Is there a standard way of reporting emmeans results? How should correct report look like? Should it include the estimate, the SE and the p.value for the constrasts that were found to be significant?
There are some broad principles that should be followed to make your analysis results informative. My list of four desirable items to include are: 1/ write a plain English statement of your result, 2/ quantify the effect (this could be an estimate or a test statistic) 3/ quantify the amount of data informing your result (e.g. degrees of freedom) 4/ quantify the uncertainty/strength of evidence (this could be a 95% confidence interval, or a p-value, or a standard error although I tend to prefer 95% CI over a standard error since it has a clearly relationship to the p-value).
Thanks for the feedback. Use the CLD() function in emmeans to produce groups from a Tukey pairwise comparison. You will need to install the multcompView package for this to work.
Hi Meiling Ming, It is the full fitted model, but it is not the model you have given in your example. This video forms part of a course at UCD. It follows on from this video ruclips.net/video/Rs3OjuvWrII/видео.html which presents the fitted model.
while I was running emmeans, because the number of observations exceeds 3000 and then it's fail. what should I do? or is there other ones suit for running Linear Mixed Models post hoc test? PLEASE~and thank you
For the emmeans question I would post this onto a forum (for example a very common forum is stackoverflow stackoverflow.com/questions/tagged/r). multcomp is another very commonly used package for multiple comparisons. Not quite a user friendly as emmeans, but very flexible. Good luck with your issue.
Dear GourmetGringo Information on the data set used for this video (and download link) is available here: www.ucd.ie/ecomodel/Resources/datasets_WebVersion.html#wolfcsv The video is a very quick run through because it is supplementary media for the stats modules we run at UCD. All the best Jon
![Summer Walker - Heart Of A Woman [Official Lyric Video]](http://i.ytimg.com/vi/SFddctbZzw8/mqdefault.jpg)
Thank you so much for this video. It helped me a lot with my master thesis!
Is there a standard way of reporting emmeans results? How should correct report look like? Should it include the estimate, the SE and the p.value for the constrasts that were found to be significant?
There are some broad principles that should be followed to make your analysis results informative. My list of four desirable items to include are:
1/ write a plain English statement of your result,
2/ quantify the effect (this could be an estimate or a test statistic)
3/ quantify the amount of data informing your result (e.g. degrees of freedom)
4/ quantify the uncertainty/strength of evidence (this could be a 95% confidence interval, or a p-value, or a standard error although I tend to prefer 95% CI over a standard error since it has a clearly relationship to the p-value).
Thank's a lot for this video!!
Very helpful indeed. How can you then group these means, say by letters in the case you prefer to present letters on plots?
Thanks for the feedback.
Use the CLD() function in emmeans to produce groups from a Tukey pairwise comparison. You will need to install the multcompView package for this to work.
@@UCDecomodel Thank you very much, this was so helpful. Looking forward to many more videos from you.
Thanks a lot - very helpful!
Can you please specify what is 'm1' in emmeans(m1, spec='Population'). Is m1 the full model (e.g., m1
Hi Meiling Ming,
It is the full fitted model, but it is not the model you have given in your example. This video forms part of a course at UCD. It follows on from this video ruclips.net/video/Rs3OjuvWrII/видео.html which presents the fitted model.
THANK YOU!!
while I was running emmeans, because the number of observations exceeds 3000 and then it's fail.
what should I do?
or is there other ones suit for running Linear Mixed Models post hoc test?
PLEASE~and thank you
For the emmeans question I would post this onto a forum (for example a very common forum is stackoverflow stackoverflow.com/questions/tagged/r).
multcomp is another very commonly used package for multiple comparisons. Not quite a user friendly as emmeans, but very flexible.
Good luck with your issue.
Is the data file you use available? It is difficult to follow your example without it.
Dear GourmetGringo
Information on the data set used for this video (and download link) is available here: www.ucd.ie/ecomodel/Resources/datasets_WebVersion.html#wolfcsv
The video is a very quick run through because it is supplementary media for the stats modules we run at UCD.
All the best
Jon