Thanks for the helpful video! Are both of your mediators considered "partial" mediators because your direct effect is still significant? For example, would a "full mediation" be when the direct effect is NOT significant and the indirect effect IS significant, while a partial would be when the indirect and direct effects are both significant? Or am I interpreting it wrong?
Partial and Full are the old way of thinking about it, but you are correct. For more info on the new way of thinking about mediation, check out Zhao, X., Lynch, J. G., & Chen, Q. (2010). Reconsidering Baron and Kenny: Myths and truths about mediation analysis. Journal of consumer research, 37(2), 197-206.
Hello James, thank you for your video. Helps out a lot. Two questions: (1) Should I include estimator = ML in addition to bootstrap = 5000 under the analysis command or should I only include the bootstrap options; (2) Do you know what the difference is between BOOTSTRAP versus BCBOOTSTRAP for Cintervals? From my understanding, it is better to use bias-correct confidence intervals. Any guidance would be helpful. Thanks!
Yes, bias corrected is better. It reduces artificial inflation of estimates. I think ML is the default estimator, but it won't hurt to include it. Yes you should include the bootstrap = 5000 to let it know you want 5000 bootstrap samples.
Hi James, very helpful video. Does this model assume that the mediators are independent? What if the mediators are correlated - does this need to specified in the syntax? Thanks.
Thank James for such a great video. I would like to know how to write Syntax if we have only exogeneous variable (e.g. the effect of feedback on job satisfaction through empowerment?
If you have only one IV, one Med, and one DV, then it would be just two lines: one with DV on the Med and IV, and then one with the Med on the IV. Then if you want to estimate mediation, make sure to include the INDIRECT line.
Hi James, thank you for the video!! I have a short question: if the ''Total indirect effect'' is not significant, but the ''Specific indirect effect'' of both of my mediators are significant, can I say that the IV still has an indirect effect on the DV? Thank you!!
Thank you, Professor with video. I have a question. The standard error, t value and p value do not appear in my standardized model results. Only non-standardized values appear. What can I do in this situation?
Hi, James, thank you for the helpful video! I have a question when analyzing indirect effects. My input file is like as follow. variable:names= EDU CA1 CA2 CA3 CA4 CA5 SA1 SA2 SA3 SA4 SA5 SA6 DK IQ EI1 EI2 EI3 EI4 AH1 AH2 AH3; analysis:type = general; bootstrap = 5000; model: CA by CA1 CA2 CA3 CA4 CA5; SA by SA1 SA2 SA3 SA4 SA5 SA6; DK; IQ; EI by EI1 EI2 EI3 EI4; AH by AH1 AH2 AH3; IQ on SA EI AH DK; CA on SA EI AH DK IQ edu; model indirect: CA ind IQ SA; CA ind IQ EI; CA ind IQ AH; CA ind IQ DK; output:sampstat stdyx cinterval(bootstrap); After analyzing, there is a warning. *** WARNING in OUTPUT command BOOTSTRAP and BCBOOTSTRAP confidence intervals are only available for the regular results of the model and MODEL INDIRECT. No confidence intervals are printed for the standardized results of the model and MODEL INDIRECT. 1 WARNING(S) FOUND IN THE INPUT INSTRUCTIONS Do you know what the input command wrong? Thank you!
If I'm interpreting your model correctly, IQ is the mediator. If this is the case, then remove IQ from the indirect statements. For example, change "CA ind IQ SA" to "CA ind SA"
![[DOKKAN BATTLE] Worldwide Campaign Announcement Video Part 2!](http://i.ytimg.com/vi/JpT5Voak6WA/mqdefault.jpg)
Thanks for the helpful video! Are both of your mediators considered "partial" mediators because your direct effect is still significant? For example, would a "full mediation" be when the direct effect is NOT significant and the indirect effect IS significant, while a partial would be when the indirect and direct effects are both significant? Or am I interpreting it wrong?
