Interpreting confidence intervals for the odds ratio

--Reading research, biostatistics
--Risk factors and protective associations
--Associated hypothesis tests
-- USMLE biostats.

[@collegestatistics]

Thanks for the kind words. I wasn't sure whether people were liking the videos or not. Let me know if there's a specific topic that you'd like a video on. 😊

[@wisdompeprah8242]

Hi, please can you share with me a video on steps involved in using SPSS to conduct principal component analysis on over 100 food items, to get at most two dietary patterns.

[@bryanc.6532]

Very helpful! Thank you for putting this up, looking forward for more of your video. 👍👍👍

[@uamykalsoomkalsoomjutt4859]

Yes sir kindly upload a video on logistics regression its components or example please ...... thankuu

[@freezee7547]

The best straight to the point explain ever on KZbin!

[@miapassas7303]

Hands down the BEST video I have come across to explain OR and 95% CI. This is exactly what I was looking for and you explained it brilliantly. Thank you, thank you, THANK YOU!

[@collegestatistics]

You're very welcome! Let me know if there are any other topics you might be interested in. Because most of my videos have very few views.

[@ambrosecarrejo3395]

Its been over 40 years since I took stats in school. That was an a very succinct summary of some complex concepts. Thank you for making it easy for me to grasp the meaning of these important statistics.

[@collegestatistics]

Glad it was helpful!

[@tusnimmokhtar7394]

The explanation was on point! Thank you!!! I have been looking for a clear explanation with examples like this for months😭 THANK YOU!!!!

[@collegestatistics]

Thanks for watching!

[@danielbl2874]

When you are lucky enough to find vids like this... Statistics is a pleasure! Great explanation!

[@collegestatistics]

Thanks for watching and the kind words. My goal is always to share that statistics is an amazing subject.

[@prosivymishecknyandoro5011]

You're a legend! Many thanks for the legendary explanations 🙌 🙏

[@KuriGohan-ss4oy]

Thank you for this! This was the most practical and coherent explanation I found after watching other tutorial vids.

[@collegestatistics]

You're very welcome!

[@melindajvelez7775]

I'm a physician and review medical articles for consulting that I do. I wanted to review the interpretation for CI since it's been 30 years since I took college statistics. This video was the most helpful out of all the you tube videos that I've watched on this topic. It answered exactly the questions that I had re CI interpretation. I hope the statistician doing these videos continues to create more. I'd also like to know if there is a direct contact to send questions that i'd like answered or covered in a video. thanks again for an excellent explanation of CI interpretation. mjv

[@collegestatistics]

Thank you for your extremely kind words of encouragement. If there's another video you want I'd be happy to do it for you. Or even a few questions that you want answered that might be answerable in a video?

[@pendapalanamundjebo5328]

Agreed! I watched like six before I stumbled across this one and I 95% confident I could explain this to someone as well :P

[@thehoodedvagabum7375]

I don´t even speak english (in daily basis) and I understood this waaaay better than in hispanic videos lol This is one of the most practical examples I´ve seen and the way you explain it´s so clear. Thank you so much!

[@collegestatistics]

Wow, thank you!

[@jacobdeguire5754]

As someone coming from taking multiple statistics classes in undergrad and now having to re-learn... Wish you were teaching those classes! Your passion shows. If not for statistics, teaching. Thank you!

[@collegestatistics]

Wow, thank you!

[@candaceleigh20]

Thank you! I have taken multiple stats classes at the college and grad school level, but being out in the work force for 5+ years, I had gotten rusty. I recently switched to a more academically focused position & really appreciate the refreshers.

[@collegestatistics]

Great job! Thank you for that awesome comment!

[@jenny_gil]

Thanks for uploading this video! It's really helped me navigate statistical analyses and being able to interpret findings. Even 4 years later, I appreciate it!

[@collegestatistics]

You're very welcome!

[@TheJProducti0ns]

There needs to be more statistics videos like this. Less calculations and more understanding the concept. I just hate being thrown formulas. This is good stuff. Thank you sir!

[@collegestatistics]

The funny thing is that for real stats, hand calculations are almost impossible. Tell your biostats prof you want to see her run through a logistic regression by hand to see if she really understands it. It'll never happen!

