5. Hayes Process Macro - Model 4 |
ฝัง
- เผยแพร่เมื่อ 17 ก.ย. 2024
- The tutorial guides on Model 4 of the Hayes Process Macro for Mediation Analysis with Single Mediator.
Concept of Mediation
• What is Mediation Anal...
For Multiple Mediation using Process Macro, watch
• 6. Hayes Process Macro...
For lecture notes and much more, visit
researchwithfa...
#SPSS #ProcessMacro #HayesMacro #MediaitonAnalysis #ConditionalAnalysis #DataAnalysis #MediatingVariable #ResearchWithFawad #SocialSciencesResearch
Very extremely helpful video. Browsed through 10 videos to understand mediation and moderation analysis and understood it from here only.
Pleasure. I am glad you liked it. Please do share in your research circle.
As usual, clear, precise, and detailed. This is very helpful. Thank you Dr. Fawad!
Pleasure. I am glad you liked it.
Thank you so much. I've been struggling to get my thesis analysis done but this video has helped me a lot. Thank you
Glad it helped!
Your style of explanation was very clear. Thank you very much
Pleasure. I am glad you liked it.
You are excellent at what you do Sir. Thank you.
Thanks. I am glad you liked it.
Amazing video! Helped me so much for my SPSS data analysis, thank you so much!
You're very welcome!
I really appreciate this, particularly the last part, I understand the results but wording them was what I was sturggling with - massive help thank you
Thanks. I am glad you liked it. Please do share in your research circle.
@@researchwithfawad Can you please tell the reference of the mediation table (the pattern that you are followng is of APA or what). Thank you
Your video helped me to interpret the table. Thank you so much!
Glad it helped!
This was so detailed and excellently explained. Thanks so much
You're very welcome!
Thank you for the video, was very easy to follow and explained everything clearly.
Pleasure. I am glad you liked them.
This is very clear, and helpful!
Glad you think so!
Sir ur vedio is very informative, doing PhD analysis by watching ur vedio 🙏
Thanks. I am glad you liked it
Sir, thank u so much for such a wonderful explanation. how to calculate the t staistics which u have mentioned at the end of the tabel sir.
Thanks for watching. You can divide effect size by Standard Error to get t value.
@@researchwithfawad thank you for this response. I was about to ask same question
insightful work. thank you so much
Glad you enjoyed it!
Thank you for this wonderful tutorial👍🏼
Glad it was helpful!
@@researchwithfawad Can you please make another one on Process Model 21? I could not find it in your playlist. Your explanations are the best
Thanks. Hopefully soon.
Respect and gratitude. Thank you sir.
You are very welcome
Thank you very much for your clear explanation, it helped me a lot. I only have one thing I can't figure out. At the end where you show all the results in the table, you show the t-statistics for the indirect effect. How/where do you find that number (5.583)?
Pleasure. I am glad you liked it.
Yo get t statistics, kindly divide the Path Coefficient by Boot SE
Really fantastic video. Thank you!
Glad you liked it!
Hi, thank you so much for making this video. It's really helping me through my thesis. One question, In the final write up, the t value for indirect effect is listed as 5.583 but I cannot find that value in your shown outputs. Maybe I missed something. Could you please advice where to find the t value for the indirect effect? Thank you so so much!!
Pleasure. I am glad you liked it. To get t value, divide Effect by Boot SE.
Should mention this in description @@researchwithfawad
@@researchwithfawadGod bless you for this response. Was about to ask the same question.....
😊😊
Twas truly a remarkable video for learning Process Macros. You simply demystified the once dreaded tool. Thank you soooo much. I ❤❤❤ this video
it really helped a lot, thank you so much sir
Thanks for watching
Excellent...
Thanks for watching.
Dear Mr.Fawad, thank you so much for the tutorial. It helped me a lot! I had a question, how to adding control variables in mediation analysis? Thank you Mr.Fawad
Pleasure. I am glad you liked it. You can use covariates in Hayes Macro. If there are more than two categories, create dummy variables and use them as separate variable. However, keep 1 category as reference.
@@researchwithfawad Thank you for your prompt response, Sir. I would like to ask another question.
Is it possible to add non-linear variables (for example: GDP squared) in Model 4 Hayes Macro? If yes, how can I proceed? Thank you!
Brilliant teaching...making thing easy
Glad you liked it
I appreciate this video, to deepen my understandings. In my findings, direct effect is not significant, but indirect effect is significant, also total effect :P=0.523. Can I interpret that this mediator is a perfect mediator? Thank you.
