Compare curves in SAS JMP

[La82]

Hi Everyone,

I am trying to compare a set of curves against a set of references.

The curves are determined using fixed x-points. What would be the correct way of analysing this? And if yea is there a way to do this in SAS JMP?

Best Regards

Lonnie

 

[hlsmith]

Not a JMP or curve person, but what are you trying to test or get out of this?


Different slopes, different areas under the curves, etc.?


Also, can you talk about these data in more detail?

 

[La82]

Hi,

Me neither - but trying to get there

I am doing a test where I get Measurement at 5 different frequencies (frequency sweep). I have 9 approved curves at this frequency, and I want to test if the 3 replicate Measurements I have made on a different material fits the references. So I could do a t-test at each frequency to test if the Measurements were ok, but I would like to test all 5 (the entire curve) at once.

I have tried a nested design, but this still checks the interaction between the frequencies which I know is going to be a significant factor since this is the purpose of the test.

I don't know if this makes sence, but pleace dont hesitate to ask more quenstions

Best Regards

Lonnie

 

[hlsmith]

I think you are using industrial language that I am not following.


So you are interested in the between group effects not the within group effects?

 

[La82]

Hmmm I think so, but I must admit I am not completely sure, my statistics is a bit rusty

But I have two goups I would like to compare across the same 5 datapoints (1, 2, 3, 4, 5).

In Group 1 i have 9 repetitions pr datapoint (9 x 5 = 45 observations), in Group 2 I have 3 repetition pr datapoint (3 x 5 = 15 observations).

Does that make more sense?

 

[hlsmith]

So you have two groups (your variable of interest) for 5 data points. And at each data point you have multiple measures from the two groups. Though the number of observations is not consist over time, but it is consist between the two groups (e.g., 5:5, 3:3, etc.).


Now what is your study question, exactly?

 

[La82]

Yes that is correct... And I would like to see if there is a difference between the two Groups at all 5 data points together...

 

[dungdn93]

Not a JMP or a curve, but what you're trying to check out, or get out of this. So you have two groups (your variable interest) to 5 data points. And at each data point you have multiple measures from two groups