[@patalinisalka8189]

Thank you soo much sir. I was really struggling with this topic in my epidemiology class. fewer calculations and more explaining are exactly what I needed. thank you.

[@collegestatistics]

You are most welcome

[@aaronmoonga3669]

the best tutorial I have ever found on odds ratio and interpretation. thank you so much for the tutorial

[@collegestatistics]

You're very welcome!

[@joykiden8514]

I got this just right on time when I'm struggling to understand OR and CI interpretations, I love your explanation , thank you very much. I'm a medical research student, watching from Kenya. Kindly recommend other similar videos to watch, regards.

[@collegestatistics]

Happy to help. Good luck on your exams!

[@huwaayman8783]

Me too !

[@madolo0978]

I love a good KZbin video, one beaming with personality! Immediately going into my Favorites folder.

[@collegestatistics]

Thanks for the support!

[@tapiwamashingaidze9317]

this is very good. i have been suffering with my assignment but you have opened my eyes. keep up the good work. i really appreciate your well informative video

[@collegestatistics]

Happy to help!

[@BME_RX]

Very clear explanation!!! Please make more videos.

[@collegestatistics]

will do!

[@bibigulaltayeva8253]

Thank you so much! From all 10000 videos I found your explanation is a best one!

[@collegestatistics]

You're very welcome!

[@DarineAlem]

This helped me a lot in understanding the OR, thanks a lot for such a great explanation, I'm an African Student, watching this from Sudan.

[@collegestatistics]

Truly amazing that the Internet allowed me to share this material with somebody so many miles (or kilometers) away. I wish you success in all you do!

[@rajinraihananikislam5432]

That was very encouraging and easy-going. While working on a research paper, I wanted to use the logistic regression. This video has been very helpful.Thanks

[@collegestatistics]

Would a video about adjusting odds ratios using logistic regression be helpful? In very general conceptual terms?

[@rajinraihananikislam5432]

@@collegestatistics Yeah, it would be. I would be very grateful to you if you upload the video about adjusting odds ratios.

[@ramamudarres2705]

Awesome explanation! Finally, I can 'connect the dots'! Thank you

[@collegestatistics]

Glad it was helpful!

[@punyasloksenapaty4460]

The best on internet for me❤

[@collegestatistics]

Thanks for commenting! (It helps the channel.)

[@linwang6975]

Thanks a lot, you explain this concept so clear and neat! I love it. Could you go deeper? for example, which OR is clinically significant? and which one is statistically significant at the 0.05 level?

[@collegestatistics]

That could get pretty technical, but thank you so much for your suggestion. Most of the time when you look at confidence intervals for the odds ratio or relative risk, you're really only concerned with one direction but the confidence interval gives you a two tailed test. Your idea makes me wonder if I should teach how to turn two tailed confidence intervals into one tailed confidence intervals and maybe write a little program to do it, and put it in the description. That's a pretty interesting video idea but would take me a little bit of time. Is that what you are getting at?

[@bob_the_bomb4508]

@@collegestatistics for me this question is, for example, “what is the difference between an odds ratio of - say - 0.2 and 0.8?

[@mustafeibrahim-xx1fk]

​@@bob_the_bomb4508 It is impossible to see a odds ratios which takes negative sign.

[@bob_the_bomb4508]

@@mustafeibrahim-xx1fkthat’s not a negative sign. It’s a parenthesis around ‘say’

[@AbiezerD]

Absolutely best explanation of OR ever !! Thank you

[@collegestatistics]

Glad you think so!

[@ashirar410]

Excellent video, very easy to understand, Thank you

[@collegestatistics]

Glad it was helpful!

[@cooljoe355]

非常感謝，我們是北京市中心醫院的一群醫生，我們感謝你們製作了精美的數學視頻，它幫助我們通過了執業醫師考試，並繼續我們美好的職業生涯，我們北京市中心醫院的醫生們，感謝你們的稱讚 擔任統計學家

[@collegestatistics]

I hope you become great doctors and help the quality of the lives of many!

[@kaoulkae]

Thank you very much. I am just curious about science and now I can understand the research papers results better !