Thanks. I am glad you liked the video. Yes, there is complete mediation.
Have binge watched many videos in the past few days to understand mediation through Hayes Process and this by far is the *most* clear one👍 thank you so much, going to subscribe and share this for sure. One question: is the write up at the end in APA style? I have been looking for tables in that format but can't seem to find any
Thanks. I am glad you liked it. Please do circulate in your research circle.
Tables in this video may not be in complete APA Style. However, you can watch this video for APA guidelines
th-cam.com/video/mLKVPXwqT9o/w-d-xo.html
@@researchwithfawad Ofcourse! Thank you for the link. Will take a look.
Excellent explaination👍
Glad it was helpful!
excellent work sir!
Thanks. I am glad you liked it.
very well explained. Please also explain the difference between Beren and kenny moderator/mediator and Process macro
Thanks. I am glad you liked it. Please watch
th-cam.com/video/Zu-P_KE0x4c/w-d-xo.html
Thank you very much for your presentation, it is very helpful and informative. Please, I have 2 questions:
-Is the significant outcome of the path (a) and (b) a condition of mediation (as is the case with the conditions of Baron and Kenny (1986)? If not, what are the conditions of mediation according to Hayes' method?
-According to the results of my thesis: paths a and b, as well as indirect effect, are significant. But Total effect (c) and direct effect (c') are not significant. Are we talking about mediation here? If so, is it total or partial?
Thanks. I m glad you liked it. In your case it is full mediation as the indirect effect is significant but the direct effect is insignificant
Thank you. Very helpful.
Thanks. I am glad you liked it.
Thank you so much for this video
Pleasure. I am glad you liked it.
Sir, where did you get the final value of t-statistics when reporting?
Divide the effect size by standard error.
Sir, could you please explain further where is comes from the t-statistic value 5.583 at the mediation analysis summary in Table 1. Thank you
Divide the indirect effect (0.2736) by the standard error (0.0490) and this will give you the t-statistics.
@@researchwithfawad tq sir for the clarification
thank you so much sir for this amazing video, I have a question about my findings SIR in my results " the Total effect and Direct effect is significant while Indirect effect contains NEGATIVE sign, so is there is mediation exist or not or we can say that its a competitative mediation" thanks again sir
Yes, it is competitive mediation. If indirect effect is significant, mediation is significant
@@researchwithfawad sir my indirect effect value is=( -. 0635),
bootSE=(. 0404)
BootLICI=(-.1466)
BootULCI=(.0099)
sorry to bother u again sir, so is it my indirect value is significant as u can see the all values above. Please correct me if m wrong sir
Thank you for the explanation!
I have a question. Path a, b, and c are significant, while c' is insignificant.
From my understanding, path c here shows that IV and DV has a direct relationship.
Since c' is insignificant, is there full mediation?
Pleasure. I am glad you liked it.
What about the indirect effect?
If Indirect effect is significant and C' is insignificant. This is full mediation.
Thank you for this! I have one question, how will I know the % that accounts for the mediator?
Pleasure. I am glad you liked it.
Please check the section where Mediator is the outcome and the R sq in that section.
Dear Mr.Fawad, I love your video.
It is very easy to understand and has helped me tremendously in my thesis reporting process. I did however struggle to find the t-statistics number that you put in the table 1 in minutes 15:27.
Would you be so kind to help me by pointing out where I could find the t-statistics number to be put in table 1? Thank you 😁
Thanks. I am glad you liked it. T statistic is calculated by dividing the Path Coefficient (Estimate) by the Standard Error (SE). Estimate/SE will give you t-statistic.
@@researchwithfawad Thank you!
@@researchwithfawad Hi, thank you for the video. however I still dont quiet get it, what exact estimator and Standart Error you use for the calculation of T statistics? can you point out it location?
thank you.
Dear Dr, How did you get the value of 5.583 for t statistics that you have shown in your Mediation Summary Table? I hope to get an answer for this soon. Thank you.
Thanks for watching. If you divide the Beta value by BootSE, you will get the t statistics.
Hope this helps
@@researchwithfawad Thank you for your answer sir. Really appreciate it.
Hello sir Thank you for explaining so easily. I have a question kindly answer this. My total effect is sig and positive. My direct and indirect effect is negative and also direct effect value is LLCI-.0127 and ULCI.3075
Is this relation is significant or not? Also if it is not significant can i still run a moderation analysis using model 7?
