[Feature Request]: Table of excluded variables in stepwise linear regression

[jirigumancik]

Description

When conducting a stepwise linear regression, I would like for JASP to generate a table with variables that were excluded from the final model. Something like SPSS does via the 'excluded variables' table.

Purpose

Reporting of non significant values

Use-case

No response

Is your feature request related to a problem?

No

Is your feature request related to a JASP module?

Regression

Describe the solution you would like

When conducting a stepwise linear regression, I would like for JASP to generate a table with variables that were excluded from the final model. Something like SPSS does via the 'excluded variables' table.

Describe alternatives that you have considered

No response

Additional context

No response

[tomtomme]

@jirigumancik
just to be sure - the note below the table is not sufficient?

what exactly would you want to see in said table. can you paste a picture of how spss is handling this?
thx

[jirigumancik]

Hello,
So this is what I would like to see, which is the table will all the variables which were excluded from the stepwise regression. I need to see the non-significant values so I can report them as well. JASP does not produce this.

[jirigumancik]

Has this been looked at yet?

[TarandeepKang]

I don't think it has been looked at yet, else Johnny,- who has been assigned to this- issue probably would have responded with his thoughts. I see that you have mistakenly closed this issue and open another one on the same topic?

However, I am now going to close your other issue because it was assigned to the wrong module i.e. ANOVA instead of regression while I will ask a team member who can reopen this issue @tomtomme to do so for me and @JohnnyDoorn to provide an update for you if and when he is able on your request?

Thanks in advance colleagues.

I'm sure we will get you an update soon although I am only a project volunteer and not a full team member, so let's wait to hear from one of them!

Tarandeep

[JohnnyDoorn]

Hi @jirigumancik,

Sorry for my late response - I had my summer break and then got hit by a teaching tsunami.. Is your case about using the forward stepwise function? Because if you use the backward function, you do get a table with the coefficients of all specified predictors, which then gets trimmed one variable at a time. Is that something you can work with?

Kind regards,
Johnny

[TarandeepKang]

Hi Johnny, I'm not the OP but I think we are having some crossed wires with terminology! :-) So this case relates to (I think) the hierarchical functionality using blocks that you implemented a couple of versions ago. If you do this in SPSS using the default forward entry method and some variable(s) is/are rejected, you get a table of variables such as the one given above. It is my understanding that this is what is requested here.

This would make the the same as SPSS. To be clear I think it produces the same kind of table for the other entry methods, but I've never actually used them.

Best,

Tarandeep