Interpreting Confidence Intervals EXPLAINED in 3 Minutes with Examples

Learn how to interpret confidence intervals and why it's often so confusing to do so. This statistics tutorial uses an example to explain the biggest mistake students make when concluding on confidence intervals. Then we discuss the correct confidence interval interpretation and touch on the difference between probability and confidence. Never make the wrong confidence interval conclusion using tricks from this video!

[gamingSlasher]

Super good. Finally I got it. It is surprising that statisticians have such a hard time explaining these concepts so "normal" people (engineer) can understand them.

[RubinhoX]

Thank you so much! This explanation was immensely helpful

[VenkataSaiMounishEnnamuri]

please do post more of these interpretation videos. they really are helpful

[mosama]

Man you are a LEGEND.
I really wish you have more videos for statistics. Thank you! 😊😊

[dovidasm]

Thanks for sharing!
Although the 2nd interpretation is the right one, I find it really akward saying kind of: "From the sample we calculated the range 7.5-8.5 hours; we are likely to have the true mean in there; I cannot tell you the probability, but ranges obtained this way have 95% chance of having the true mean inside". So, yeah, I totally agree with the 3rd version saying: we are 95% CONFIDENT (warning key word) that the true mean falls between 7.5-8.5 hours.
However, I think there's a 4th version which might be useful: "Any mean between 7.5 and 8.5 is "compatible" with the observed average at a confidence level of 95%, i.e., the 95% probability interval for any mean between 7.5 and 8.5 contains the observed average". Hard to put it into words, what I mean: [7.5-8.5] = {μ : | observed_average - μ | ≤ 1.96 σ/√n}. μ is fixed and unknown, it's clear we have no randomness in this set, we cannot therefore calculate a probability, but there's still this 95% probability issue in its construction. So, it might make sense to refer to some kind of pseudoprobability for this set, let's call it ""CONFIDENCE"" 😛

[Vladyslav_Rehan]

Oh, you have such high-quality videos, it's a pity that there are so few subscribers ...

[a_programmer]

Bro, continue please, Your videos are so helpful and clear. THANKS!

[anthonycarbonaro]

Your presentations are excellent. Thank you 🙏. Looking forward to seeing more from you.

[murielsamuels]

Thank you. I am telling everyone about you. Thank you.

[gooddeedsleadto]

Simple explanation. Thankful
Could u also solve few problems to see how it is used and also using p instead of mu?
Could u explain how to use it to predict election results through confidence intervals?

[mainakdey]

this is an awesome explanation. thanks sir 👍👍

[medhatkmeel]

Really, I am ok with the interpretation, BUT why in the fig. shown (1:42) there are some confidence intervals of different length? >>> How the same formula (for ex.: x-bar +/- critical val. * SE) would generate intervals varying in width in the presence of only one x-bar (the sample mean as our point estimate) and only one critical value (reliability factor) that determined by the confidence level we choose and lastly one SE the depends upon the std and the sample size?

[lipe]

But are they 'approximately the same thing, ' or is the problem just choosing the right words? By the way, thanks for the video; it helped a lot!

[arthurrichard]

Please solve this example questions in your videos.

[seliassiva]

This feels more like arguing about semantics more like. Maybe we should ask the question, how wrong is the wrong answer? Does it actually affecting the final result by a significant amount or can it potentially lead to false calculations?

[nature_through_my_lens]

Can you do a video on Maximum Likelihood Estimation?

[quantstyle]

That second interpretation "If we repeatedly found confidence intervals..." is not an interpretation of 7.5 and 8.5 hours. It's a general definition.

[leomorlacos]

Thank you so much! Why no patreon account?

[maxywaxymus]

Then what do statisticians mean by the word 'confident'? What is making them 'confident'?

[spyhunter]

You haven't mentioned about how to find the confidence interval. As far as I am aware, it's the value we choose to decide, right?