Linear Mixed Model

[Andreas Erik Lindahl]

Hi

I'm doing a LMM for the purpose of investigating if burn size (exposure) has a causal effect on cortisol slope (Outcome).

Burn size is a between-subjects variable, it does not change over time for a given patient.
cortisol slope is a within-subjects variable. It is measured each day for seven consecutive days in every patient.
I also put other covariates in the model to be adjusted for. Some are measured daily like cortisol slope. some are fix like burn size.

Also, burn size is dichotomous. cortisol slope is continous. the other covariates are in some cases dichotomous and in some cases continous.

My question is:
What output from the model is relevant to present in a scientific paper using this model?
I look at the p-values, and when significant for burn size, I interpret this as: Burn size explains changes in cortisol slope.
What other than the p-value is of importance to a referee and reader?

regards,
Andreas

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Bruce Weaver]

Andreas, your question has nothing to do with SPSS.

Have you looked at other articles that use this type of model?  Depending on your discipline, their may be a set of guidelines (e.g., the APA Publication Manual for Psychology).  If "scientific paper" means journal article, you should also consult the Instructions for Authors on the journal's website.

Andreas Erik Lindahl wrote

--- snip ---
My question is:
What output from the model is relevant to present in a scientific paper using this model?
I look at the p-values, and when significant for burn size, I interpret this as: Burn size explains changes in cortisol slope.
What other than the p-value is of importance to a referee and reader?

regards,
Andreas

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Ryan]

The first question you should ask yourself is whether you parameterization the model correctly. I notice you did not provide the linear mixed model/equation or syntax. How much do you know about linear mixed models? Why have you decided to employ such a model? How exactly do you interpret the parameter estimates derived from the model you parameterization. My point? You might want to start by studying/reading about the theory and application of linear mixed models before fitting one and being concerned about which output to report. To a large extent, you will know what to report when you understand the  model you have employed.

Ryan

On Jan 17, 2013, at 4:04 PM, Andreas Erik Lindahl <[hidden email]> wrote:

> Hi
>
> I'm doing a LMM for the purpose of investigating if burn size (exposure) has a causal effect on cortisol slope (Outcome).
>
> Burn size is a between-subjects variable, it does not change over time for a given patient.
> cortisol slope is a within-subjects variable. It is measured each day for seven consecutive days in every patient.
> I also put other covariates in the model to be adjusted for. Some are measured daily like cortisol slope. some are fix like burn size.
>
> Also, burn size is dichotomous. cortisol slope is continous. the other covariates are in some cases dichotomous and in some cases continous.
>
> My question is:
> What output from the model is relevant to present in a scientific paper using this model?
> I look at the p-values, and when significant for burn size, I interpret this as: Burn size explains changes in cortisol slope.
> What other than the p-value is of importance to a referee and reader?
>
> regards,
> Andreas
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD