I-mr VS X-mr Chart - What's their difference?

Hi experts,
I have a Q regards to IMR chart. I went around to look in the net and found there are 2 versions of it.

1) IMR only can have individual data which means we will tk the constant which carries n=2 only everytime.
2) IMR can have subgroups. thats the reason we have a constant table for individual with sample size till 10 (the table of constant we I got it from (ASTM publication STP-15D, Manual on the Presentation of Data and Control Chart Analysis, 1976; pp134-136.)

My Q, which is the right one?If the version 2 is right,I wonder how the way to calculate the UCL & LCL and how this will differ with the Xbar R chart or X bar S chart (which we knows carries the subgroup ).Thanks a lot for your explanation.

2) IMR can have subgroups

Click to expand...

Wrong, are subgroup charts, not IMR (Individual moving range)
There are many different kind of control charts, the individual a point on the chart represent one data point, the subgroups charts one point represent several data points

with sample size till 10

Click to expand...

We are now talking about X-Bar & R Control Charts, if the sample is bigger, X-Bar & S Control Charts.

Wich is better?, it depends on the sample size, and the sample size depends on the process wich is meassured

Good Day Sitiz,

Not sure I understand your question, if it is in fact when to use a I/MR chart vs X Bar/R chart; here are some of my notes/guidelines, if you use Minitab, the control charts tutorial is very good at drawing a dinstinction

"When to Use an Individual-X / MR Chart
Individual-X / Moving Range charts are generally used when you can't group measurements into rational subgroups, when it's more convenient to monitor actual observations rather than subgroup averages, or when the process distribution is very skewed or bounded. Each subgroup, consisting of a single observation, represents a "snapshot" of the process at a given point in time. The charts' x-axes are time based, so that the charts show a history of the process. For this reason, you must have data that is time-ordered; that is, entered in the sequence from which it was generated. If this is not the case, then trends or shifts in the process may not be detected, but instead attributed to random (common cause) variation.

Advertised Reason for using Individual charts– Where the cost of inspection is high such as when destructive testing is involved…..Real Reason it is a tool with which to loosen the criteria for assessing stability, particularly if small mean shifts are actually common over time."


And by extension X-bar R charts are "typically" used when:
- X-bar/R-Chart – when subgroup size is between 2 & 10 use X-bar/R-chart
- X-bar/S Chart – when subgroup size is > 10 use X-bar/S-chart

There is a vast pool of academic and professional knowledge here, hopefully others will add to or revise my input.

Hope this helps.....

Regards,
Marty

The confusion may arise from the Moving Range sample size. While the X - chart is always an individual point, the Moving Range may be calculated from 2 successive points, 3 successive points. n successive points.

Miner said:

The confusion may arise from the Moving Range sample size. While the X - chart is always an individual point, the Moving Range may be calculated from 2 successive points, 3 successive points. n successive points.

Click to expand...

OK, But I don't remember to have seen tables for constants for such charts (moving ranges for more than two data points), can you post the tables or a link to them?

Darius said:

OK, But I don't remember to have seen tables for constants for such charts (moving ranges for more than two data points), can you post the tables or a link to them?

Click to expand...

Don't have a table I can post but they are in the AIAG SPC manual, 2nd edition (probably other editions too but thats what I have.......)

Darius said:

OK, But I don't remember to have seen tables for constants for such charts (moving ranges for more than two data points), can you post the tables or a link to them?

Click to expand...

This is a link to an article from George Mason University and table for I-MR constants for moving ranges of varying sizes.

Thank you all for your kind reply but I think we still have not come to an conclusion.
Some saying IMR not subgroup chart.
Miner,
The table is the same table I was talking on the ASTM std. But question here is we still use the constant n=2 from the table.So when actually we will use the other sample size from the table for IMR/XMR chart?
Sry guys coz I am still not clear on this.

The X or individuals chart is ALWAYS one single data point plotted on the chart. The n number comes from how many consecutive points are used to calculate the Moving Range.

In other word, the x chart is plotted using individual points (n = 1) while the MR chart is plotted using subgroups of n = 2 or more.

One thing that made a lot of sense to me when being taught the IMR and other control charts was that when you look at subgrouped data - say on an X-bar/R chart, the questions you've asked of your process are these:
What's changing between subgroups? (differences in the X-bar chart)
What's NOT changing between the subgroups?
What's changing within the subgroups? (look at the R chart here)
What's NOT changing within the subgroups?

So it's a "within v between"discussion on process and sampling variation.

When you don't have subgrouped data and are looking at the IMR charts, the questions shift a little and you're now asking about longer term variation on the I or X chart and about short term variation on the MR chart. The reason you can select n=2, n=3 n=6, etc for the MR chart is because you need to define what short term really is for your process.

If you're in an industrial setting and looking at an assembly process, n=2 is pretty common. We used to look at banking center activity where the feature of interest might happen every 5th day, so we set n=5 for our charts - just as an example.

Hope it helps!

Cheers,

Adam