Almost 8 years later and this video is still helping people! Thank you for such an informative and digestible method of teaching me how to do this. MUCH appreciated.
A couple other commenters caught this, but this can't be overstated for your viewers: THIS VIDEO CALCULATES CONTROL LIMITS INCORRECTLY AND WILL LEAD TO THE CREATION OF MISLEADING CHARTS THAT ARE USELESS FOR PROCESS IMPROVEMENT. Using any STDEV function calculates the dispersion of all data relative to the overall mean, which exaggerates the effects of outliers and small shifts, thereby inflating the limits and hiding the signals that they are supposed to detect. The key insight of Walter Shewhart was that our statistics need to account for the element of time: this isn't a sample of numbers representing a population, but rather, an ordered sequence of data produced by a causal system over a period of time. The CORRECT way to calculate the control limits is to calculate the point-to-point differences (the "moving range"), average them, and divide the average of the moving range by 1.128 (bias correction for subgroup of 1, which is what the individuals chart is), which gives you Sigma. Control limits are 3 Sigma on either side of the mean, unless there is a barrier on one side of the mean (e.g., if your UCL is less than zero when a negative value for the process variable would be nonsensical), in which case you just get a one-sided chart, and you can detect "Rule 1" (Western Electric) signals on only one side of the average. I would recommend that your viewers look up Dr. Donald J. Wheeler's FREE articles on Quality Digest, where he clearly and accessibly explains the statistical underpinnings of SPC and how to use it to drive process improvements. I would also recommend that your viewers calculate control limits in the way I described, also calculate control limits according to the method described in this video, and then try different modifications to the data stream to experiment with how different kinds of outliers affect the calculated limits in both approaches. It will give you a real hands-on sense of what makes the Process Behavior Chart so unique and why a population-statistics approach falls short.
This is the best training video I have ever seen, and I have seen lots. Doug is great at providing the information in a logical, very understandable way.
Hi Doug, I really appreciate the video and can you please consider next time adding the excel file in the description so we can try it ourselves. Thank you.
Thank you for doing such an outstanding job explaining without adding any distractions or going into any tangents. This is perfect for my operations course!
@@DougHExcel that’s brilliant. I have another question, if that’s okay? How do we prevent the UCL and the LCL from moving along with the mean? As the current way would move the LCL and UCL, so how would I keep the control the same? I’m sorry if this doesn’t make sense - but I’m wondering how I can keep the lines constant as the test progresses to measure the performance in regards the initial control levels.
In real application, the mean should not be the population mean but rather the true value of the CRM. Is that correct? Moreover, which standard deviation should be use for 1sd, 2sd and 3sd? Is it from the SD from measurements or SD from the CRM itself?
Thank you for an excellent tutorial. Since the mean and standard deviation change with the data, how do you identify data that is outside of bounds, hence a suggestion of process instability?
Careful when calculating control limits...it is not technically correct to simply use sample standard deviation. You should really use the short term standard deviation calculated from moving range and statistical constant 1.128. :)
Doug That was great presentation but could not understand the yellow highlighted on may be its to show but when i change the chart has no reaction and the avreg mean doesnt change can you explain
Muito Bom ... Optima explicação. Consegui fazer o exercício visualizando apenas uma vez o vídeo. A dificuldade que tenho na língua inglês não constituiu uma barreira para entender como fazer o chart control. EXCELENTE ... PARABÉNS E SUCESSOS Muito Obrigada Doug H.
Not a terrible piece on charting, but it is not a control chart. Using 3 standard deviations is an incorrect method for calculating control limits. You have to use a local measure of dispersion; in this case it would probably be a moving range (or maybe you could use a c-chart, if these are counts that could be characterized by the Poisson distribution).
It is nice video program to create control charts. But my problem is that I am the user of this control charts now it is deactivated and no graph of or result checking the interred value. so how to help me.
Hallo Mr. Dough H, Thank you for posting this video. It was helpful for me. Actually I am doing some studies with the help of control chart. I have a question for you: to get the value for UCL and LCl you have multiply with 3 std and you are getting a perfect chart. However, in my case for LCL, when I multiply with 3 std it is showing negative value in LCL. Can you kindly explain me, what does it mean? Is it fine to multiply with 2 std instead of 3 std?
Hi Kunchok, what are you measuring? Or what is the unit of measure? Usually if you a control limit is negative you manually change it to zero (if it is time or length). Note: if you are measuring temperature then negative values are valid.
can this only work with a lot of data? i tried with only 3 row of data and the chart is mess up. i do my control chart yearly. so i have only 3 years of data.
