Confidence Intervals: Crash Course Statistics #20
Vložit
- čas přidán 12. 06. 2018
- Today we’re going to talk about confidence intervals. Confidence intervals allow us to quantify our uncertainty, by allowing us to define a range of values for our predictions and assigning a likelihood that something falls within that range. And confidence intervals come up a lot like when you get delivery windows for packages, during elections when pollsters cite margin of errors, and we use them instinctively in everyday decisions. But confidence intervals also demonstrate the tradeoff of accuracy for precision - the greater our confidence, usually the less useful our range.
Crash Course is on Patreon! You can support us directly by signing up at / crashcourse
Thanks to the following Patrons for their generous monthly contributions that help keep Crash Course free for everyone forever:
Mark Brouwer, Glenn Elliott, Justin Zingsheim, Jessica Wode, Eric Prestemon, Kathrin Benoit, Tom Trval, Jason Saslow, Nathan Taylor, Divonne Holmes à Court, Brian Thomas Gossett, Khaled El Shalakany, Indika Siriwardena, SR Foxley, Sam Ferguson, Yasenia Cruz, Eric Koslow, Caleb Weeks, Tim Curwick, Evren Türkmenoğlu, D.A. Noe, Shawn Arnold, Ruth Perez, Malcolm Callis, Ken Penttinen, Advait Shinde, Cody Carpenter, Annamaria Herrera, William McGraw, Bader AlGhamdi, Vaso, Melissa Briski, Joey Quek, Andrei Krishkevich, Rachel Bright, Alex S, Mayumi Maeda, Kathy & Tim Philip, Montather, Jirat, Eric Kitchen, Moritz Schmidt, Ian Dundore, Chris Peters, Sandra Aft
--
Want to find Crash Course elsewhere on the internet?
Facebook - / youtubecrashcourse
Twitter - / thecrashcourse
Tumblr - / thecrashcourse
Support Crash Course on Patreon: / crashcourse
CC Kids: / crashcoursekids
I'm 95% confident that students like us cannot retain information verbally
I have my statistics exam in 2 days and believe me, I'm not 95% confident of succeeding at all
was this hypothesis accepted or rejected?
just took the semester final
lmao same
mine is in two hours
I have an exam in 6 hours and I don’t know what I’m doing
i am 95% more confused than I was before
Crash course should have the math written down or drawn out instead of just telling us the process of the math. It's hard for me to imagine the math in my head. The graphs help though!
Nasia furthermore, she talks way too fast to adequately absorb the information.
I usually cross reference with Khan Academy when I need to see the calculations.
yess
started off okay and just devolved into noise where i would occasionally hear 'distribution'
im 95% sure I'm gonna fail my exam tomorrow ahahaha kms
4:10 calculate 95% CI
7:00 t distribution
to calculate t distribution, use program
10:40 margin of error
ah stats... always very intuitive at the beginning. and then becomes less and less intuitive. it's just new terms and formula at some point. there is no "big picture" to help you find where you are or why you are there. anyways, great explanation as always.
statistics lesson on crash course should be more like written rather than oral explanation.
I am 95% confident there won't be gold lame pants next week.
Confidence and Stastics are words that people usually wouldnt put in the same sentence
"I have no confidence about thees statistics"
Who here is cramming for tomorrow??
HOLYSHIT ME
Cramming for tonight!!
I’m here for fun and ‘background learning’ et. ‘playing games while absorbing knowledge’
i have a stats exam in 3 days and you taught me more in one video than my professor has taught me in two weeks. He didn’t even teach us about confidence intervals but put it in our homework. So i thank you so so much for this, but i still think i’m gonna fail my exam.
Should have added more graphics to this episode. It's hard to follow along
Thank you so very much !!!! super awesome way of breaking confidence intervals in easy to understand way !!!
glad I'm not the only one more confused after watching this video than before. stats its the only time where crash course has failed me.
oh i dunno, crash course physics is pretty bad too. barely made it out of kinematics and they toss you into calculus
i understood every lect in this series before this lect.When the t- dist started, it got too complex.
Confidence interval formula is : the mean, plus or minus (standard deviation / sqrt(sample size)) * the Z statistic.
The Z statistic in this example 1.96 because of the desired 95% confidence interval. If it was exactly 2 then your confidence would be 95.45%. If it were 3 you'd have 99.73% confidence. So basically, the wider the range (or the greater your sample size), the great your confidence. The narrower your range (or the smaller your sample size), the less confidence you have in the accuracy of the mean.
Keep it up! If I remember accurately there is not much left for this section of the course :D
Thank you thank you thank you! This helps a lot!
In a paper almost certainly prepared for a statistics class, someone found the mean length of 30 randomly selected top 100 songs from iTunes was 226.93 seconds. Seventy songs of that length will last you through a 4 hour 24 minute marathon.
I’d have to buy the whole music store to finish my marathon. Which consequently would make me sure to finish it. Nice trick!
I've got every Grateful Dead studio album on my phone, so there's no way I'll be finishing a marathon.
