Interpreting PROCESS mediation output (help!)

Ayuk

New Member
#1
Hi there! I'm using the PROCESS tool in SPSS to test the following mediation model:

X (Mindset) ---> M (Coping strategy) ---> Y (Quality of life)

All variables are measured on a continuous scale.

I'm Using ''Discovering statistics using IBM SPSS'' by Andy Field as my bible, and now i am trying to interpret my output. However, I found a discrepency between the way the output is presented in Field and the way it is presented in my analysis. I followed the exact same steps.

Field: In the second model, where both X and M are added to the model, the mediator is the first variable added to the model. Therefore, we can conclude whether X significantly predicts Y, even with M added to the model.

My output (see photo): in this same model, where both X and M are included, the X variable was added to the model first. Therefore, I can only conlcude whether M significantly predicts Y, with X added to the model?

This doesn't make any sense to me and I am also wondering why the X and M variable switched places in my output. I see output similar to Field's when I look at other examples of PROCESS mediation analyses.

Someone please help!
 

Attachments

Karabiner

TS Contributor
#2
It is not quite clear to me what your problem is. In the first model, Y is regressed on the predictor X, and predictor X has a small regression weight which is not statistically significant. So there's no need to add a mediator. Anyway, after adding M, predictor X is still not significant, while M has a statistically significant relationship with Y. Don't know what you mean by switching places, though.

With kind regards

Karabiner
 

Ayuk

New Member
#3
Thank you for your reply. You say that after adding M, predictor X is still not significant. However, it seems as though in the second model, the X variable is added first. If I'm reading it like I would a simple regression, I would interpret that as: after adding X (mind_TN) mediator M (COP) is still significant. Tot conclude what you concluded, we would expect to see the M variable (COP) first, followed by the X variable (mind). That's what I meant with 'switching places'
. Maybe I'm overthinking or reading it wrong, would love to hear your thoughts.