How to test this model?

[naamalo]

Hi everyone,
I wish to know how can I statistically test this model (better with spss process if possible) that was suggested to me be one of my teachers.
the data has been collected with self report questionnaire
participants reported at the same time about depression, which ways of coping they have used, threat perception, and another two personality traits. all variables are continuous variables.
the suggested model is that coping mediates the relationship between threat perception and depression ( as well as anxiety, in separate model that is identical)
and that personality and threat appraisal have mutual effect on each other (correlation? ) and at this point we can say that one of them was before the other, so they both in the same "time" at the sequence (it is correlation model and mediation, no casual due to the method)
please help me to know how can I test this model (and how to report according to apa style
thanks!

 

[hlsmith]

First flip this sucker around so it goes from left to right. Also, per this model the three nodes (exogenous terms) have no direct effect on the outcome, correct? Are you sure you want to commit to that? So they aren't partially mediated, they are fully mediated. Also, doe the top exogenous term have an effect on the bottom one? If so this isn't in the model as well.

Thanks.

 

[naamalo]

sorry, it is partial mediation you are right
I've corrected it

 

[hlsmith]

Also, I am not sure if the bidirectional heads will make this complicated as well. For additional clarification, does trait one also need an arrow to threat perception?

Of note, I am just trying to help you better define this, though I have never done mediation analysis with trying to address multiple variables mediated through a single mediator - I am guessing it isn't too much different.

@spunky any input on this post?

 

[EagerInStatistic]

Dear Naamalo,

Put it clear, your main objective is to examine coping as a mediator between threat perception and depression/ anxiety and depression. Also, you would like to see the effect of personality traits in this model. In my opinion, personality traits seem like more stable compared to perceived threat which can be changed by situational factors. Therefore, I do not expect a mutual effect between personality trait and threat perception or anxiety.

If I were you, I would use personality traits as moderator variable that may affect the relationship betwen X (perceived threat) and M (Coping). This model is called "moderated mediation". For the first stage, you could separate your model into some small parts to easily interpret and comprehend your model and then integrate them together. Below, you can see some bacis models and for finding more models you can have a look at Hayes' templates in order to clarify your model in your mind. About how to report conclusions of the analysis, you can check attached file.

kind regards

 

[Karabiner]

This is a bit surprising, though. The theoretical model would say that
threat perception causes coping, and coping causes depression?
I have never heard that coping is a causal factor for depression.
I would rather assume that coping acts as a suppressor.

With kind regards

Karabiner