SPSS - Mediation with PROCESS Categorical Variables (Model 4)

Hi, could you advise how to calculate effect sizes? If I understand the output correctly, there can be find R2 for total effect and direc effect, but what about indirect effect? How would you compare which relationship is "stronger" If it is the one with mediator or the one without the mediator?


I have few more questions, can I get somehow in contact with you?

Thank you so much for you nice tutorial. Can you plz tell me, how do I understand that the moderation effect is significant????

Hi.
Just a question, I hope you could help me. I have developed an intervention to my students. This intervention functions as my independent variable. How could this be encoded in spss? Does that mean I input 1 for the student who undergone the intervention and 0 for those who are under control group?

Thank you.

Hi, I have a question. I am looking at whether age (categorical, e.g 1 = 20-25) effects scores on a scale (continuous). I then want to see if career stage (categorical, student = 1, qualified = 2) will act as a mediator for these results. Is my data appropriate to follow the steps in this video? Thanks

I have a question for my masterthesis. So i need to use an moderated mediation model 14 Process Hayes in SPSS.

X = colour (green, Blue, yellow)
Y = attitude
M = sustainability perception
W = Age (generation X, Y, Z)

So X and W are multicategorical. But it is really not clear for me how to understand these results.

Do you have a tutorial video on this? Or could help me understand the results.

Thank you in advance.
You would really save me and my thesis.

Hello, thank you for your support. If its posiible I have a question; Can I analyse categorical dependent variable in Model4 with Process?

Dear Dr. Buchanan,
Thank you so much for this video!
I only have one question left: Is it possible to include categorical covariates (2 and more categories) using process in SPSS?
Thank you in advance,
Sonja

Thanks so much! Is this the same process outline in Hayes & Preacher (2014; full cite below)? Just curious for citing this procedure in a paper.

Hayes, A. F., & Preacher, K. J. (2014). Statistical mediation analysis with a multicategorical independent variable. British journal of mathematical and statistical psychology, 67(3), 451-470.

Hi, thank you so much for your helpful videos! I have two quick questions if you have a few minutes to spare. I'm using a categorical IV (attachment categories, grouped as 1,2,3,4) and I'm doing model 7 with continuous M, W, and Y. My values for X1, X2 and X3 do not equal (and are not even close) to the value I get from subtracting the means. For example, for X1 for path 'a' I get a completely different score than the difference between means of 2-1 with M as an outcome. Do you know why that might happen? Also, is mean centering of other variables not necessary when the predictor is categorical? Thanks again!

Getting MEans without Split file -> go Means mpg by cyl.

Hi Dr. Buchanan,
Thank you so much for this video. Not only had I understood how to interpret a mediation with a categorical X, but I also learned about an easier way to draw the relationship diagram.
I have an additional question though, what if you have a categorical X (3 levels as in your video), a categorical Y (3 levels, or dummy-coded as 1=Presence vs 0=Absence for each single level), and a continuous M? I tried to put them into the same Model=4 and ran this analysis. How would the interpretation change?
In my case, I have X = culture (A,B,C), Y = style of advice (direct, multiple, indirect), M = responsibility for bad advice.
Thank you!

Thank you so much for the video. That was so helpful. Do we need to check the assumption of linearity?

Dear Dr Erin Buchanan,
I have a question regarding the interpretation: you always say there is a difference mediation 4vs6 and 4vs8. Is there a way to also compare 4vs6? 
Thank you in advance.

What should I do when we have two categorical mediators?

Thanks for this clear tutorial. Could I ask you if PROCESS handles a binary independent variable, a mediator which is continuous and a binary outcome? If so, what is the set up for this analysis? Thanks in advance

This is the MOST HELPFUL video. Thank you SO MUCH! Please don't ever delete this!

