JASP 0.14 Tutorial: Analysis of Covariance (ANCOVA) (Episode 26)
HTML-код
- Опубликовано: 15 мар 2021
- In this JASP tutorial, I go through a quick data library example of Analysis of Covariance. Most of the output is similar to the ANOVA (one-way) module, but this video adds the covariate and explanation of the output.
Video of JASP ANOVA: • JASP 0.10.1 Tutorial: ...
JASP: jasp-stats.org
NOTE: This tutorial uses the new preview/beta build of 0.14.1. This build contains slightly more functions/features than the previous builds used for tutorials on this channel, but it is functionally the same for the purposes of this tutorial.
Find me on Twitter: / profaswan
Go to my website: swanpsych.com
Twitch streams on psych & related topics: / cogpsychprof
Discuss this video and others on my Discord channel: / discord
Super helpful, thank you!
great tutorial thank you very much
What should you do, if the data is not normally distributed? Jasp recommends the Kruskal Wallis as a non-parametric option. Though this test does not provide information on Interaction and Covariates, or does it?
Can I still use that test to analyse the effects of my independent variables on my dependent variable?
K-W does not provide detailed info like the ANOVA would. However, the GLM is generally robust to non-normal distributions, so I’d still run the ANOVA. Just be careful in your interpretations of the results.
Why is it that the model differs depending on whether the covariate or fixed factor term is entered first? Is there a correct way to go about this?
It actually shouldn't matter in a GLM model as far as I'm aware, so this must just be a JASP issue. So knowing that, put your covariates in last
@@AlexanderSwan great, thanks so much - and for all your excellent content!
I have been picking my brain about this and I believe you made an error at the end of the video. It seems that JASP ANCOVA graphs don't adjust and basically only show the ANOVA graph. On your exemple both high and low dose have the same mean which shows on the graph but the marginal means are different so if the graph was actually showing the ANCOVA results with the marginal means, the high dose dot would be higher than the low dose, which it isn't. I tried several times on my data and no matter if you add or remove covariates you always end-up with the exact same graph which is quite frustrating. Or am I missing something ?
That seems like it might be a bug with the module on JASP. If I made an error, that’s my bad, I probably didn’t look too closely as I was going through the tutorial. I always recommend making your own graphs in Excel!
how about when the covariates have not been included, the main effect is significant, but after the covariates are included, the main effect and covariates are not significant.
This means your relationship is not significant after partialing out this covariate as a control variable, but it also means that covariate is not a significant predictor for the DV
@@AlexanderSwan Thank you very much for your explanation Sir.