Thank you. I'm not fluent in english but your explanation are good. Just one thing missing for me : why do you say r = .856 is a large effect ? Is there a table to know the strength of effect by its number ?
Can you tell me which test to use if I have two separate groups (control and exp) tested on two occasions (pre- and post tests) , sample size of each group is 62, non-parametric data, one independent variable and one dependent variable (ordinal)
@@PatLalka Thank you for your prompt reply. I am still unsure whether a mixed ANOVA is appropriate because my data fails one of the mixed anova assumptions (dependent variable should be continuous)
Hey, I think you can use this formula here z = (x - μ) / σ. But also to my understanding the normality, Shapiro-Wilk, test does a similar thing in telling you how well distributed your data is. Hope that helps
i love this guy rn for this
OMG Pat enjoy is hardly likely on my part ;). BUT THANK YOU you are an angel, so well described and helpful. Hope you have a lovely summer, thanks!
THANK YOUUUU!!! You're a life-saver.
This is amazing!
Thank you. I'm not fluent in english but your explanation are good.
Just one thing missing for me : why do you say r = .856 is a large effect ? Is there a table to know the strength of effect by its number ?
Is this correct ? -> Cohen's classification of effect sizes = 0.1 (small effect), 0.3 (moderate effect) and 0.5 and above (large effect).
Sorry for the late reply! Yes that's perfect
Very nicely explained 👏
Thaks Bro, you save me haha
I am having a difficulty interpreting my man Whitney u test huhuhu
What does it mean if p value is 0.388 and statistic is 326 using Mann Whitney U Test?
If p value is above 0.05 it means there is no statistical difference between your variables
@@PatLalka Thank you so much !! Great Tutorial
Can you tell me which test to use if I have two separate groups (control and exp) tested on two occasions (pre- and post tests) , sample size of each group is 62, non-parametric data, one independent variable and one dependent variable (ordinal)
Think you’re looking for a mixed factorial anova here 💪
As long as the groups tested don’t have the same subjects in it - if they do use a repeated measures
@@PatLalka Thank you for your prompt reply. I am still unsure whether a mixed ANOVA is appropriate because my data fails one of the mixed anova assumptions (dependent variable should be continuous)
If you send me your data on my email I can have a look for you
@@PatLalka great!
Do you know how to calculate z when using jamovi to run MWU test?
Hey, I think you can use this formula here z = (x - μ) / σ. But also to my understanding the normality, Shapiro-Wilk, test does a similar thing in telling you how well distributed your data is. Hope that helps
@@PatLalka THANKS a lot, it's really helping!