Yes, I suppose we have all found using the summary function for models messy at some point (although R still makes it easy compared to other software). These "flaws" are common in learning, teaching, and even authoring, so hopefully we will not repeat them again. In fact, thanks to this video I am in the process of correcting them in my current R project. PS: Mr. Zablotski, as always, your content does not disappoint, the quality of your videos is solid. Not a single video on your channel loses its consistency. Really useful and profitable. Every time I get the notification of a new video I am 95% sure that it will reach the high bars. God bless you man! PS2: Your videos are addictive, each one makes me want to watch another HAHAHAHA.
Wow, this comment absolutely made my day! :) I'am so happy the video was helpful. But even better - I absolutely love hearing that you're already putting what you learned into practice with your current project - that's fantastic, and that's what I wish people do - apply this knowledge directly! And thank you so much for the kind words about the channel! I try to deliver high-quality content, but never know whether it's good. Thus, your feedback means the world to me and keeps me motivated to keep creating!
I just ran into the problem you described at the start of the video a few days ago...namely trying to report confidence intervals, interactions, and so on and was struggling to do so using only the summary function. Thank you so much for this video, it had some really great tips in there and you explain it so concisely and clearly that it actually becomes interesting (and I never thought I would say this about statistics to be honest :D). You're a natural teacher and i'm sure your work and effort is greatly appreciated, at least it is on this end! Thanks!
Thank you so much for such a nice feedback! And I am very happy I am not alone :) who finds the summary function fairly limited. I am also glad my content it's helpful for more then just me. That motivates me to produce more. Cheers!
Thanks a million for this super-helpful video. I learned about the emmeans package feom your emmeans video and i was blown away. Academic publications usually don't include the means of each category and i dont know why it's NOT required. I recently published a paper using emmeans to show the group estimated averages based on the model and i got a lot of push backs. Please, when you're able to get to it and if it interests you, could you please review ggpredict, ggemmeans and emmeans, their similarities and differences and when appropriate to use one or the others. Thank you.
that's an a great idea! I do usually compare those functions and most of them use emmeans in the background. For instans, ggemmeans and tbl_regression with contrasts all use emmeans. So, getting back to the sorce - emmeans is the way to go. The results could be tricky though, because the gtsummary never say it uses Tukey correction by default, so the results were strangely different from the summary and co. But yeah, thanks for the suggestion, I'll consider to build them into the next videos and thanks for watching and commenting, this is the best support!
oh, and by the way, I do publish a lot of academic papers where I use contrasts and I usually use prediction plots (effect plots) + the table with contrasts and it usually has no problems with reviewers. The pushback happens sometimes when you hit an old reviewer, who used the same sh...t the whole life and thinks he/she known everything better. just politely ignore and continue to contribute to science ... that's what I tell myself every time I get stupid requests, like ... why don't you just use ANOVA ... for everything. That really happens a lot!
I played with the gtsummary: tbl_regression function, but there is no add_pairwise_contrasts parameter in the add_ dropdown list. BTW, thanks for the effort.
sure, it's not explicitely there, but if you run the code, it works. the package improves steadily and new functions are added all the time. another dependent and very cool package you can a look at is broom.helpers. they work together in parrallel, so if you did not installed it already, please do, and explore
I really like the content, not easy to find about packages like sjplot. But, to me, there is too much video effects. This is not necessary for such high quality content.
Thanks, Thiago! Your feedback is greatly appreciated! I already waited for somebody to say that the video might be overstimulating, so, you are the first who noticed. Congrats. I am still learning creating content. One of the principles is - show, don't tell. Therefore, I try to show everything, even text. Which, even for me is a bit too much. So, when you could point out which parts of the video, or which elements were distracting, and which parts were very good, I would be able to improve. Thanks!
Thank you for the video. I have always enjoyed watching your videos and blogs. But I just have found out that the blog webpage does not open nor the link function. Is your blog shut down, if so are you planning to open it again?
thanks for the feedback. yes, they shut down my blog since they want me to pay. so, it'll take some time to find another alternative, since I am not an IT guy. thus, yes, I plan to reopen it, but have no idea when this will happen. but since RUclips is still free, feel free to use videos, since my blog is actually the script for the video, word by word, code by code. thanks for understanding
most of the function yes. there will be some nuances though. if you try and let forks know here in the comments, that would be awesome and helpful for others! thanks for watching!
