Is there a way to calculate 1 sigma shift?

[#1]

I have a data set that includes 20 part and each part is measured 20 times for its thickness. I also have a reference mean and std.I would like to compare each part's mean to the reference mean and to see if their difference is more than my reference point's std. Is there a way to do this in JMP? I am using JMP 11.2. Thanks.

[#2 · Accepted solution]

There are a variety of ways one could approach this.  My proposal would be to go to the putlldown menus and select Tables==>Summary.  There you select the 20 different collumns of data and specify to calculate the mean of each column.  It will create a new table with one row, which will have one column for each input column.  Then, a new column can be added for each of the columns, and then specify a formula for each column, which makes the comparison to see if the reference std is greater than the absolute difference between the calculated mean and the reference mean.

Rather than doing separate calculations I would add a step that would allow the calculation to only have to be specified once.  What I would do, is once the summary data table is created, I would then go to Tables==>Stack and stack all of the summary columns.  Once this is done, only a single additional column would have to be created to determine if the reference std was greater or less than the difference .

[#3]

Hi Jim,

Thanks for your response.

My data is in two columns. Below is an example of how my data looks like.

Part     Value

1           1.2

1           1.3

1           1.5

1           1.4

1           1.6

2           1.3

2           1.2

2           1.5

2           1.6

2           1.6

I think the Summary works. I can summarize the mean and std from each part, then create a column with a formula to calculate the sigma shift. Thanks a lot.