Complaints time series significant change

@Bev D thanks, the book is not that expensive.... and winter is coming here so plenty of time to read
@Steve Prevette thank you I will have a look and further process and also look what the suggestions of Bev D bring.

Thank you both

Bev D said:

In my experience this isn't unusual. The c chart can work but with the number of uses of various parts of the software that might generate complaints being susceptible to non-homogenous variation the I, MR chart is more likely a better indicator of stability - it certainly matches what our eyes tell us.

As Dr. Wheeler pointed out the c chart and (broken link removed) are subject to many 'assumptions' that are really requirements and so the I, MR chart is better choice.

Click to expand...

Thank you for the advice @Bev D , with parts do you mean modular software? That is (not yet) the case. Unfortunately tagging of risk and complaint type is also not yet done but will be set-up soon because it is a mandatory requirement for our products as PMS activity. This in itself will bring ofcourse more insight and meaning to the complaint data.

Not knowing what the software is makes it difficult to be specific and correct. But the use of the software may be non-homogenously variable due to many factors: - from the number of uses to what the use cases are to the 'sensitivity' to complaining of the users.

You can also get the book from Amazon as a kindle book if you use kindle...

Bev D said:

Not knowing what the software is makes it difficult to be specific and correct. But the use of the software may be non-homogenously variable due to many factors: - from the number of uses to what the use cases are to the 'sensitivity' to complaining of the users.

You can also get the book from Amazon as a kindle book if you use kindle...

Click to expand...

Yeah I saw that option but I don't use kindle, I prefer getting a copy of the book from SPC press

01mercy said:

Yeah I saw that option but I don't use kindle, I prefer getting a copy of the book from SPC press

Click to expand...

There is something about being able to hold the book isn't there? Also, Kindle is fine for reading novels, but it doesn't display graphical elements very well.

Miner said:

There is something about being able to hold the book isn't there? Also, Kindle is fine for reading novels, but it doesn't display graphical elements very well.

Click to expand...

Yeah I prefer a book.... I think (at age of 44) I'm old fashioned

@Bev D @Steve Prevette can you tell me why you have chosen also for the I-mr-chart to be used since the count of complaint is descrete but this chart is pointed out to be used for continuous data?

As I understand correctly? the I-MR chart sets the control limits based on the variation in the Range between the points and the c-shart sets the limits based on the standard deviation over the individual measurements?

In my reply on Friday I refer to two articles that Dr Wheeler wrote. The p chart and the I, MR chart. They should be highlighted in the post. This means that if you click on the word you will be directed to the article. (The words are links to the articles). The articles are free to read and explain the situation quite well. The I, MR chart actually works for any data type. If you still have questions after reading the articles let us know.

I default to the c-chart (which assumes a Poisson distribution) whenever I am counting events, defects, injuries. Most times (in my experience dealing with such data) each individual event is independent from each other. The closely related u chart also works well, for example, OSHA recordable injuries per 200,000 hours worked.

BUT in this case, the sum total of the complaint data showed a variation higher than the Poisson. Therefore, I tried the iMR chart (and probably woudl not be a bad idea to plot the R chart, but that usually ends up confusing folks more than enlightening them, IMO). Since it fit, I conclude that either there are "lumpings" of complaints (one initiating event begats multiple complaints), OR there is a mixture of different variations in the individual feeds. Thus, a good thing to do is start going down to the individual categories and sources of complaints to see what that tells you.

On the flow chart, I initially followed the "discrete data" path, then defects, then c-chart.

Bev D said:

In my reply on Friday I refer to two articles that Dr Wheeler wrote. The p chart and the I, MR chart. They should be highlighted in the post. This means that if you click on the word you will be directed to the article. (The words are links to the articles). The articles are free to read and explain the situation quite well. The I, MR chart actually works for any data type. If you still have questions after reading the articles let us know.

Click to expand...

My apologies @Bev D I missed that, I will read it.