No problem. I had it too on my new computer, but then I installed one of the dependences (just another package) of the {performance} package and it worked. Forgot which exactly it is. So, just find out all the dependences of perfomance and sjPlot packages, install them and it will work. Please, write here in the comments for other folks which might have the same problem what package is was. Thanks forward.
thanks for the feedback. yes, they shut down my blog since they want me to pay. so, it'll take some time to find another alternative, since I am not an IT guy. thus, yes, I plan to reopen it, but have no idea when this will happen. but since RUclips is still free, feel free to rewatch videos, since my blog is actually the script for the video, word by word, code by code. thanks for understanding!
Mr Zablotski, I just found a video about the R packages "marginaleffects" and "modelsummary" on this college conference video ruclips.net/video/uyNhK18LLYU/видео.html. Have you heard about them? What do you think about these packages and their features? Would you consider creating a package review video for them? No hard questions, just wanted to share this information and get more knowledge about it. Greetings from Colombia.
Hey mate, thanks for the useful info! I did touch them once, but wasn't diving deep into it, because they were (despite being amazing tools!) not as quick and as intuitive as for example emmeans or gtsummary. But I'll have a look at the video and revisit the packages, and put them on the list, when I see some utility for data science. Thanks for watching by the way!
Yes, I suppose we have all found using the summary function for models messy at some point (although R still makes it easy compared to other software). These "flaws" are common in learning, teaching, and even authoring, so hopefully we will not repeat them again. In fact, thanks to this video I am in the process of correcting them in my current R project.
PS: Mr. Zablotski, as always, your content does not disappoint, the quality of your videos is solid. Not a single video on your channel loses its consistency. Really useful and profitable. Every time I get the notification of a new video I am 95% sure that it will reach the high bars. God bless you man!
PS2: Your videos are addictive, each one makes me want to watch another HAHAHAHA.
Wow, this comment absolutely made my day! :) I'am so happy the video was helpful. But even better - I absolutely love hearing that you're already putting what you learned into practice with your current project - that's fantastic, and that's what I wish people do - apply this knowledge directly! And thank you so much for the kind words about the channel! I try to deliver high-quality content, but never know whether it's good. Thus, your feedback means the world to me and keeps me motivated to keep creating!
I just ran into the problem you described at the start of the video a few days ago...namely trying to report confidence intervals, interactions, and so on and was struggling to do so using only the summary function. Thank you so much for this video, it had some really great tips in there and you explain it so concisely and clearly that it actually becomes interesting (and I never thought I would say this about statistics to be honest :D). You're a natural teacher and i'm sure your work and effort is greatly appreciated, at least it is on this end! Thanks!
Thank you so much for such a nice feedback! And I am very happy I am not alone :) who finds the summary function fairly limited. I am also glad my content it's helpful for more then just me. That motivates me to produce more. Cheers!
I just wanted to say a big thank you for the amazing video! Your content is incredibly well-arranged and explained.
I appreciate that a lot! thanks you for watching and commenting, that's the best support!
Thanks a million for this super-helpful video. I learned about the emmeans package feom your emmeans video and i was blown away. Academic publications usually don't include the means of each category and i dont know why it's NOT required. I recently published a paper using emmeans to show the group estimated averages based on the model and i got a lot of push backs.
Please, when you're able to get to it and if it interests you, could you please review ggpredict, ggemmeans and emmeans, their similarities and differences and when appropriate to use one or the others.
Thank you.
that's an a great idea! I do usually compare those functions and most of them use emmeans in the background. For instans, ggemmeans and tbl_regression with contrasts all use emmeans. So, getting back to the sorce - emmeans is the way to go. The results could be tricky though, because the gtsummary never say it uses Tukey correction by default, so the results were strangely different from the summary and co. But yeah, thanks for the suggestion, I'll consider to build them into the next videos and thanks for watching and commenting, this is the best support!
oh, and by the way, I do publish a lot of academic papers where I use contrasts and I usually use prediction plots (effect plots) + the table with contrasts and it usually has no problems with reviewers. The pushback happens sometimes when you hit an old reviewer, who used the same sh...t the whole life and thinks he/she known everything better. just politely ignore and continue to contribute to science ... that's what I tell myself every time I get stupid requests, like ... why don't you just use ANOVA ... for everything. That really happens a lot!
