What goes wrong in analyzing Hayes PROCESS 4.1 model 1 (Graph)?

19 May 2022 4 1K Report

Hi all,

For some time now, I cannot figure out what is going on with a moderation model (model 1) with a covariate that I'm running.

IV = categorical, two groups, dummy coded (lockdown (T2) or no lockdown condition (T1))

DV = continuous (mental illness)

M = continuous (resilience)

Covariate = gender (dummy coded)

I set the conditioning values at 16th, 50th and 84th percentiles. Further, I mean centred the continuous variables.

The overall model is significant.

When I run the data for visualizing the conditional effect of the focal predictor, I get a graph that I do not think is right. Resilience is pictured on the x-axis, mental illness on the y-axis, and the lockdown or no lockdown condition is noted as two variables (0 and 1) on the upper right side of the graph. I always thought that that is where the moderator should be, not the IV. That's why I'm confused.

Can somebody explain to me what is going on or what I perhaps should do differently in the analysis? A big thank you in advance :)

Andrew Seidman

Evelien Ridder I wish I could be more helpful, but this model seems to be set up appropriately. I have not used categorical variables as an IV in PROCESS Model 1, but I have as a moderator (I assume PROCESS Model 1 allows categorical variables as an IV), and it required me to use a multicategorical coding system (i.e., Helmert, sequential coding) within PROCESS to do so that was available within the dialog button. My best guess is that there is something going on with your dummy-coding that is confusing PROCESS in terms of interpreting the moderation. You can check out this video https://www.youtube.com/watch?v=r2_zw4G55X0 for insight into coding the IV (it deals with mediation, but I would imagine the logic still holds for coding). Also, Andrew Hayes is very responsive via e-mail and could shed insight as well if the video is not helpful.

Evelien Ridder

Andrew Seidman I watched the video and found where I was looking for, as it was mentioned in the comments: The simple slopes are often separated by a categorical moderator rather than by the standard deviation. That would explain why the graph was set up differently maybe. Thank you very much for your help.

What is the ideal sample size for a Between-Subjects factorial design?

What is your perception on using visual analysis or quantitative approaches to evaluate data from single-case research designs (n of 1 trials)?

What SPSS analysis should I run with one IV (with two levels) and two DVs?

How can I make stacked bar graph with scatter plot?

Mediation path analysis - Can I run two seperate models?

SEM or Multiple Regression for an analysis without Latent Variables?

How can I set up a good fit SEM model in AMOS?

What could be reason for getting such type of graph for complex impedance in case of polymer blend films?

Can I use a graph from a another source in a thesis?

I have need of XRD graph analysis software for phase identification. Can you please provide links for software tools which can be downloaded?

Are there lab markers/pens resistant to Roti-Histol?