[@collegestatistics]

Glad it was helpful!

[@aliyeaslsonsuz7099]

It is clear and short explanation. Thank you so much. I understand by all your 4 example so easily :)

[@collegestatistics]

Glad it was helpful!

[@bezawitashine2217]

Wow very much appriciated. What a great way of explaning this topic. Thank you

[@collegestatistics]

You are very welcome

[@christianaedwards3362]

Well Done, this video has helped with my exams questions . Bunch of thanks

[@collegestatistics]

Glad to hear that

[@Victoria-vg4sg]

Thank you so much for your video. The best odds ratio video ever!

[@collegestatistics]

Glad you enjoyed it!

[@pooi5647]

I have enjoyed the video. Thanks!

[@collegestatistics]

Thank you too!

[@rumeysaeed1]

Ur the best teacher sir. After studying two years of stats, I now know that I knew nothing about this. 😂 thank you!! And I just subscribed +1

[@collegestatistics]

Thank you so much for the compliment. Is there any topic you would like next?

[@rumeysaeed1]

@@collegestatistics I would appreciate if you could show us how to calculate p value and what is the significance of it in regards to the hypotheses. Thanks sir! I like how you actually show us the real examples and not just theoretically show what stats means. And due to the fact that many students havent been shown any examples is the reason why they actually fail to understand the topic and relevance of the stats when it comes to the researches. I think whatever topic you post up here, at least with some time many students will know the channel as you simplify the concepts!

[@rachael4868]

Really great explanations here. I can't thank you enough!

[@collegestatistics]

You're very welcome!

[@cha1645]

Impressive. Good explanation. You just fully explained a statistics question in MSc selection exam in my country (#SriLanka). Thanks a lot. I was clueless. Now I have understood the concept well enough. Tc #OddsRati and #ConfidenceInterval

[@collegestatistics]

You are most welcome!

[@carlosort-ceb2101]

Very nice. The same rule apply to relative risk for confidence interval?

[@collegestatistics]

Yes, exactly.

[@ayakhanji4120]

Thank you very much, I was struggling until I found your video. keep it up =D

[@collegestatistics]

Glad I could help!

[@ekit1686]

The way u deliver is amazing ..keep it up!

[@collegestatistics]

Thanks a lot 😊

[@flexonngulinga3038]

this is a great video. it has helped me in my Epidemiology assignment.

[@collegestatistics]

Thanks so much for letting me know. Most of my other videos get very few views.

[@graindust9109]

Omg thankkk u so much whoever u are. U r really helping me sir :") I'll subscribe ur channel

[@collegestatistics]

Thanks and welcome

[@faialnashmi4710]

This is more than helpful. Thank you so much for sharing:)

[@collegestatistics]

Glad it was helpful!

[@pendapalanamundjebo5328]

This was extremely helpful. Thank you.

[@collegestatistics]

You're very welcome!

[@terrys8206]

This is an amazing video! Thank you for your explanation :)

[@collegestatistics]

So glad you liked it! And I'm really glad people are starting to watch my channel. I don't have many viewers yet. If there's anything you want to video on feel free to ask.

[@faeeza3091]

Thank you so much sir, i'm doing my thesis right now and it really helpful!

[@collegestatistics]

Great!

[@manpreetkj3]

This is such a helpful video. Thank you!

[@collegestatistics]

You're so welcome!

[@ogedegbeidemudia9419]

This is quite insightful. Many thanks

[@collegestatistics]

Glad you enjoyed it!

[@urpriceless]

Best explanation!!!

[@collegestatistics]

Glad you liked it

[@agathaloveschrist]

Thank you so much. You are the best!

[@collegestatistics]

You're welcome! And thanks for watching.

[@aidylmorningdew195]

Thank you. you are a great teacher in stats

[@collegestatistics]

I appreciate that!

[@Assassin-pg5gb]

I could actually kiss you the interpretation has had me so confused BUT NOW IT MAKEAS SENSE THANKS!!!!!!

[@collegestatistics]

So happy to help!

[@MedicalBlackBook]

A very nice explanation, I finally understood

[@collegestatistics]

Glad it helped!