Thanks for the video Dr.
I have a question please. In your example, what if in the second run (comparing 6 Vs 4, and 6 Vs 8), it shows that the bootstrap CI includes zero for 6 vs 8. Do you conclude that there is no mediation effect even though the first run (4 Vs 6, 4 Vs 8) shows that there is? Or do you focus on just the first run (4 Vs 6 and 4 Vs 8) and conclude that there is mediation? I really need your help. Thank you!

this is great, is there a video on how to conduct and report a mediation analysis with a dichotomous outcome

Very helpful and clear, perfectly what I was looking for. Thank you ❤️

Very helpful.. Thank you a lot :)What if I have 2 categorical variables X? is it the same way??

Question: Why is there only one B path coefficient in the results? there are separate A and C' paths for the different group comparisons (X1, X2), but just one B path in the output?

Thank you! This was very helpful! Do you by any chance know what could cause contradicting findings of the index for moderated mediation which is significant for one category but the conditional indirect effects do not show me significance for this category but for another one? To report, can I rather rely on values of the index or the conditional indirect effects? Thank you!

Thank you very much. I would like to ask whether PROCESS can conduct a mediation analysis with Categorical IV (four categories) and DV (three categories), but with a continuous mediator. Thanks in advance.

I've not used mediation analysis but I have to now. This was great... super thorough & helpful.

This was soooo helpful Thank you so much!!!!! You saved so much time for me.. Thanks A LOT!

Thank you for this very useful video.
I wanted to ask you how to compare 6 to 8?
How do you know which group has more effect?
Thanks in advance

thanks so muh for this video, I noticed that this analysis selected the bootstrap samples as 5000, does it mean this mediation analysis used the bootstrap method? thanks

Dear Dr. Buchanan,
Thank you very much for this detailed video! I was wondering if PROCESS can analyze a mediation model where independent variable is a binary variable, mediator is continuous, and dependent variable is continuous/categorical variable? If so, what is the setup for this analysis? Thanks in advance!

Thanks so much for this video Dr Buchanan!!! saved my thesis <3 I just wanted to ask why we don't run correlations between variables before the mediation analysis? I understand you can't with the categorical data but would you not still want to check M and Y are correlated?

Hi, sorry if this question has already been asked, but is it possible to use SPSS process to analyse a categorical mediator (M)? I have a continuous X and Y, but categorical M. Thank you.

Thank you so much! I couldn't find about how to do a mediation analysis with a multicategorical x variable anywhere! This video is a life saver <3 Quick question: I'm currently writing my thesis and I'm wondering how you'd write the results down in a clear paragraph in APA style? Any clarification would help! Thank you!

Thanks

Thank you so much for your videos! :) I watched so many already and that helped me to understand my output for my masters thesis more. However, I still have trouble understanding and interpreting some points... I use model 14 of process and have a multicategorical X (low, high, unknown), two continuous Mediators, two continuous Y (calculated separately) and one continuous W. Also I included categorical and continuous covariates. I quite dont unterstand how to interpret the constant in my output of the DV, as e.g., x1 and x2 are sign. but the constant is not. Calculating that for the second DV the constant is sign. but x1 and x2 are not. Do you have any tips or other videos which might help? Thank you so much.

Can I ask why path b is the same for 2 different mediation relationships?

this was amazing.

This is THE most amazing SPSS video ever. Thank you so, so much!!

Hello, thank you for this video. What happens when you have only 2 categories. I notice you had 3… I’m trying to run this test for my X that is one group that has received professional coaching and one that has not, how do I do this test with only two categories it won’t let me run it when I click multicategorical and indicator as it says I need a minimum of 3 categories … what would my third category be?

Could you do one with a binary DV, please? The existing videos on this are not understandable.

Hi Dr Buchanan,

If my independent variable only has two categories, is it still able to perform models 4 and 7, or do I have to use another method?
Thank you so much for answering the question.

I cannot describe how this has helped my Grad Stats anxiety. I got this! I actually got this!

Thank you so much for your video! It has helped me a lot with my research.
Just wondering what type of Hypothesis would this test help to address? Would it be something along the line of: "Is there a significance difference in weights between 4 cylinders and 6 cylinders?"

Can anyone please help me to find out any video on using scale variable as the moderator

If I want to compare 6vs8 will I need to run a separate mediation analysis i.e., a sequential mediation?