Partial and Full are the old way of thinking about it, but you are correct. For more info on the new way of thinking about mediation, check out Zhao, X., Lynch, J. G., & Chen, Q. (2010). Reconsidering Baron and Kenny: Myths and truths about mediation analysis. Journal of consumer research, 37(2), 197-206.
Thanks so much! I'll definitely take a look!@@Gaskination
Hello James, thank you for your video. Helps out a lot. Two questions: (1) Should I include estimator = ML in addition to bootstrap = 5000 under the analysis command or should I only include the bootstrap options; (2) Do you know what the difference is between BOOTSTRAP versus BCBOOTSTRAP for Cintervals? From my understanding, it is better to use bias-correct confidence intervals. Any guidance would be helpful. Thanks!
Yes, bias corrected is better. It reduces artificial inflation of estimates. I think ML is the default estimator, but it won't hurt to include it. Yes you should include the bootstrap = 5000 to let it know you want 5000 bootstrap samples.
Hi James, very helpful video. Does this model assume that the mediators are independent? What if the mediators are correlated - does this need to specified in the syntax? Thanks.
Usually Mplus automatically correlates mediators' error terms. I'm not sure why it didn't in this model.
Thank James for such a great video. I would like to know how to write Syntax if we have only exogeneous variable (e.g. the effect of feedback on job satisfaction through empowerment?
If you have only one IV, one Med, and one DV, then it would be just two lines: one with DV on the Med and IV, and then one with the Med on the IV. Then if you want to estimate mediation, make sure to include the INDIRECT line.
hi james, thank you. Do you have a model for this but with a moderator and for a within subject design?
Here is my full playlist for Mplus: ruclips.net/p/PLnMJlbz3sefL983xxEgbGjRc6mVIsIcAg&si=Y29re6hioKkDNK2n
Hi James, thank you for the video!! I have a short question: if the ''Total indirect effect'' is not significant, but the ''Specific indirect effect'' of both of my mediators are significant, can I say that the IV still has an indirect effect on the DV? Thank you!!
Yes.
@@Gaskination thank you!!!
Thank you, Professor with video. I have a question. The standard error, t value and p value do not appear in my standardized model results. Only non-standardized values appear. What can I do in this situation?
Check the menu at the top to check if standardized values are showing. There is an option for it in the menu of the output window.
@@Gaskination I'm doing bootstrap 2000. And it still does not come out. Is it related to this?
@@fatihcelik7537 If you're including the STDYX in the OUTPUT line, then it should provide that option. Make sure you have not entered it as STDXY.
Hi, James, thank you for the helpful video! I have a question when analyzing indirect effects.
My input file is like as follow.
variable:names= EDU CA1 CA2 CA3 CA4 CA5 SA1 SA2 SA3 SA4 SA5 SA6
DK IQ EI1 EI2 EI3 EI4 AH1 AH2 AH3;
analysis:type = general; bootstrap = 5000;
model: CA by CA1 CA2 CA3 CA4 CA5;
SA by SA1 SA2 SA3 SA4 SA5 SA6;
DK;
IQ;
EI by EI1 EI2 EI3 EI4;
AH by AH1 AH2 AH3;
IQ on SA EI AH DK;
CA on SA EI AH DK IQ edu;
model indirect: CA ind IQ SA;
CA ind IQ EI;
CA ind IQ AH;
CA ind IQ DK;
output:sampstat stdyx cinterval(bootstrap);
After analyzing, there is a warning.
*** WARNING in OUTPUT command
BOOTSTRAP and BCBOOTSTRAP confidence intervals are only available for the regular
results of the model and MODEL INDIRECT. No confidence intervals are printed for
the standardized results of the model and MODEL INDIRECT.
1 WARNING(S) FOUND IN THE INPUT INSTRUCTIONS
Do you know what the input command wrong?
Thank you!
If I'm interpreting your model correctly, IQ is the mediator. If this is the case, then remove IQ from the indirect statements. For example, change "CA ind IQ SA" to "CA ind SA"
@@Gaskination Very very helpful! thank you!!