Testing mediation and moderation using Hayes Process Macro (Model 5) with SPSS

[@yasinali8827]

Thank you Sir. You are the GOT of statistics.

[@YuriPavlov]

It little time, this instructor became my favorite. Thank you!

[@mikecrowson2462]

Thank you, Yuri, for your kind comment. Best wishes!

[@The_Quota_Official]

Thanks so much Mike, your videos are always helpful and clear.

[@mikecrowson2462]

Thank you for your comment! I'm glad you are finding my videos helpful. Stay tuned for more upcoming videos on Process :)

[@The_Quota_Official]

@@mikecrowson2462 it would be awesome to see you do videos on Matlab, however I’m not sure if the program falls into your area of expertise :)

[@mikecrowson2462]

@@The_Quota_Official unfortunately Matlab is not currently in my wheelhouse. But I'll keep it on my radar for something to learn in the future! Thanks for your suggestion. Cheers!

[@zewdietura3515]

Thank you for the video, I'm conducting a research with three moderators and mediator(as a set of subscales) that may have impact on relationship between X and Y. So, how could run the analysis process? can I have anything from you to support my stand?

[@mikecrowson2462]

Hello. Thanks for visiting. Your 'stand' regarding your model is a theoretical issue. The question is, "Is it defensible based on the literature and sound reasoning". If you are asking if you can test this particular conceptualization using Process, the answer is no. The reason is that Process only allows you to include two moderators (hence the W and Z boxes in the main console) in the model. If you are seeking to include a third moderator, then you would need to try something else. There aren't too many 'ready made' programs out there that will make things simple for you. You might try downloading a copy of the open software jamovi (at www.jamovi.org/download.html). This is built off R program language, but the interface is really easy to use. I believe if you download the Module, General analyses for linear models, you can get an interface for performing mediation, moderation, and also testing moderated mediation. I don't really do anything with this, but it's something you might be able to do with your model. Best wishes.

[@fatmayagmurevcil2526]

very instructive, thanks

[@dsavkay]

Amazing, thanks 💯

[@서승경-l1e]

[Seo Seung KyeongBell] [오후 12:42] Thank you for your explanation of model number 5. I wrote my thesis as model number 5 in Korea. However, there are not many studies, so could you introduce me if you have a paper that you used as Moden number 5? And model number 5 is the model made from the number of cases? Or is it a model for utilization?

[@RybatQ]

Hi. I am also working on model number 5. Can you share your thesis with me, I'd like to read them. Unless of course they're not in English.

[@서승경-l1e]

[Seo Seung KyeongBell] [오후 1:27] Can you present a paper using Model 5?

[@MrArdahazal]

Hi, I need an urgent help I have a question: 1. Since indirect effects are not normally distributed, significance tests are evaluated according to bootstrap confidence intervals. When deviating from normal, a larger sample may be required. 2. I know that Hayes' script is not longitudinal for cross-sectional data. Thank you very much in advance

[@azadehart]

Thanks for the video. I wonder how we can analyse (ex) company size as a demographic variable? Should I use latent result? Thanks

[@DrJeZeus]

Thanks for this video Mike, finally someone has addressed model 5! One question: if partial mediation occurs in model 4, but is then not significant when the moderation is present (i.e. model 5) and significant , how would you explain this, for example if it occurred with this data?

[@mahiratabassum2863]

I want to test this model on Amos. Please share 1 vedio How to test model 5 of Hayse in Amos?

[@dhanyapravin]

Hi Could you please tell what is the purpose of bootstrapping in PROCESS and how can we interpret in report about why we have chosen the bootstrap sample as 1000 or 5000.

[@seonahlee7091]

This video is a very informative and friendly explanation. I have one question. As a result of applying Model 5, the indirect effect of X on Y was not significant. In this case, I wonder if the thesis has no value.

[@mikecrowson2462]

Hello Seonah, I'm glad you found the video useful. I'm not really sure I understand with your question about the 'thesis having no value'. If you are asking if it is possible for a previously significant indirect effect might could become non-significant after adding in the moderated effect of path c', I would assume this is possible. In that second regression model where Y is regressed onto M, X, the moderator (W), and the X*W interaction, the effect of the mediator on Y is controlling for the other predictors in the model (i.e., X,W, and XW); whereas if the moderator was not previously included (using Model 4 for the standard mediation model), then the only variable being controlled for is X. I cannot speak to the issue of 'why' this might have occurred from a theoretical standpoint. I only highlight that it is possible. Cheers!

[@AZADEHMOOSAVIUPM]

hello, why macro heys is not working when put empty variable m , as my model is moderator not mediator?

[@janinedebart2226]

Does anyone know how to find the standardized beta coefficient in the outcome when using process model 5?

[@jalan2dibelanda]

Very clear video Sir! I have a question about mean centering (minute 5.38 of your video). What if my X is a categorical variable? (M and W = numerical, Y= also numerical). I tried to mean centered all variables (Process mean centered my X and W). I also tried only continuous variable that defines products (Process only mean centered my W). Which option is actually fitting for this case Sir? Thank you and I wish you a great evening:). Regards, IB

[@mikorees5853]

How to interpret if there is mediation in this model or not?

[@dezariemoskal9724]

Thank you for the video! I have the same question as DrJeZeus (in the comments below), but I am unsure whether you typically address these types of questions. I am thinking that the moderator (not accounted for in model 4) may be contributing to the mediation effect (confounding 3rd variable) which when accounted for in model 5 pulls the variance from the mediation effect. Thanks in advance for any assistance you can offer.

[@darkcrow4671]

Sir your voice sound like K Billy's announcement in Reservoir Dog

[@dr.hariharasudharevathi8528]

Thank you so much sir. Kindly explain model 80 as well sir. Looking for that as it will definitely help me for my thesis.

[@mikecrowson2462]

Hi there. I don't ordinarily drop everything to take a request, but this seemed like a good video to put together. I hope you find it helpful! Cheers. Go to: kzbin.info/www/bejne/qpXOln2ZhsuFrs0

[@dr.hariharasudharevathi8528]

@@mikecrowson2462 Thanks a tonne sir. Extremely helpful.