How to perform precision analysis and ANOVA having four nested factors?

[#1]

Hi!

I am quite new to JMP and I would really appreciate your input regarding the following analysis.

I have to establish the precision of an assay.

I performed a test that consisted in measuring my sample in x replicates, twice a day for 20 days, on 3 instruments located in 2 different sites (1 in site A, 2 in site B; site A had different environmental conditions and operators performing the test compared to site B).

I assume that these 4 factors are nested as follows:

• Site
• Instrument (nested in Site)
• Day (nested in Instrument and Site)
• Run (nested in Day, Instrument and Site).

I would like to have the following outputs (expressed as CV%):

• Precision within the same run
• Precision between runs
• Precision between days
• Precision within the same instrument
• Precision between instruments
• Precision between sites
• Total precision.

Can this be done in JMP? What approach would you recommend me to use?

I am also supposed to analyze the data via ANOVA. Would I have to perform a four-way nested ANOVA? How can I do this in JMP?

Thank you very much in advance.

C.

[#2]

First, layout out the JMP data table so that each factor and the assay are in separate columns. You can use the Variability Chart platform or the Fit Least Squares platform to estimate the variance components with the nested model.

I attached a data table that is probably like your study so that you can see how to set up your data. I saved a script for the two platforms so you can study the results. The Assay data is simulated with all the nested factor random effects equal to zero.

[#3]

Hello,

I am also somewhat new to JMP and have a similar question as above. I did try using your data table and scripts

that you provided, but had trouble modifying the table "Assay" column.

I would like to determine mean, SD, and %CV.

My experiment was similar but was only within one site

-3 Samples (2 level for each)
-4 replicate per day per sample
-2 shifts per day (4 runs per shift)
-12 days (8 runs/day)
-2 operators (8 runs/day total)
-1 site
-2 instruments (1 instrument per operator)

I would like to determine mean, SD, and %CV for the following with the following assumptions:

-Within site (nested with: b/w operator, b/w day)
-Within Operator (nested with within instrument)
-Within Day (nested with: b/w run)
-Within Shift
-Within Run
-Total

Can you please direct me how to do this within JMP?
I realize that it will likely involve a very similar approach to the last question.
Can I simply modify the table you provided or do I also need to modify the scripts?
Again I tried modifying the table but the "Assay" column was locked.

This is the table I hope to create.

Sample

N

Avg Result Value

Within Site

Within Operator

Within Day

Within Shift

Within Run

Total

SD

%CV

SD

%CV

SD

%CV

SD

%CV

SD

%CV

SD

%CV

AL

96

AH

96

BL

96

BH

96

CL

96

CH

96

Thank you in advance,

Mark

[#4]

Sorry the table didn't past correctly. I've attached it.