Why Confidence Intervals are better than P-values

[@bugbumble3]

I have been following you since last 7years and your lectures are still the best for the topics provided.

[@yahyaal-anbaki9279]

No one could explain it as good as you did! thank‘s alot.

[@MicaelaWeinberg]

Outstanding video! Clearest video on CI i've watched, thank you!

[@melissarobinson4125]

Loved this video. Calm voice. Thank you!

[@66IOU3]

Thank you Mike Wazowski, that was super helpful.

[@brandomiranda6703]

confidence intervals don't say for specific values of the CI that the true mean is in that range. It instead says that if you repeated the survey (e.g. obtained 100 data sets of size n ideally n >= 30) then the mean would be inside those CI 95% of the time. But you don't know for which one and CI is a random variable/quantity. But I suppose you could just approximate that what you said is true since 95% it will be true but as long as you know that is not the proper way to interpret CI I guess it's fine to do that.

[@vijaynagaonkar9087]

Patwari Sir... U r genius... Thanks from India....

[@ignketone2823]

Excellent, thank you very much. Good day.

[@asmaabonbona8409]

Thank you so much! You ROCK!

[@EmmanuelKyaloMuli]

great video!!

[@eddiewilliams2108]

Agreed. C.I. is better than p-value, especially in light of all the p-hacking that been going on for years now, especially in the social sciences and very sadly the medical/drug industry. Are you suggesting that you use the highest C.I. that does not contain 0 or 1 for “difference” and “ratio” systems respectively?

[@ammaraldawoodyeh2078]

Amazing, thank you!

[@joshuafancher3111]

Thank You. Great Explanation.

[@zeelpatel8049]

Thank you so much. Very helpful.

[@Beghnksl8798]

Thank you so much Sir. Very clear

[@tadessegudeta9715]

Thank you! It's a very helpful video!

[@LEARNING-MEDICAL-EDUCATION]

plz explain :if confidence limit is increased, previously significant data now becomes less significant

[@doctorjakkrit]

what is your program or application use in your presentation ?

[@jimmccrory4647]

Why not look at a lower one-sided confidence interval that infers, say a 90% Confidence that the true value will be X or greater? Wouldn’t that a more useful way to present the inference?

[@pixzers]

Thank you. This was very helpful!

[@zakirzak1494]

Great ...... very informative , thank you

[@genovevaaguilar520]

Hard to hear but really good video, thanks!

[@MrDrjonty]

Very cool. What program do you use to create the cool hand drawn slides?

[@sonikagoel4686]

very good explanation.. thanks

[@shafiqahmadchughtai8382]

Excellent.

[@user-uj4xl5ct5y]

what if a CI ranges from 0.002 to 0.21? Will it still be statistically significant ?

[@hakobharutyunyan5972]

thank you,very helpful

[@nourashaheen487]

I cannot thank you enough 🙏 If I can know the name of the software or program you are using to present this lecture?

[@oliviao4311]

really helpful! thanks

[@kershalramchurren5538]

Good stuff...

[@svenhuber6533]

well explained

[@harnett1011]

excellent.

[@drdebdattabhattacharyya904]

That's great

[@raecooker]

Why did you pick 5 and 44 as your interval limits?

[@gevxgu250]

probably by bootstrapping

[@Persason]

Lovely video. Thanks. One question. So you can have a P value that is statistically significant and at the same time get a CI that is not as it crosses the point of no difference? Which CI are typically accepted in journals 95% or 99%? Please do speak up or get a better microphone for you next video then I had to turn up windows and my headset and my amp to 100% just to barely hear you :D.

[@nombuzoconini8324]

can i get midwiry video

[@alaachammakhi931]

the content is good but try to speak louder

[@hc2159]

can't hear anything. your voice is barely present

[@Commando303X]

I barely can hear this damned thing.

[@bhanund]

Content is good but audio is bad.

[@akshaypatil4480]

too low voice. poorly audible even after full volume

[@oldblueday]

+Akshay Patil I put a copy up that is louder. It's here: kzbin.info/www/bejne/Z6W6mZhqm7yjY9E