Jamovi Tutorial: Mediation

Download data here: drive.google.c...

[@mariagracielanajes5833]

Hi, Sir Rene! I want to say thank you for these videos. Since the day I started writing a research paper that includes mediation, I have been watching your videos. This helps me a lot, especially in understanding mediation and its paths.

[@DarthKidd]

Thank you sooooo much! My professor kept telling me something was partially mediated in our data set, and I was confused since there was a statistically signifiant path. I didn't know about if both direct and inirect effects were significant that that meant it was a partially mediated. THANK YOU!!!

[@vitalisugwu1429]

Thanks a lot for supporting your video with a dataset so that your viewers can follow up.

[@douglasthai1960]

Hi, may i know how to interpret below for a simple mediation: Indirect Effect: 0.47 component: 0.31 component 0.15 Direct .043 Total .09 All with significant results but seems the numbers are very small Does this considered as a weak mediation? Do I have to address in my paper? Thank you so much

[@renenob]

You can watch this tutorial: kzbin.info/www/bejne/noXUlJiIe6yKa9Usi=EGnPKfq1DJy8l1_o

[@yingxuanzhu8564]

Thank you for such a helpful video! I wonder if you can also give a lecture on the detailed regression process, and that would be great!

[@keishapatrice8200]

This is so helpful sir Nob!

[@lahoktc2019]

thank you so much sir!! very helpful

[@lorenzomiguelgarcia842]

Hello sir Nob, may I ask what it will mean if the estimates are all negative and all effects have an insignificant p-value? How may I interpret that and put it into context? Thank you

[@renenob]

Well, if the coefficients are negative (and significant), that simply describes the direction of the relationship. Specifically, the increase in one variable predicts a decrease in the other variable, and vice versa. If the results are not significant, it is useless to interpret the direction (positive or negative). It simply means that you do not have evidence to support the claim that the variables are related to one another.

[@sg7696]

Does data need to be normally distributed for mediation analysis?

[@renenob]

More often than not, our data is not normally distributed. Since this is regression, I think the more important assumption to satisfy is the normality of the residuals.

[@compaq2804]

What happens when the indirect effect is insignificant? How do we report it?

[@renenob]

you can simply say the the IV no longer predicts the DV after accounting for the mediator. Or your can say that the mediator full accounts for the influence of the IV on the DV.

[@claravassal9081]

Hello !! Thank you for this very helpful video… I still have a lot of questions for a university work in statistics… if you could help me ?

[@renenob]

Sure, I will help if I can.

[@melisturkoglu2384]

do we need to standardize though?

[@renenob]

I guess it's not really a need. It would have just been nicer to have it.

[@melisturkoglu2384]

@@renenob yes i agree thanks