Thanks for the great video! I actually did not know that you can create a new spss data file by copy-pasting the values on syntax, that was also a life saver!
Dr Mike, thankyou for such a detailed explanation. I just needed some help. If my interaction is significant and positive but the INDEX f Moderated mediation is negative, how would I interpret it? If I plot it, the slope comes out to be positive, but index of moderated mediation part I'm still confused, how to interpret that. If you would be kind enough to help
Hello, Do you have an example of a moderated mediation using the process model 18 ? I searching how to construct the graphic representation of the indirects and directs effects and how to calculate the coefficients. Thank you.
Thanks for the amazing presentation Mike ! just want to ask, what if we conduct the analysis with all latent variables (e.g. V1 V2 V3, T1 T2 T3 indicators of them ) and put these into analysis instead of means, do Process still works well ?
Could you specify, what is the final verdict in this case? As the conditional and conditional indirect effects are significant at their respective values yet the index of moderated mediation nonsignificant?
Hello, when i conduct a moderated mediation using Hayes Process macro model 16 - can i just click on "model number": 16 and put a second moderating variable in the field "Moderator variable W/Z?" Or does the conduction process of model 16 differ a lot from the process of model 14? I really would appreciate an answer! Thank you in advance.
Hi, thank you for the great tutorial! It was really useful and easy to follow. I was wondering, is there a way to verify the pairwise contrasts of indirect effect when you have a multicategorical x (they say that this option is not available)? Thank you!
Thanks for the teaching video, Professor Crowson.. I am just wondering if there were typos on ,Model 2 regarding XW (interaction term) and slope for XW** on 2:16? Shouldn't they be M instead of X? Thanks..
Thanks! super useful. In my data, instead of having one score for each subjects, I actually have 60 trials for each subjects, so I would like the model to take into account that points come form different subjects, is there a way for that? thanks!
Hi Valentina, thanks for visiting and for your question. Process really isn't set up for repeated measures data of the type you are describing. It sounds like you have panel data with repeated measurements. If you only had a few time points I might suggest a cross-lagged panel design using SEM and then test indirect effects that way. [David McKinnon (2008) does a nice job of talking about ways of testing longitudinal models out using cross-lagged designs.] It might also be possible to test mediation in Process using syntax.] But with 60 repeated measurements per subject, I'm drawing a blank on alternatives for you. I wish I could be more helpful on this for you. Cheers! *MacKinnon, D. P. (2008). Introduction to statistical mediation analysis. Taylor & Francis Group/Lawrence Erlbaum Associates.
@@mikecrowson2462 thanks Mike! I may can generate a model for each subject and then test whether estimates are different from 0. This is a technique I use to obtain a beta for each participant in regression analysis...
If the IMM is not significant there's really no need to interpret each conditional indirect effect. The video is just walking through different pieces of output. If IMM is not significant you might rerun the analysis without the conditioning of the indirect effect and proceed to interpret that single indirect effect...being sure to report on the original model and respecified model.
thank you very much for your reply. If the index of moderated mediation is significant, i should skip the single indirect effects and interpret only the conditional ones?
Thanks for the great video! I actually did not know that you can create a new spss data file by copy-pasting the values on syntax, that was also a life saver!
Dr Mike, thankyou for such a detailed explanation. I just needed some help. If my interaction is significant and positive but the INDEX f Moderated mediation is negative, how would I interpret it? If I plot it, the slope comes out to be positive, but index of moderated mediation part I'm still confused, how to interpret that. If you would be kind enough to help
Hello, Do you have an example of a moderated mediation using the process model 18 ? I searching how to construct the graphic representation of the indirects and directs effects and how to calculate the coefficients.
Thank you.
Thanks for the amazing presentation Mike ! just want to ask, what if we conduct the analysis with all latent variables (e.g. V1 V2 V3, T1 T2 T3 indicators of them ) and put these into analysis instead of means, do Process still works well ?
Could you specify, what is the final verdict in this case? As the conditional and conditional indirect effects are significant at their respective values yet the index of moderated mediation nonsignificant?
Hello,
when i conduct a moderated mediation using Hayes Process macro model 16 - can i just click on "model number": 16 and put a second moderating variable in the field "Moderator variable W/Z?" Or does the conduction process of model 16 differ a lot from the process of model 14?
I really would appreciate an answer!
Thank you in advance.
Hi, thank you for the great tutorial! It was really useful and easy to follow. I was wondering, is there a way to verify the pairwise contrasts of indirect effect when you have a multicategorical x (they say that this option is not available)? Thank you!
If I do as the same yours, do I need to use AMOS further? Thanks very much!
Thanks for the teaching video, Professor Crowson.. I am just wondering if there were typos on ,Model 2 regarding XW (interaction term) and slope for XW** on 2:16? Shouldn't they be M instead of X? Thanks..
I'm wondering the same thing!
if i dont have the data kist free in the output, what can i do?
Thanks! super useful. In my data, instead of having one score for each subjects, I actually have 60 trials for each subjects, so I would like the model to take into account that points come form different subjects, is there a way for that? thanks!
Hi Valentina, thanks for visiting and for your question.
Process really isn't set up for repeated measures data of the type you are describing. It sounds like you have panel data with repeated measurements. If you only had a few time points I might suggest a cross-lagged panel design using SEM and then test indirect effects that way. [David McKinnon (2008) does a nice job of talking about ways of testing longitudinal models out using cross-lagged designs.] It might also be possible to test mediation in Process using syntax.] But with 60 repeated measurements per subject, I'm drawing a blank on alternatives for you. I wish I could be more helpful on this for you. Cheers!
*MacKinnon, D. P. (2008). Introduction to statistical mediation analysis. Taylor & Francis Group/Lawrence Erlbaum Associates.
@@mikecrowson2462 thanks Mike! I may can generate a model for each subject and then test whether estimates are different from 0. This is a technique I use to obtain a beta for each participant in regression analysis...
If the index of moderated mediation is not significant, why do we have to interpret the conditional effects at different levels of the moderator?
If the IMM is not significant there's really no need to interpret each conditional indirect effect. The video is just walking through different pieces of output. If IMM is not significant you might rerun the analysis without the conditioning of the indirect effect and proceed to interpret that single indirect effect...being sure to report on the original model and respecified model.
thank you very much for your reply. If the index of moderated mediation is significant, i should skip the single indirect effects and interpret only the conditional ones?