Almost 8 years later and this video is still helping people! Thank you for such an informative and digestible method of teaching me how to do this. MUCH appreciated.
Glad you found it useful!
This video was THE most helpful video out of all the "control chart in Excel" videos I have come across! Thank you!
Perfect demonstration! Easy to follow. Thank you !!
You are welcome!
Thank you SO much. You just turned a very frustrating assignment that the professor showed absolutely nothing about into a great learning experience.
Glad it was helpful!
A very clear explanation of how to create a control chart- thank you!
You're very welcome!
Thank you so much. Your video has made my semester. I have understood everything
Glad it helped!
Probably one of the best lectures in RUclips. ❤️❤️❤️
Hi Nedwin L.H., thanks for the kind words!
A couple other commenters caught this, but this can't be overstated for your viewers: THIS VIDEO CALCULATES CONTROL LIMITS INCORRECTLY AND WILL LEAD TO THE CREATION OF MISLEADING CHARTS THAT ARE USELESS FOR PROCESS IMPROVEMENT.
Using any STDEV function calculates the dispersion of all data relative to the overall mean, which exaggerates the effects of outliers and small shifts, thereby inflating the limits and hiding the signals that they are supposed to detect. The key insight of Walter Shewhart was that our statistics need to account for the element of time: this isn't a sample of numbers representing a population, but rather, an ordered sequence of data produced by a causal system over a period of time.
The CORRECT way to calculate the control limits is to calculate the point-to-point differences (the "moving range"), average them, and divide the average of the moving range by 1.128 (bias correction for subgroup of 1, which is what the individuals chart is), which gives you Sigma. Control limits are 3 Sigma on either side of the mean, unless there is a barrier on one side of the mean (e.g., if your UCL is less than zero when a negative value for the process variable would be nonsensical), in which case you just get a one-sided chart, and you can detect "Rule 1" (Western Electric) signals on only one side of the average.
I would recommend that your viewers look up Dr. Donald J. Wheeler's FREE articles on Quality Digest, where he clearly and accessibly explains the statistical underpinnings of SPC and how to use it to drive process improvements. I would also recommend that your viewers calculate control limits in the way I described, also calculate control limits according to the method described in this video, and then try different modifications to the data stream to experiment with how different kinds of outliers affect the calculated limits in both approaches. It will give you a real hands-on sense of what makes the Process Behavior Chart so unique and why a population-statistics approach falls short.
Matthew Lemieux, that is very interesting...thanks for letting me know.
This is the best training video I have ever seen, and I have seen lots. Doug is great at providing the information in a logical, very understandable way.
Hi Pat Davis, glad you liked it, thanks for commenting!
Thanks for the clear and concise, detailed explanation. It made a world of difference to me.
You're welcome!
Hi Doug, I really appreciate the video and can you please consider next time adding the excel file in the description so we can try it ourselves. Thank you.
Hi Sam Usamah Zagaar, thanks for the comment! Will consider it!
Thank you for posting this video my company loves charts and graphs and that is not an area of my expertise. This was extremely helpful!
:-D
Glad you liked!
Hi sir good night. Thank you so much sir your class nice. And I learned how to make control chart
This is great. Please let me know how I go about the line chart if my LCL is negative. Thanks
Bravo. Your video really helped my assignments.
Thank you in splendour.
You’re welcome, glad it helped!
Great~! It's the simplest way to understand the basic logic of a control chart. Thanks a lot~!
Quite easy to understand and relate. Thankyou
Hi Deepak Thakuri, thanks for the comment!
Thank you for doing such an outstanding job explaining without adding any distractions or going into any tangents. This is perfect for my operations course!
Glad it was helpful!
This is very very nice video, very simple and informative. Such a confusing topic explained in minutes...Wonderful Doug H !!!!
Thanks for your comment!
Thank you! I, as a beginner, understood it very well.
Glad it was helpful!
Excellent vid. Clear explanation and easy to follow. Thumbs up and subscribed.
Awesome, thank you!
Awesome Video Doug. Helped a novice SPC man a bunch.
You're welcome, thanks for the comment!
This is a great video. To keep measuring progress as your data collection goes on, would you just refresh your ranges each time you add more data?
That is correct either by updating the range from your chart or if the source range is using the table format it should update dynamically.