If there is a 0 between UL and LL CI, this shows insignificant results. Yes, you can run and check the results of Model 7.
Thanks for this video. I had one question left. If we want to perform a regression analysis, should we perform this with the use of PROCESS when we also want to check for mediation. Or is it possible to check the effect of X on Y with just SPSS first and then the mediation effect with PROCESS? Because PROCESS gives slightly different values for my unstandardized coefficient than SPSS does.
Use Process for direct relationship as well when testing mediation. Difference in results is because Process macro utilizes bootstrapping technique.
Great Video! Thanks a lot. I have a quick question, however. If I want to test for a mediation effect and I have multiple IVs in my conceptual model. Do I run separate regressions for each IV or one multiple regression that includes all IVs and my mediator variable?
One Regression. If you are using Process Macro for mediation, follow the steps in the video.
@@researchwithfawad Assalamua'alaikum Dr. Can you explain more about this because I saw in the video that we can only put one IV under X variable. And can I still use Process Macro if my DV is dichotomous? If not, what other options do I have? Thank you!
This is an excellent video, well explained and robust. Please, how did you get the result for the t-statistics (5.583)?
Thanks. I am glad you liked it. To get t statistics, Divide the Boot Estimate by Standard Error
Thank you for your great teaching in the video. I have a question. What if my a path is not significant, but the b path and c' path are significant. Is it acceptable to say there is a mediation? Appreciate if you could answer this
Pleasure. I am glad you liked it. For mediation check for specific indirect effect.
@@researchwithfawad I have the same problem. Path a not significant but b and c' are significant. What do you mean by checking for specific indirect effect? Thank you so much for such helpful video sir.
Thank you Dr Fawad, this was helpful. But in my model I have four IVs with one M and one DV. Please suggest where to put three additional IVs in MACRO? I have used SEM but want to test the results with Process Macro as well?
Thanks for watching. This is a limitation of Hayes Macro. Please use SEM tools like Amos or SmartPls
@ResearchWithFawad Yes that is correct but I was told to put all other IV as Covariates. Is that correct?
Sir. Thank you so much for simplifying the method. Sir, I have one more doubt how to calculate EFFECT RATIO. Plz guide me sir.
You need to calculate VAF values. I am sorry I don't have a video on VAF yet.
thankyou for a detailed video. Can you please explain how will we calculate the effect size in this mediation analysis?
Pleasure. I am glad you like it. The results do provide an indirect effect: a*b
Your videos is really helpful for my research...sir i have one question, can i used mutiple regression to test the relationship and process macro to test the mediation.??
Thanks. Yes, but Process macro also provides details of individual relationships.
Really helpful! I do still have a question. Can I also use PROCESS model 4 to test partial mediation? Or should I also perform linear regression to test the direct effect?
Thanks. I am glad you liked it. Yes, partial or full is contingent to the results. It does provide the direct effects.
How did you get the value of t-statistics that is presented in the table? Is it the t-statistics of direct effect of X on Y? I am confused because in the mediation analysis table the value is t = 5.583 while in the matrix output the closest to it is t = 5.5426.
Thanks for watching. T statistics is calculated by dividing the effect by BootSE.
Thank you for the great video ! I am running the process for mediating testing of my dissertation. I have some problems with the output interpretation. My output did not display the indirect effect, but it reported the error # line..... Cold you please explain what does this mean? Thank you.
Please email me the screenshot on kh.fawad83@gmail.com
@@researchwithfawad thank you very much. I did email you.
Hello - thank you for your video, it has been really helpful. My findings show negative coefficient values for my total, direct and indirect effect all of which are significant (my b path is positive and also significant). What does a negative coefficients mean for an indirect effect? (ie. are both my IV and M increasing and my DV decreasing or is it something like my IV is increasing my M is decreasing?)
Interpret the relationships between IV and MV and further how MV influences DV.
Secondly, if path a * b * c' is positive, this is complimentary mediation otherwise competitive.
Please refer to Chapter 7 in
Hair et al (2021). Partial Least Squares Structural Equation Modeling (PLS-SEM) Using R: A Workbook
Thank you very much for the informative video! It was really helpful. If you don't mind me asking a short question: If my current research (mediation - model 4) uses binary dummy coding on the X-variable (i.e., 0 & 1), instead of a continuous predictor. Can I still run Process Macro the same way you did? If not, how would you suggest I do it? I am aware that for X with 3 categories, I would need to tick the multicategorical box before running the analysis, but mine only has *2 categories* as aforementioned. Would love to hear from you!
