The best lecturer I have ever met, still coming back to your videos even I am now doing honours in another uni. But due to my experimental error and other stuff, the two-way RM ANOVA I conducted in sigmaplot reports back violation of normality and equal variance... no way of saving me from this disaster. The data collection is done, but a sample size was only 10... such a shame ... I know I should have done better. LOL
thank you so much for the straightforward explanation. I got one question and I would appreciate it if you could answer it for me. you mentioned that regardless of the scale used in the study, the effect size can be measured. I am confused about this point as I am trying to combine five different scale for quality of life in heart failure that follow different scoring system in one meta analysis using Cohen's d. Some scales relate high scores to good quality of life while some relate high scores to poor QoL. I was advice not do it as the scales are totally different. statistically, dose the method of measurement used affect the final conclusion about effect size?
In this case, where you are combining different Cohen's d values together, you have to decide the direction of effect you want to interpret and discuss. So, in a meta-analysis, you would decide to flip a negative Cohen's d to a positive Cohen's d to be consistent with the interpretation of the other Cohen's d values, for example.
Thank you so much for making this series! I've been trying to get my head around Hedge's G and whether or not to use it.
The best lecturer I have ever met, still coming back to your videos even I am now doing honours in another uni. But due to my experimental error and other stuff, the two-way RM ANOVA I conducted in sigmaplot reports back violation of normality and equal variance... no way of saving me from this disaster. The data collection is done, but a sample size was only 10... such a shame ... I know I should have done better. LOL
Very helpful concised presentation!
Many thanks!
You are amazing !!! Thank you so much !!!
Thank you. Very helpful
Doesn't Hedge's g use wrong values at 4:40? Shouldn't it be n1 + n2 and not SD1 + SD2?
yes you are right
not s*s but n
what if I get a D's Cohen value of 1.6548? what is the interpretation?
thank you so much for the straightforward explanation. I got one question and I would appreciate it if you could answer it for me. you mentioned that regardless of the scale used in the study, the effect size can be measured. I am confused about this point as I am trying to combine five different scale for quality of life in heart failure that follow different scoring system in one meta analysis using Cohen's d. Some scales relate high scores to good quality of life while some relate high scores to poor QoL.
I was advice not do it as the scales are totally different. statistically, dose the method of measurement used affect the final conclusion about effect size?
In this case, where you are combining different Cohen's d values together, you have to decide the direction of effect you want to interpret and discuss. So, in a meta-analysis, you would decide to flip a negative Cohen's d to a positive Cohen's d to be consistent with the interpretation of the other Cohen's d values, for example.
excellent
for cohens d at 4:33 in the video i keep getting +0.705 :S