@@DougHExcel that’s brilliant. I have another question, if that’s okay? How do we prevent the UCL and the LCL from moving along with the mean? As the current way would move the LCL and UCL, so how would I keep the control the same? I’m sorry if this doesn’t make sense - but I’m wondering how I can keep the lines constant as the test progresses to measure the performance in regards the initial control levels.
Thank u very much ..💓..I followed the steps an boom ..made my first control chart
You’re welcome!
Made sate homework a breeze. Thanks!
You're Welcome!
Clear and easy. Well done.
Hi B, thanks for the comment!
Really appreciable to explain it in such a simple way
In real application, the mean should not be the population mean but rather the true value of the CRM. Is that correct? Moreover, which standard deviation should be use for 1sd, 2sd and 3sd? Is it from the SD from measurements or SD from the CRM itself?
Thank you so much for the video.. this video helped even a dunce like me to understand with ease..
You're very welcome! You're no dunce...we are all students of Excel 😀🤓📚
Thank you for an excellent tutorial. Since the mean and standard deviation change with the data, how do you identify data that is outside of bounds, hence a suggestion of process instability?
Careful when calculating control limits...it is not technically correct to simply use sample standard deviation. You should really use the short term standard deviation calculated from moving range and statistical constant 1.128. :)
Hi The Engineering Toolbox Channel, thanks for adding to the thread!
Thank You So much .. This was really helpful
You're welcome!
Excellent tutorial, many thanks
Hi chris, glad you liked it, thanks for commenting!
Really Appreciate this! Fantastically explained
Glad you enjoyed it!
love your tutorials on excel, thanks!
Hi Jon Manilenio, glad you liked it, thanks for commenting!
really good video, thanks a lot!
Thanks!
thanks your explanation has really helped me big time
Thanks Ngosa Royd Mutale, glad it helped!
@@DougHExcel I can't thank you enough...
Excellent Video, very very very helpful and explained so easy way. Millions of Thanks
You're welcome, thanks for the comment!
Very helpful, good job.
Glad it was helpful!
Thank you so much for this tutorial, I was able to create a control chart for my process variation assignment
Hi Susan Wanjiku, glad you liked it, thanks for commenting!
well explained....and easy to follow also..thanks...
Hi kishore pk, thanks for the comment!
Thank you so much sir for sharing this. This really helps a lot.
You are most welcome
Is multiplying and dividing by 3 for control limits consistent across all control charts?
This was great I was looking for a minitab example and came across this. I really like how you did it. Good job sir!
Thanks this was great!!!
Hi AB Newman, you're welcome!
Thank you! very clear and easy to follow. Question: what type of control chart that is?
Hi Haasan, thanks for the comment!
just what I needed, this was informative and I am glad you took the time to show us how to do everything!
+Yasmin Ali glad you liked!
Thank you so much. This video is very helpful.
You’re welcome
Thank you! Helped me create a control chart for my operations management assignment. Easy to follow and everything worked on one try! Awesome
Also I learned new things in excel =>
Glad it helped!
Excellent video. I have a question....If I have a goal this will replace my mean?
Thanks! I will review them.
PD: I love your videos. Your awesome teacher.
A great help..and a very nice way to make me understand ..thank you so much...Harish
You're welcome!
Thank you so much !! super clear !!
Hi MIGUEL MORA MENDOZA, glad you liked it, thanks for commenting!
Thank you very much.. it is really helpful
Thanks Roula Kholoud, glad it helped!
hello please tell what measurement is in this , it is cusum points, or x-t values or simple x
This helped very much. glad i found it.
Nice job, simple and clear.
Thanks,
Hi jhourani, thanks for the comment!
great work, very clear!
Hi aaryanna nijjar, thanks for the comment!
Excellent. Very helpful. Thanks
Hi Dr sai mangala Divi...you're welcome, glad you liked!
Awesome tutorial Doug. Thanks!!!
You're welcome!
Doug That was great presentation but could not understand the yellow highlighted on may be its to show but when i change the chart has no reaction and the avreg mean doesnt change can you explain
Hi Nirmal Joseph, sorry 🙁....but try a post on the mrexcel.com forum!
Simple and great
Thank you! Cheers!
That was great , this will really help me. Thank you
Hi @Dough H, I learned that there are many types of control charts. May I know what kind of control chart is this?
It’s closest to the Shewart chart
Muito Bom ... Optima explicação.