Thanks for your interest. Yes you can still use Model 4 for Binary IV.
Thank you so much for this insightful and informative tutorial! In addition, can you tell me know name and link of thesis that you used in video?
Thanks for watching.
Dear authors, thank you so much for this insightful and informative tutorial! In addition, may I know how can we compute the p-value and t-value for indirect effect path? Really appreciate if you could assist on this
Thanks. I am glad you liked it. To get t statistic, divide Estimate by Standard Error. To get p value, use p value calculator online.
AOA. Does Hayes Process take in account the sign of the direct and indirect effect for computing total effect size? For instance, when the signs for both direct and indirect effect are different, does that explain *why* sometimes the total effect is LESS than the direct or indirect effect? Because logically it doesn't make sense that the total effect would be less than either of the two. If not, what other reasons could that be? Suppressor effect?
WAS. Please go through this thread.
www.researchgate.net/post/Is_it_statistically_possible_to_have_a_indirect_effect_value_more_than_total_effect
@@researchwithfawad Thank you, you are so helpful seriously
Dear Fawad, Thank you very much for your videos. I have a doubt. How did you get t= 5,583 for the indirect effect? I don´t see this value in the tables, and I don´t know where does it come from.
Thanks for watching. Divide the effect size by standard error.
@@researchwithfawad Thank you very much for your quick and clear answer!
Hello ! Thank you for this well explicated tutorial :) I have one question and I hope you can answer. My mediator variable is a score on pre and post test, so if I want to make this analysis, do I have to do it using the post-pre scores ?
You can run two model and assess the mediation in both conditions.
Thank you!
You're welcome!
Salaam Aleykum, thanks for this great video again. I would like to ask how did you decide on a t value of 1.96?
Thanks for watching. For Two Tailed Test at 95% CI, the threshold is 1.96.
Sir, please let me know from where you get the value of t=5.583??? (t-statistics) last part
Thanks for your interest. You can divide the Estimate by Standard Error (SE).
I have a question: where does the t-value of 5,583 come from in the report? I see a t-value of 5,543 with direct effect. Is this a typo? Or should this value be obtained from somewhere else?
Thanks for your interest. T value is estimated by dividing the Estimate by Standard Error (BootSE)
Respected Mr. Fawad small clarification from the interpretation part from where did taken the t value as 5.583 and when i have 2 IV and 1 MV and 1 DV how to performance a test????do you workout like that models ?
I sincerely appreciate and thankful for your response in advance
Hi. To calculate T, divide Estimate by BootSE. If you have multiple IVs, use AMOS.
Thank you for the helpful video. My question is that from where have you calculated the value of t statistics
Thanks. I am glad you liked it.
You can divide the Estimate (Est) by standard error (se) to get t-value
@@researchwithfawad could you please elaborate by the example from this video
Thank you so much sir, very informative lecture. I have a question, how to do mediation analysis when we have one categorical indipendent variable and more than one continuous dependent variable?
Pleasure. The process is the same. You just need to create dummy variables if there are more than 2 categories in the IV.
Thank you very much for your presentation, it is very helpful and informative. Please, I have a question: -According to the results of my thesis: paths a and c are significant and path b, as well as indirect effect, are insignificant and includes a zero. But Total effect (c) and direct effect (c') are significant. Are we talking about mediation here? If so, is it total or partial? m quite confused about results as it shows mediation or not? kindly guide me
Pleasure. I am glad you liked it. It is mediation analysis. Please refer to this video for further details
th-cam.com/video/8lDwmMF9aN8/w-d-xo.html
Thanks for responding but M asking about mediation analysis in process macro. I have mentioned results in above comment. Can u plz tell me does my results show mediation?
Watch the video shared, plus the following video. Your answer is in the videos
th-cam.com/video/Zu-P_KE0x4c/w-d-xo.html
please this is important, so my question is how did you get the t-value in the last final table?
it would be a huge help Sir. Thanks.
You can divide the path coefficient (effect size) by Standard Error.
Is the p value 0.0001 or is it p 0.05 ? I am a bit confused on this on one side you have written 0.001 but during your explanation you quoted 0.05. also one of my variable has a p value of 0.0021 does it mean its significant ?
P value less than 0.05 is significant.
Why is t in the table 5.583? Where does that value come from or how is it calculated?
You can divide the Estimate by the Standard Error to get t-statistics.
Where does the t-statistic (5.583) value come from in the latest report table, sir? I could not see the value 5.583 under the t values in the hayes program.
