-277 of Economic Control of Quality of Manufactured Product. On these pages Shewhart seems to toy briefly with the idea of using the probability approach. Paraphrasing his argument on these pages, he points out that if a process were exactly stable, i.e. did unwaveringly fit some precise mathematical model, and if we knew the details of its underlying (fixed) statistical distribution, we could then work in terms of probability limits. However, he notes that, in practice, we never know the appropriate type of statistical distribution. While statisticians usually plump, almost as if it is a foregone conclusion, for their favorite distribution, the Gaussian or normal distribution, Shewhart disposes of the use of the normal distribution on page 12 of his 1939 book, Statistical Method from the Viewpoint of Quality Control." (Shewhart's Charts and the Probability Approach, Wheeler and Neave, Ninth Annual Conference of the British Deming Association, 1996)
On pages 334–335 of Out of the Crisis Deming says:
“The calculations that show where to place the control limits on a chart have their basis in the theory of probability. It would nevertheless be wrong to attach any particular figure to the probability that a statistical signal for detection of a special cause could be wrong, or that the chart could fail to send a signal when a special cause exists. The reason is that no process, except in artificial demonstrations by use of random numbers, is steady, unwavering. It is true that some books on the statistical control of quality and many training manuals for teaching control charts show a graph of the normal curve and proportions of area thereunder. Such tables and charts are misleading and derail effective study and use of control charts.”
To ADD regarding auto-correlation:
"Thus, a major problem with the probability approach to control charts is that it is totally out of contact with reality. The assumptions used for the mathematical treatment become prohibitions which are mistakenly imposed upon practice. Restrictions such as the following are commonly encountered. “The data have to be normally distributed.” “The control chart works because of the central limit theorem—therefore you have to have subgroups of at least five observations.” “The chart will not work with serially-correlated (autocorrelated) data—the observations must be independent of each other before you can use a control chart.” These, and others like them, are examples of the tail wagging the dog. The assumptions selected for the convenience of the mathematician are turned into prerequisites for the use of the technique. In the case of control charts, this reversal is both inappropriate and misleading.
@statman Thanks you very much for your kindly instruction.
I would like to share my understanding for apply control chart. The reason that we need to check normality is to make sure that the data that we use to calculate control limit at start is appropriate. We may found that the data has abnormal by mix data from 2 machine, there are some data from different operator. In case of mix by 2 machine, it better to make 2 control chart. If the abnormal by different group of OP then we should cut it off because it is temporary problem.
For autocorrelation, this is new for me. I never use it before. I guess it will use for checking abnormality too. I will try with some data to understand it more.
Actually I plan to apply control chart to my process. I have to train my young engineer. After I study autocorrelation, if I am not 100% clear on it, I will not explain that and will use only normality test and also control chart rule test.
Thanks you very much again to get suggestion and knowledge from expert. Any addition advise, please don’t hesitate to let me know.