SPSS #22 Sobel Test For Mediation - Multiple Linear Regression

Hi guys and welcome to the 22nd episode of my SPSS Tutorials. If this video is helpful to you, please like and subscribe.

[@amilachamali]

This was really helpful. Very clear and on point. Im doing my dissertation and I searched for this test. This was the clearest one I found. Thank you so much !!

[@natinamckessey8242]

Super helpful! Made my night much easier.

[@DS-ux5bf]

According to the online calculator, values for path b are obtained by running a regression that includes both the IV and mediator as predictors (not the mediator alone). You need to control for the effect of the IV on its own in order to determine whether its effect on the DV is being mediated by your third variable.

[@baspsstutorials9783]

Thanks for pointing that out, you are very much right! Mistake on my behalf

[@baspsstutorials9783]

As a viewer rightfully corrects me, at 4:32 I should have added job as well as an independent variable, so both job and salary as independent. In the output you only have to look at the values of salary though

[@codehanieh9965]

I was about to write a comment on that.. thank you for the clarification. However, if you have 2 independent variables you would have two t(b). I do not know how to run the sobel test on the webpage that you have introduced. On that page, there is only one filed for t(b).

[@joyiedeman]

Bedoel je dat je ze dan allebei tegelijkertijd toevoegt of bedoel je dat bij de de eerste t waarde die je berekent, een baan ook onafhankelijk had moeten zijn?

[@joyiedeman]

Top uitleg! Helpt goed bij een beeldvorming van de formules etc.

[@zulhilmizulkurnain6823]

Thank you so much, this video really help me a lot

[@mindfulisle]

Thank you so much.

[@yijiang2546]

Thanks! How do you explain the sign of the Sobel test statistics? Especially if the test statistic is negative.

[@baspsstutorials9783]

What do you mean with sign?

[@yijiang2546]

@@baspsstutorials9783 From the video you said that if the Sobel test statistic is above 1.96 or below -1.96, it is significant. My question is do above 1.96 and below -1.96 have a different meaning? Thanks!

[@baspsstutorials9783]

@@yijiang2546 oh now I get it! No it has no important difference. Above 1.96 or below -1.96 both mean that the result is significant.

[@ismailj.ismail4921]

Members! How can i get a SOBEL plugin for SPSS?

[@kossonouprunelle7576]

My mediating variable is ordinal variable, the dependent and independent variable are also categorical variables. can i use this method to test the modiation effect? thank you .

[@kelseycoulthard8798]

Hello, thank you for explaining this! If I have covariates to control for, should I include these in my models before running the Sobel tests? Right now, when I include these covariates, my Sobel test is not significant (.0544) but when I don't include them, similar to this video, the test is significant. Which is the appropriate way to run this?

[@chrisknowles7471]

Did you ever get a reply for this? I have the same problem...

[@rajeevdutraj8642]

After putting the values I got the p-value as 0.0, what does it mean?

[@Danielle-cw3gy]

Dag Bas, ik heb een vraag over bootstrapping. Ik heb voor een simpele mediatie analyse de CS methode gebruikt van Baron & Kenny (1986). Nu wordt mij aangeraden om de significantie van het indirecte effect te berekenen d.m.v. bootstrapping i.p.v. de Sobeltest. Ik kom er niet uit hoe dit te doen. Kun je mij op weg helpen?

[@baspsstutorials9783]

Hi Daniëlle. Ik moet eerlijk bekennen dat bootstrappen nog op mijn to-learn lijstje staat. Ga er wel videos over maken maar dat duurt helaas nog even, ga eerst alle anova's doen. Maar er is vast een ander youtubekanaal dat het je kan uitleggen! Ik leer mezelf ook alles aan via youtube namelijk. Voor niet-bootstrap vragen kan je me altijd bereiken, groetjes, Bas

[@Danielle-cw3gy]

@@baspsstutorials9783 Dankjewel Bas, komt goed! Succes met je kanaal👍

[@baspsstutorials9783]

Na deze video nog steeds hulp nodig bij SPSS? Bekijk dan deze video: kzbin.info/www/bejne/hJm5e4pmgdullacsi=MRha5EE03xcIItjE