T statistic is calculated. Estimate/SE will give you t statistics
@@researchwithfawad Thanks a kot sir but where can i find the estimate ?
@@researchwithfawad SIR, I CAN FOUND THAT FINALLY, I looked at the comments and your old replies. THANK YOU SO MUCH FOR EVERYTHING! 🥹🥹🥹🥲🥲🥲😅😅😅😅😅😅😅🌼🌼🌼🌼🌼🌼🌼🌼🌼🌼🌼🌼🌼
Where do you find the value of the t statistic for the indirect effect?
If you divide the indirect effect by Standard Error you will get the t value.
when i'm putting the values, output says "set printback=off" and no results are shown, can you help?
Please reinstall the plug in.
I have question about objectives to put in the analisis. As first, I put all participants including target population(for exampe: score
I am sorry, I did not understand the question.
@@researchwithfawad If my interst is depression, I put all depressed people in the mediation analysis, is this correst?
Dear Mr.Fawad, thank you for the explanation it helped me a lot.
However I have one question, if path a was significant and path b was not significant we can say that path ab is not significant. Is that right?
It is safer you use the Bootstrap Upper & Lower bounds to establish if it is significant or not. So that rule of thumb doesn't lead you to a wrong decision & conclusion
No, better to check the significance of the indirect path.
What if the indirect effect is negative but significant? How do I interpret it?
Please watch
th-cam.com/video/8lDwmMF9aN8/w-d-xo.html
I appreciate this video sir. I have one question. In my findings, the relationship between IV and Mediator is non significant. Both the total effect and direct effect are significant and but indirect effect is non significant. and Preacher and Kelley (2011) Kappa-squared is also significant. is there still exist mediation?
Thanks. I am glad you liked it.
In simple terms if the index of moderated mediation has a 0 in between LL and UL, this is insignificant. Otherwise, moderated mediation is significant.
@@researchwithfawad sir this is about mediation run Model 4. is there exist mediation?
If indirect effect is insignificant, there is no mediation.
On the final where you reported the key statistics of the analysis, how did you get 5.583 the t-statistics. I guess I can’t trace it to the output of the analysis.
I get it now. You divided the indirect effect size by the Boot standard error. Thank you for your videos.
Perfect
How did you get the t-statistic of 5.583 for the indirect effect?
Thanks for watching. Divide the Effect by Standard Error to get t statistics
Hi I have a question I was wondering where you got your t-statistic from in the final table?
Thank you
Please divide Estimate by Standard Error to get t statistics
@@researchwithfawad Thank you, this video really helps for my thesis that is due soon!
PLEASE HELP I HAVE A DEADLINE: my indirect effect is significant. my direct effect from X to Y is not and my total effect is not either. is this a mediation then?
Yes. If the indirect effect is significant, there is mediation.
Sir , Independent variable is having 5 items.... So how to have one value of X to run Process?
Take the mean of the 5 items.
Sir thanks for the video it's really so simple and understandable ..I wanna ask my study have 5 IV and one mediator which is a mediator of all these 5 IV and there is one DV so which model number I use in hyes process ..do I do it separately one by one by placing One IV one M and one DV or I go for one step ..by placing all IV in one box at a time ...I have covariates as well ..kindly guide ..I shall be thankful
Thanks. I am glad you liked it. Please use Model 4.
Thanks for reply sir ..then I did it separately by adding one by one IV and do separate analysis .and I made list of 5 tables or only one table ??
Make one table.
At the of the video, you showed t-statistics in the mediation report. Where does this t-statistics come from?
*at the end of the video..
Thanks for watching. You can have t-statistics by diving Path Coefficient/Effect size by BootSE.
Sir, what is the difference between standardized and unstandardized coefficients?
This link may help in providing a more detailed explanation.
www.statology.org/standardized-vs-unstandardized-regression-coefficients/
Sir if both CI values (upper and lower) are negative, does it indicate there is no mediation?
Yes
I have a question: My p value in direct effect of X on Y is not significant. In Indirect effect of X on Y, the BootLLCI is -2.6148 and BootULCI is -.2033. In this case, i have a mediation?
Yes, there is mediation.
@@researchwithfawad I really appreciate your very quick response. Just one more question:
For example, if BootLLCI was -.2567 and BootULCI was -.0728 would there still be a mediation?
Yes
Very very helpful thank You.. just wondering if we have two groups in the IV (experimental vs control) and using mediation analysis how to know which group has the most effect on the DV through the mediator
If the effect size is positive, The IV that has been given the value one, carry as significant impact on the other variable.
