I see that adapting pharmaceutical industries data is becoming more necessary, because the guidelines and general chapters in pharmacopoeia is making it necessary to assess the variability of the manufacturing and analytical test method. And we have decided that when doing comparison of two methods to prove equivalence, we choose equivalence margins (preferably based on the variability) and check if the confidence intervals are within the equivalence margins. It has been refreshing to study up on statistics and your videos are interesting and have been a great help!
This is wonderfully presented and makes a fairly complicated subject understandable to non-statisticians who need to see through data. Wonderful work. Thank you.
Thanks for this video. The CIs require the same "testing" supplement that significance tests do, that is one cannot avoid the falacies and weaknesses of statistical tests.
Very nice video !! however, most of biological parameters (mainly those used in the drug discovery) do not follow a normal distribution. Its is a log-normal distribution. Normaly in this case we perform a geomean with confidence intervals for discribe the data and non parametric methods for analysis. Or, that is more recomended, transform the parameters to a log of the parameters, and in this case the distribution become normal, and apply the parametric analysis. What is more recomended for use this cohen's d: brings the parameters to a normal distribution or keeps the log normal distribution?
Wish you were our Biostatistic professor at school of pharmacy back in the days. However, it's never too late I'll turn on notifications button and stay tuned.
Thank you for your video! Complicated ideas were elaborated clearly. Could you please make a video to compare the difference between ANCOVA and Regression when we apply them to quasi-experimental design study? Many textbooks suggest that when we cannot meet parallel slopes assumption of ANCOVA, we can use Regression. What are pros and cons?
Why not using CI on the Cohen's d as well? Reporting just a point score is highly misleading, no matter if its Cohen's d or p or t value. Moreover, since H1 is usually two-tailed, there is no real difference between CI and p value; if the Ci includes H0 (usually 0), then the P value is larger than a-prior alpha (usually 0.05), snd vice versa: if the result is significant, p value smaller than alpha, then CI does not contain H0. This was s nice intro to stats video, but i dont see snything new here. Its merely a few guidlines for better standards in scientific publicstion.
smells like anti-science, to say that p-value is so complex. Science is about robustness, assure about something, not about turning simple to anyone understand. Are you sure your are not anti-science? Try watching Idiocracy (the movie), and guess how far you need to go to be simple enough.
If you like, please find our e-Book here: datatab.net/statistics-book 😀
I'm so glad TH-cam recommended me this video. I love the editing style and the structured explanation.
Awesome, thank you! Regards Hannah
Thank you DATAtab team!
I see that adapting pharmaceutical industries data is becoming more necessary, because the guidelines and general chapters in pharmacopoeia is making it necessary to assess the variability of the manufacturing and analytical test method. And we have decided that when doing comparison of two methods to prove equivalence, we choose equivalence margins (preferably based on the variability) and check if the confidence intervals are within the equivalence margins. It has been refreshing to study up on statistics and your videos are interesting and have been a great help!
This is wonderfully presented and makes a fairly complicated subject understandable to non-statisticians who need to see through data. Wonderful work. Thank you.
Thanks for this video. The CIs require the same "testing" supplement that significance tests do, that is one cannot avoid the falacies and weaknesses of statistical tests.
Great video on a modern approach to data analysis. My question is what if we don't have a second study for meta analysis?
Which editing software you used to making this video's. How you make this videos, plz share this process.
Please share
Interested too
Very nice video !! however, most of biological parameters (mainly those used in the drug discovery) do not follow a normal distribution. Its is a log-normal distribution. Normaly in this case we perform a geomean with confidence intervals for discribe the data and non parametric methods for analysis. Or, that is more recomended, transform the parameters to a log of the parameters, and in this case the distribution become normal, and apply the parametric analysis. What is more recomended for use this cohen's d: brings the parameters to a normal distribution or keeps the log normal distribution?
Well done 💯👍
Thank you! Cheers!
Wish you were our Biostatistic professor at school of pharmacy back in the days. However, it's never too late I'll turn on notifications button and stay tuned.
Thank you for your video! Complicated ideas were elaborated clearly.
Could you please make a video to compare the difference between ANCOVA and Regression when we apply them to quasi-experimental design study? Many textbooks suggest that when we cannot meet parallel slopes assumption of ANCOVA, we can use Regression. What are pros and cons?
Ask to micro soft copilot
The nutritional and exercise sciences are going to hate this video
Hmm, why? : )
Thanks for keep us informed about new statistics method vs tradational.
Most welcome
U should show ur face in all videos 😊
Why not using CI on the Cohen's d as well? Reporting just a point score is highly misleading, no matter if its Cohen's d or p or t value.
Moreover, since H1 is usually two-tailed, there is no real difference between CI and p value; if the Ci includes H0 (usually 0), then the P value is larger than a-prior alpha (usually 0.05), snd vice versa: if the result is significant, p value smaller than alpha, then CI does not contain H0.
This was s nice intro to stats video, but i dont see snything new here. Its merely a few guidlines for better standards in scientific publicstion.
I get this book for free
smells like anti-science, to say that p-value is so complex. Science is about robustness, assure about something, not about turning simple to anyone understand.
Are you sure your are not anti-science?
Try watching Idiocracy (the movie), and guess how far you need to go to be simple enough.