What inspired this Wish List Request?
This is just my personal suggestion.
Please describe the new or improved feature or solution you would like to see.
Fit Model's Response Screening supports Switch columns. In the results, it outputs all tests for both model effects and Switch variables. But I would like to get FDR-adjusted test results only for Switch variables.
I assume the following senario. A healthcare scientist wants to get FDR p-values for age-gender adjusted risk factors. They don't care about the effect of age and gender because they are just covariates (variables for adjusting). They are interested only in risk factors.
Please run the following script. It outputs all tests for both Age, Gender and Risk Factors 1-10. I would like to get only FDR p-values only for Risk Factors 1-10.
Random Reset(11111111);
age = J(1000, 1, Random Integer(20, 70));
gender = J(1000, 1, Random Integer(0, 1));
x = (age||gender)*J(2, 10, 1) + J(1000, 10, Random Normal());
y = age + gender + J(1000, 1, Random Normal());
dt = As Table(age||gender||x||y);
Column(dt, 1)<<Set Name("Age");
Column(dt, 2)<<Set Name("Gender");
Column(dt, 2)<<Modeling Type("Nominal");
For(i = 3, i <= 13, i++,
Column(dt, i)<<Set Name("Risk Factor"||Char(i-2))
);
Column(dt, NCol(dt))<<Set Name("Y");
dt << Fit Model(
Effects( :Age, :Gender ), Personality( "Response Screening" ),
Y( :Y ),
Switch(
:Risk Factor1, :Risk Factor2, :Risk Factor3, :Risk Factor4, :Risk Factor5,
:Risk Factor6, :Risk Factor7, :Risk Factor8, :Risk Factor9, :Risk Factor10
),
Run
);
What would the value be to you and/or other users if this new feature is provided?
It might be useful when epidemiologists find new risk factory in their exploratory research.
Is this an external customer request?
No, it is not.
Is there a competitive software package that has this feature or solution that we should be aware of?
I do not know.
Will this impact a renewal, expansion, or adoption at a customer, and does this put an account at risk if we do not address this?
No.