Consegui fazer o exercício visualizando apenas uma vez o vídeo.
A dificuldade que tenho na língua inglês não constituiu uma barreira para entender como fazer o chart control.
EXCELENTE ... PARABÉNS E SUCESSOS
Muito Obrigada Doug H.
Doug did you create a demonstration of the control chart in Microsoft 16 yet? Would be a good reference.
Hi Jess Cotten, thanks for the comment!
Thank you very much for sharing this instructional video.
You're welcome!
Very useful Vedio....I m happy to learn from it
Thanks Dipak Pingale, glad it helped!
THANK YOU SO MUCH FOR THIS.
You're so welcome!
Not a terrible piece on charting, but it is not a control chart. Using 3 standard deviations is an incorrect method for calculating control limits. You have to use a local measure of dispersion; in this case it would probably be a moving range (or maybe you could use a c-chart, if these are counts that could be characterized by the Poisson distribution).
Rip Stauffer, that is very interesting...thanks for letting me know.
Excellent ..Very helpful
Thank you Sir
Very good, thank you.
You're welcome!
yes this helped for sure, learned some new excel tricks :) thanks
Hi PetStuBa, glad you liked it, thanks for commenting!
Great advice on how to navigate around MS Excel for Control Charts! Was a lifesaver for a project I am working on. Thank you!
Nice! Can I change the days to months and also use larger nos. Eg - 70/, 80/, 90/
Yes, here's some videos for ideas ruclips.net/user/dough517search?query=axis
very helpful.. but i want to manage the data limits . dont want to starts from zero i want my own limits to set.
It is nice video program to create control charts. But my problem is that I am the user of this control charts now it is deactivated and no graph of or result checking the interred value. so how to help me.
Hi tolina jira, sorry I don't do consulting :-( ....but try a post on the mrexcel.com forum!
Thanks Doug, really helpfull
Hi Fish, you're welcome!
How impossible you can make sb. who's low in excel understands how to do the control chart this easily? Thank you.
You can do it!
Can I do the same if the data is in line in place of columns?
Sorry, don’t understand the comment/ question...
Hallo Mr. Dough H,
Thank you for posting this video. It was helpful for me. Actually I am doing some studies with the help of control chart. I have a question for you: to get the value for UCL and LCl you have multiply with 3 std and you are getting a perfect chart. However, in my case for LCL, when I multiply with 3 std it is showing negative value in LCL. Can you kindly explain me, what does it mean? Is it fine to multiply with 2 std instead of 3 std?
Hi Kunchok, what are you measuring? Or what is the unit of measure? Usually if you a control limit is negative you manually change it to zero (if it is time or length). Note: if you are measuring temperature then negative values are valid.
Hey Daniel Solomon, in the video he has multiplied mean with 3. Is that a relation to get SD?
yes, me too!
Hi Kunchok Tharlam, thanks for the comment!
thank you soo mch Mr Dough
You're Welcome!
this is really helpful. is it possible to do this with spss??
sorry I'm not familiar with spss; though it's an enterprise level stat package, I'd be surprised if it didn't have this feature.
Great video!
Thanks!
thanks this was a great help
Glad it helped!
Helpdful thanks
You're Welcome!
Is there a tutorial video about how to make inventory control chart in excel, anyone knows? Please help me.
Hi Muhamad Ramdan, thanks for the feedback!
Very helpful
can this only work with a lot of data? i tried with only 3 row of data and the chart is mess up. i do my control chart yearly. so i have only 3 years of data.
Hi Zack Neo, thanks for the comment; yes it could but it'll be crowded.
Thank you so much.. this really helps.. Awesome, cheers..!!
+Arunagiri P Glad it helped you out!
Hi,
This is an excellent video. Liked it very much.
Can I use it in my training presentations, with due credits to you?
Request a revert pls.
Thanks
Hi Sadashiv Borgaonkar, thanks for the comment! Sure thing feel free to use it with the attribution.
Thank you for the kind permission. Will do the needful definetly
Thanks for nice video
Hi Sucharita Mitra, thanks for the comment!
Thankyou, really helpful.
You're welcome!
THANK YOU
You're Welcome!
Great , thank you
You're Welcome!
fantastic!
is the 3 sigma fixed value or it will depend on the std dev?
std dev
Thanks a lot
Thank you..
You're Welcome!
Thank You
You're welcome!
Thank you much.
Is this chart known as that I-MR Chart?
how do press F4 on macbook
Try CMD+T