❤
I am impressed, huge respect for you, doing excellent work.
I am also adding a request make a video on metafor R package
wow, thanks for such a generous feedback! Great idea! I'll put metafor on a to-do list for sure!
Excellent one. Thanks. Help us with the codes.
thanks for watching!
Thank you very much!
You're welcome! Thanks for watching and commenting!
Great content as always. Very informative
Much appreciated! Thanks for watching!
Very useful. Thanks
Glad you think so! And thanks again for watching!
I played with the gtsummary: tbl_regression function, but there is no add_pairwise_contrasts parameter in the add_ dropdown list. BTW, thanks for the effort.
sure, it's not explicitely there, but if you run the code, it works. the package improves steadily and new functions are added all the time. another dependent and very cool package you can a look at is broom.helpers. they work together in parrallel, so if you did not installed it already, please do, and explore
I really like the content, not easy to find about packages like sjplot. But, to me, there is too much video effects. This is not necessary for such high quality content.
Thanks, Thiago! Your feedback is greatly appreciated! I already waited for somebody to say that the video might be overstimulating, so, you are the first who noticed. Congrats. I am still learning creating content. One of the principles is - show, don't tell. Therefore, I try to show everything, even text. Which, even for me is a bit too much. So, when you could point out which parts of the video, or which elements were distracting, and which parts were very good, I would be able to improve. Thanks!
Great video. Thank you for that. Can it also be applied to GLM models such as Poisson regression?
Yes, most of the functions can! Try them all out. Thanks for watching!
is there any update on when your blog will be back up? I have been checking over the last couple weeks
not yet. I am working on it, but don't know when this will be fixed. I am not the IT guy myself, so, it takes time. Sorry!
Awesome all methods in one place ❤
I thought this could be overwhelming? But after watching some videos which speak half an hour about summary function, I couldn't hold back anymore ;)
Thank you for the video. I have always enjoyed watching your videos and blogs. But I just have found out that the blog webpage does not open nor the link function. Is your blog shut down, if so are you planning to open it again?
thanks for the feedback. yes, they shut down my blog since they want me to pay. so, it'll take some time to find another alternative, since I am not an IT guy. thus, yes, I plan to reopen it, but have no idea when this will happen. but since RUclips is still free, feel free to use videos, since my blog is actually the script for the video, word by word, code by code. thanks for understanding
Are these packages suitable for mixed effect models, such as GLMM, CLMM, or models based on special distributions, such as beta regression?
most of the function yes. there will be some nuances though. if you try and let forks know here in the comments, that would be awesome and helpful for others! thanks for watching!
2:54 Why does my residuals plot lack confidence intervals?
No problem. I had it too on my new computer, but then I installed one of the dependences (just another package) of the {performance} package and it worked. Forgot which exactly it is. So, just find out all the dependences of perfomance and sjPlot packages, install them and it will work. Please, write here in the comments for other folks which might have the same problem what package is was. Thanks forward.
I recently attempted to visit your blog. However, I encountered a ‘Site Not Found’ error. Could you please assist me in resolving this issue?
thanks for the feedback. yes, they shut down my blog since they want me to pay. so, it'll take some time to find another alternative, since I am not an IT guy. thus, yes, I plan to reopen it, but have no idea when this will happen. but since RUclips is still free, feel free to rewatch videos, since my blog is actually the script for the video, word by word, code by code. thanks for understanding!
Great video as always. I could help with the blog issue. Care to discuss?
Thank you so much! That’s very lovely from you! But a friend of mine who is an IT guy, will have a look at it.
👏👏👏👏
🙏 🎉 👍
Mr Zablotski, I just found a video about the R packages "marginaleffects" and "modelsummary" on this college conference video ruclips.net/video/uyNhK18LLYU/видео.html. Have you heard about them? What do you think about these packages and their features? Would you consider creating a package review video for them? No hard questions, just wanted to share this information and get more knowledge about it. Greetings from Colombia.
Hey mate, thanks for the useful info! I did touch them once, but wasn't diving deep into it, because they were (despite being amazing tools!) not as quick and as intuitive as for example emmeans or gtsummary. But I'll have a look at the video and revisit the packages, and put them on the list, when I see some utility for data science. Thanks for watching by the way!