[@isabelladicorrado6153]

keep working hard professor! see you in class

[@collegestatistics]

Thanks!

[@nehacharak6190]

thanks a ton, its an amazing explanation!

[@collegestatistics]

Glad it was helpful!

[@insideexercise]

Great explanation but id be interested to hear of diseases that are increased by exercise! You need to watch my Inside exercise podcast!!

[@bob_the_bomb4508]

Take up playing squash when you have an undiagnosed heart condition?

[@mariosgrigoriou7596]

simple and clear, thank you.

[@collegestatistics]

You're welcome!

[@vicmamawow6176]

Awesome! This video is really helpful to me.

[@collegestatistics]

Thank you for watching and glad it was helpful.

[@hirjani1389]

Is there a video that explains interpretation of OR? This one is really helpful :)

[@EzrhaGodilano]

This is very useful. Thank you!

[@collegestatistics]

Glad it was helpful!

[@wouter55555]

this was actually amazing!

[@collegestatistics]

Thanks for watching!

[@suvanibaral6780]

The best explanation ever. Can u please explain how did you get that CI . Also can u make video on p values

[@collegestatistics]

I have a lot of P value videos in my hypothesis testing playlist.

[@Amir_LNH]

Very much impressive

[@collegestatistics]

Hope I helped.

[@walaaalnabhan8341]

Outstanding explanation, thank you a lot :)

[@collegestatistics]

Glad it was helpful!

[@amirpri-gal6970]

good and clear explanation, thank you friend

[@collegestatistics]

You are welcome

[@SS-fd8kh]

Very useful. Thank you a lot!

[@collegestatistics]

Glad it was helpful!

[@debrajanetmalagaespichan1952]

Thank you for your explanation!

[@collegestatistics]

Glad it was helpful!

[@Zalmitechsolutions]

It was a clear explanation thanks for the video.

[@collegestatistics]

You're welcome!

[@JAMScover1]

This is so helpful! Thanks

[@collegestatistics]

You're so welcome!

[@jwilson3663]

Thank you for this!!

[@collegestatistics]

No problem 😊

[@ashleyroughton6814]

Very helpful. Ty.

[@collegestatistics]

You're welcome!

[@tatianathomas2572]

Great video! Helped a lot.

[@collegestatistics]

Glad you liked it!

[@marieofstad6984]

Thank you for this video!! I couldn't for the life of me understand why the 1 in the CI would equal to it being not statistically significant, but now I get it!😊 However, I still struggle to understand where the CI is coming from... Is it a new CI we need to calculate specifically for the OR, and that the formula for this is different to the CI we calculate for the mean?

[@collegestatistics]

Thank you so much for watching my channel. I stopped making videos because it didn't seem like many people were watching but this particular video does have a lot of views. Anyway, to answer your question...You might be able to think about it this way. Look up the formula for the confidence interval for a population mean and you'll find different formulas: one for when the standard deviation is unknown and another for when the standard deviation is known. Then look up the formula for the confidence interval for a difference of two means. You'll find a similar thing… different formulas. The confidence interval for the odds ratio requires the standard error for the odds ratio estimator. And this standard error is pretty darn complicated and requires a computer algorithm! You can estimate it by hand on a logarithmic scale but that's only an estimate. But the short answer is yes, it's different.

[@marieofstad6984]

I'm very grateful this video exist! I had spent hours trying to understand this, rereading my textbook, online resources and trying to google my way to an answer... Non of them made any sense to me! But your video is gold, so well explained, and now I'm sitting here wondering how I didn't understand it right away! Haha Ooh okay, it all makes sense then! Thank you so much, I really appreciate it!!😊 I think this is the end of statistics for me for this semester (we only have to learn the very basics to be able to understand and interpret the results in scientific publications!), but if I ever have to learn more statistics, I'll definitely check if you have a video on it!

[@foreer4627]

If we have 1 in the upper limit in CI, are the findings significant?

[@collegestatistics]

I would need a bit more detail in the question. Maybe put some numbers in it?

[@foreverarmy_ot7]

thank you. excellent explaining

[@collegestatistics]

Glad you liked it

[@EmmanuelKyaloMuli]

from Kenya. Great video!!

