231383_en-US

liliconcom

Re: model equivalency with SAS

Hi @Woodstock 

my best guess is that your issue is caused by the way jmp builds the contrast across ordinal and nominal independent variables categories.

have a look at these two links:

https://www.jmp.com/support/help/14-2/coding-for-nominal-effects.shtml 

 

https://www.jmp.com/support/help/14-2/models-with-nominal-and-ordinal-effects.shtml 

 

first compare models with just continuous independents. then, when you add binary independents define them as ordinal in JMP. this should produce parameter estimates that are equivalent to SAS (in the case of 0 and 1).

if you have a nominal effect with many categories, you can produce n-1 dummy variables in JMP (as is done the old fashion way in SAS) and define each one of them as ordinal.

since you mentioned the factor of 2 this is most likely the case.

you can save the prediction formula to the table or ask for the prediction expression. this may help clarify how JMP is estimating the equation.

https://www.jmp.com/support/help/14-2/show-prediction-expression.shtml

also have a look here:

https://community.jmp.com/t5/Discussions/Logit-why-aren-t-my-exponentiated-coefficients-equaling-my-odds/td-p/15661 

overall, prediction and conclusions should be identical between SAS and JMP (or any other software for standard procedures).

Re: model equivalency with SAS

See Help > Books > Fitting Linear Models > Statistical Details. It explains the differences between the way JMP and SAS parameterize linear models. I would not expect the estimates to be the same. They estimate different parameters.

model equivalency with SAS

HI there,

I have a mixed effects model which I am trying to show equivalence  with SAS for... Was hoping to get a better understanding of how to do that... 

 

JMP:

 

Fit Model(
Y( :log_val ),
Effects( :Group, :Analyst ),
Random Effects( :Sample[:Group] ),
Personality( "Standard Least Squares" ),
Emphasis( "Effect Leverage" ),
Method( "REML" ),
NoBounds( 0 )
)

 

 

SAS:

proc mixed data=WEBWORK.MYDATA method=reml plots=(residualPanel) alpha=0.05;
class Group Analyst Sample;
model log_val=Group Analyst / solution;
random Sample(Group) /;
run;

 

My parameter estimates for fixed params is off by a factor of 2 in some cases

Re: model equivalency with SAS

Hi Mark,

Thanks for the reply. Help me understand what you mean by "I would not expect the estimates to be the same. They estimate different parameters.", Given I need to show equivalency between parameter estimation, what are the differences, and how important are they to the overall, and individual estimations?

 

I found this: https://www.jmp.com/support/help/14-2/nominal-factors.shtml#65535 from your Help menus breadcrumb suggestion, but I'm not sure how to use it to obtain equivalence between the two platforms.

 

Thanks for your time

 

Re: model equivalency with SAS

Hi @ron_horne ,

Thanks for your reply. I'm wondering what you mean by the statement: "you can produce n-1 dummy variables in JMP (as is done the old fashion way in SAS) and define each one of them as ordinal."... specifically, how do I do this? 

 

When I export the model specification directly from JMP as SAS code, I don't get the same results. See attached

Re: model equivalency with SAS

Hi @Woodstock 

as for creating dummy variables from a categorical column try the following:

https://community.jmp.com/t5/JMP-Academic-Knowledge-Base/Make-Indicator-Dummy-Variables/ta-p/22636 

with respect to the output that you sent, perhaps someone else will answer.

ron

 

Re: model equivalency with SAS

@ron_horne ...  I should have GTS'd first instead of asking... I found the same page about the same time you replied... Sorry about that. Handy little add-in

Re: model equivalency with SAS

When you fit the model in JMP, go to the red triangle and choose Estimates > Indicator Parameterization Estimates. That will give you a report that matches SAS.

JMP output:

 

SAS output:

Re: model equivalency with SAS

Hi @ron_horne, The dummy approach got me equivalent estimates compared to SAS across all levels and factors. Appreciate your help.

 

One question: the output indicates that the results are 'biased'. What does "biased" mean in the context of the JMP output, relative to the dummy variable approach (numeric, ordinal), as compared to the character, nominal approach which gave the difference between JMP and SAS? (side note: the SAS code generated from within the Model Specification window gives different results between JMP and SAS) 

 

Interestingly, R shows the same results as well, except for the intercept. (Group is a factor) 

Re: model equivalency with SAS

you are welcome @Woodstock ,

as for the bias issue, i would assume at least one of the categories has just one observation or that two categories are actually the same observations.

see if this discussion helps:

https://community.jmp.com/t5/Discussions/I-am-receiving-Lost-DFs-Biased-and-Zeroed-parameter-estimate/td-p/13769 

Re: model equivalency with SAS

The models are equivalent in their predictions. Given the same predictor values, both models predict the same response. The models are different in their parameterization. I do not expect the same parameter values.

 

If I fit a simple linear model with one term for a continuous predictor, then center and scale the predictor, and re-fit the model with the transformed predictor, the parameters are necessarily different yet the two models predict the same response. The parameters are different. The interpretation of the parameters is different.

 

 

I do not like the previous reference to the "old fashioned" parameters in SAS. They are simply different. They represent a different interpretation and test of the effects of changing the factor level. No parameterization is inherently superior. The best parameterization depends on how you think about the effects. The choice is situational.

 

This issue is the same as the one about different types of sum of squares. Most of the time I prefer Type III. That doesn't mean that Type I are "old fashioned" or wrong or of no use. There are times when they provide exactly the right information, like the test for poolabilith of multiple batches in determining the expiry of a drug. It just depends.

Re: model equivalency with SAS

The matter of parameterization is covered in most textbooks about regression models. They also cover the least squares solution and the appearance of biased or zeroed parameters. JMP help also covers this situation. It says, "Note: Estimates are obtained and tested, if possible, even when there are linear dependencies among the model terms. Such estimates are labeled Biased or Zeroed."

 

It also defines the conditions that lead to such a result:

 

"When XX is singular, a generalized inverse is used to obtain estimates. This approach permits some, but not all, of the parameters involved in a linear dependency to be estimated. Parameters are estimated based on the order of entry of their associated terms into the model, so that the last terms entered are the ones whose parameters are not estimated. Estimates are given in the Parameter Estimates report, and parameters that cannot be estimated are given estimates of 0.

 

However, estimates of parameters for terms involved in linear dependencies are not unique. Because the associated terms are aliased, there are infinitely many vectors of estimates that satisfy the least squares criterion. In these cases, “Biased” appears to the left of these estimates in the Parameter Estimates report. “Zeroed” appears to the left of the estimates of 0 in the Parameter Estimates report for terms involved in a linear dependency whose parameters cannot be estimated."

 

So your choice of parameterization 'over-specifies' the model given the data or, as Ron expresses it, you do not have sufficient levels to estimate all the effects.