Thank you for this tutorial! Very helpful..
In fact, I have a question : could I use Process Macro to test mediation even if my variables are kind of higher order construct? If yes, how can I proceed? Thank you!
Pleasure. I am glad you liked it. Yes you can, but you have to create a composite variable.
Hello Dr. Fawad. If I have 6 IVs, do I have to run the process one by one? Its different from using multiple regression right? In multiple regression (SPSS), we can enter all the IVs in the Block 1 (Linear Regression ). Correct me if Im wrong.
Yes, you will need to run the macro multiple times. It is better to use SEM tools like SmartPLS4
Thank you, Dr. for your reply. Credit to you for making this video. It helps a lot! I have one more question. Hope you dont mind. May I know how to get the t-statistics value as shown in the mediation table? I've read your reply to the same question from the comments. But, I couldn't find the way to calculate it. Please assist. Thank you, Dr.
Thanks for watching. Divide the path estimate by Standard error
How you write t-statistics in last table ? I don’t find this value in your video
To get t statistics, divide Effect size by Boot SE.
I'm so very gratefull see this video and i want to asking you sir, my path B its not significan with p = 0.0192. but for indirect effect its significan. Can i say my mediating variabel its significan and can i say its mediating?
Thanks. 0.0192 is significant.
Hi, can I use this analysis also when my DV is transformed into a logarithmic variabele? Because of a violated homoskedasticity assumption, I transformed the DV into a log, but not sure is this is necessary for Process Macro Analysis.
Yes, you can still use it.
Very helpful thank you! Do you have any videos on Testing Assumptions for Regression Models?
I am glad you found the video helpful. I am afraid, nothing yet on Assumptions. Hopefully soon.
Is there any prerequisite for IV or DV? IV and DV should be the factors after they are signidicantly associated with the outcome by regression? Or significant relationships with the outcome by Pearsn Correlation Coefficient ? Thank you. Would love to hear from you.
Thanks for your interest. I guess this series will help.
10MinResearch: th-cam.com/play/PLb7vm6tsQ3KtGP-Aq1Hx4IQltUyPnpqLl.html
@@researchwithfawad Thank you so much other your lectures, I watched it and they are also good and help me a lots. My question is that the IV and DV should be checked by regression that they are significant relationships between the outcome before put in the mediation anaysis?
No, IV and DV. That was Baron and Kenny Approach. Please watch
th-cam.com/video/Zu-P_KE0x4c/w-d-xo.html
Hello very nice videos you made,,,,,,,have you covered this concept by stata software? if yes, can you please provide link for the same.
Thanks for watching. I am sorry I do not use Stata.
Hi, I have a question. Is it still possible if the Independent Variable is categorical (based on to conditions 1 vs 0)?
Yes IV can be Dichotomous.
Hi Sir, thanks for sharing!
I want to ask what I should do if the mediator is a categorical variable? Just put it on the bracket as you did in the video?
Thanks!
Pleasure. I am glad you liked it.
Yes, the procedure remains the same.
@@researchwithfawad thank you!
Excellent explanation. I have a question regarding the structure of a thesis. Should students also include descriptive statistics, tests of normality and so on, along with the results of the mediation analysis?
I will recommend that descriptive statistics are reported. As for Normality, instead of more conservative approach, one can report Skewness and Kurtosis to present distribution of data.
@@researchwithfawad Thank you very much for the reply. I will make sure to recommend your channel to my classmates.
Sorry to bother you but I have one more question. If the data from one of the variables is skewed e.g. Z score is more than +/-3.00, can I still go ahead with the mediation analysis?
Why indirect effects of X on Y does not explicitly reveals t value and p value , in other case , direct effect of x on y ,it has effect , se, t value, p value , LLCI and ULCI . pls enlighten
This is a limitation of the Macro.
How to do mediation analysis when we have more than one independent variable.
Thanks for your comment. Please use SMART-PLS or AMOS
@@researchwithfawad In this situation, can I use SPSS to do it? Moreover I also want to ask whether mediation analysis can accept or reject a hypothesis or not. Do we need to make a multiple regression analysis first before making mediation analysis
No, SPSS cannot handle two IVs at a time. Plus The analysis you are doing is regression analysis. No need for separate analysis.
I got perfect mediation(P0.05,. Can I conclude this is a perfect mediator?
If your indirect effect is significant, this shows that there is full mediation.
@@researchwithfawad Thank you so much, it helps my understnding of indirect effect.