[@collegestatistics]

Thanks! 😃

[@EmmanuelKyaloMuli]

@@collegestatistics Please help me understand how to interpret the same odds ratio in the video if the confidence interval width exceeds 5%.

[@captainmustard1]

very intersting video, i'm so thankful. But i have question, sometimes, i find a very huge number in upper CI, like 1561646. What does it mean? Is this is because of covariance?

[@collegestatistics]

I'd need to see a screenshot.

[@xoxo_1990]

I liked the video. one suggestion, please use words like strong or weak evidence as these are used in exams.

[@collegestatistics]

It's unfortunate that some instructors want evidence classified as weak or strong because confidence intervals allow evidence to be classified as non-significant or significant, and that is more rigorous and validly interpretable. But thank you so much for watching and for the comment.

[@whitneepe3890]

very clear explanation!

[@collegestatistics]

Glad to hear that! Many thanks!

[@maharudrahrishikesh3881]

That was a great video. Thanks a lot. Can you upload a video on how to calculate the 95% CI, please?

[@collegestatistics]

Great suggestion! Working on that tonight!

[@yaplishan1996]

Hello! Thank you for the great video :) Could I ask a question on how to write the interpretation in a correct format? (I am concerned about my statement not being accurate for use in exams as I am required to critically appraise statistic data in university - for example your video mentioned not to describe relationship as causation but to the data as association) For example: Vitamin D with OR = 0.71 and 95% CI is 0.68 - 0.84 Does that indicate that: vitamin D is a protective factor and it significantly decreases disease risk by 29%? Thank you for your help!

[@collegestatistics]

Yes, and if you wanna make sure you get full credit from your professor call it a “protective association” this way your professor knows that you are not trying to make a causal inference from a retrospective study. And you could put the word association in all capital letters or underline it or something like that. One of the biggest parts of introductory college statistics is to make sure the student knows the difference between associations and causations. Hope this helps. And thank you for watching my video!

[@sarawrsflove]

@@collegestatistics I agree this video was extremely helpful. For me it helped refresh how to interpret the significance. I did have one question. As a pharmacy student who is often reviewing literature from randomized control trials. Often times odds ratios will be used as a result of logistic regression. Is the interpretation the same, and would causation be an appropriate conclusion given the study design?

[@farihazafar4806]

Your teaching style is too good

[@collegestatistics]

Glad to hear that

[@ibassik]

Great video!

[@collegestatistics]

Glad you enjoyed it

[@afsanajabeen608]

thank you so much for this informative vedio

[@collegestatistics]

So nice of you

[@rositamansouri2735]

Thank you

[@hopaeds2183]

very useful video. thank you

[@collegestatistics]

You are welcome. Thanks for watching.

[@faridabashir4678]

Thank you This is very helpful

[@collegestatistics]

Glad it was helpful!

[@sumangurung6935]

Thank you, largely helpful

[@mihretabreham1055]

Thanks aloooot!keep it up!

[@collegestatistics]

Thank you too

[@reem4374]

This is AMAZING thank you What if my study is cohort; will using OR and CI appropriate or not?

[@collegestatistics]

On the medical boards you’d want to use relative risk for prospective studies. But both can technically be used for prospective studies. But relative risk cannot be used in retrospective studies.

[@zh6804]

Thanks a lot 🙏🌷🌷🌷

[@collegestatistics]

Most welcome

[@shreyaannagarajan2003]

What do you mean by odds ratio between 0.6and 1.1 were all compatible 1.0 ❓ I don't understand the 'compatible with one' part Thank you

[@collegestatistics]

That interval captures the number one.

[@felicialalremruati8478]

So helpful for writing my review article😁

[@collegestatistics]

Fantastic!

[@user-es6kn6md2b]

Thanks

[@collegestatistics]

Your Welcome!

[@mahenur]

Very helpful thank-you

[@collegestatistics]

Glad it was helpful!

[@longz2419]

Nice

[@collegestatistics]

Thanks!

[@minuteswithfrancie9742]

Thank you so much sir

[@meza4593]

you are amazing

[@collegestatistics]

thank you very much!

[@RaanitaOfficial]

This is so helpful

[@collegestatistics]

thanks