The fact there is no examples that explain exactly how to find a confidence interval is a crash course is crazy. Just explaining terminology isn't enough SHOW US HOW.
when i listen to their intro my mind screams (HANK) lol
After watching this video I came to a conclusion........ WTF IS A CONFIDENCE INTERVAL???
"Imagine you're going to run a marathon. And you don't know how long it will take..." lol :D Yes, you do. :D
Good job!!!
Great video!
Congrats on 7.8m subs
Gawd this is so much better than my college prof was
Had to look away from dentist scenes
Thank you crash crouse
Ha, "buying" music…yeah…i totally do that…
I agree with everyone's complaints that they don't show the math - but does anyone know a good alternative that does show it step by step? Would love to know!
Thanks Adriene :D
According to my confidence, the statistical quantification of my interval is more than 200% of 95 that I fail this final.
I'm pretty sure that, when you don't know the population standard deviation, then you should always use a t-distribution, even with a large sample size.
took me until just now to pinpoint the Intense Gina Linetti energy going on (and I only noticed because i started rewatching b99 to counteract the deep sadness stats is bringing me). more engaging than my college textbook though! also the number of people using the 95% c.i. in the comments but like, wrong.... concerned
Thaaat’s what she reminded me of!!! Lol does anyone else think she gives of Gina Linetti vibes? Great teacher. ^^
lol a little bit. like Gina's calmer, saner older sister
I cannot get over your decor
Am I the only one upset about not seeing those gold pants?
The sound in the background that's there during the text "Our conficdence is in the fact that the procedure of calculating this confidence.." at 3:10 is REALLY annoying. Also it would be great with some formulas, I'm sure at least 80% of your audience are people studying for exams.
the sound effects are really distracting :/
I have 95% confidence I will pass or not pass tomorrow because my exam isn't tomorrow.
There should be a link in the description to the topic play list
could you please explain the BLP model in the next video?... that would be very usefull for me..
One thing that's really confusing: given a sample and then calculating a 95% confidence interval for that sample, you cannot say that there is a 95% chance that the population mean is within that confidence interval. This is because every sample has other features besides mean and standard deviation that will determine how likely it is that the mean will fall within that confidence interval. If we just use the mean and standard deviation, the 95% confidence interval is the best guess we can make for where the mean will fall 95% of the time, and indeed for 95% of samples, the mean will fall within our confidence interval. However, if we have the most advanced statistical modelling tools known to man and other features of our sample besides the mean and standard deviation, we would know the exact probability that the population mean is within our confidence interval, which is likely not exactly 95%
I have never found an explanation of this that isn't highly confusing and vehemently disputed.
Any recommended resources that can help me make sense of it?
I was pretty confident with this topic but now I am so confused! She is going really fast and instead of showing all the dentistry stuff and some more unnecessary scenes, it would've been better if there were actual words and a graph on a part of the screen throughout that example, e.g., "mean number of cavities the avg person has in 5 yrs > 1. Calculate 95% confidence interval." or smth like that.
So is the only difference between T distribution and Z distribution that T distribution is used for smaller sampling sizes that are not quite normal?
first crash course topic I've ever watched that really isn't doing its job. I love crash course generally, but this is wayyyyyy to fast :(
Well, now you have to wear gold pants next week...
I'm 95% that there'll be more than 3 fillings needed,
so confusing in words needs to show how the numbers actually work
ie. math
The delivery guy will come by on the statistical 11ish. My impression is that computers will fight the indefinite by shooting the delivery guy, due to conservation of time and electricity.
Neat
What the heck was that boom sound at 3:42?!
Did anyone else notice the sound at 3:42?
Gummy worms are ok, I guess, but you could talk about student's actual work on delicious beer.
I want to ask whether the 1 and 2 sigmas in normal distribution the same as confidence interval?
I also think that makes sense but calculating the CI that way was never mentioned. I THINK it's because 2 sigma just represents 95% of all scores, whereas the CI is used to describe the MEAN (population mean) within a range.
So people who buy songs really exist , interesting.
gotta get that crispy 320kbps
4:20 what table??
If we're already saying that the mean of our sampling distribution will be approximately equal to the population mean, then why are we taking a 95% confidence interval to prove that the population mean will be within a particular range? except that we're trying to give an interval estimation......
01:15 its center, not it's center
she's a bit drier than hank or john but i'm definitely getting a better grasp on stat
I should have been more specific. It's the true believers in computers who will fight the indefinite to save time and or electricity, because that's where they came from in the first place.
But the p value will vary depending on whether you are using the .05 level (used here) or an .01 level which is more stringent.
the p-value does not change based on alpha level, but your cutoff would be different. 0.04 is considered "significant" if alpha = 0.05, but not if alpha = 0.01. However in both cases p = 0.04
Chelsea Parlett You are absolutely correct. Thank you very much for correcting my blunder. I must have been half asleep at the time.
9:05 - Where did the 0.033 came from?
I think... We have 50 repeated Bernoulli trials with p=0.06 (given) so the standard error is sqrt(0.06(1-0.06)/50) = 0.033
Yay statistics :):):)
Why times 1.96 around min 2?
Confidence interval for appearance of gold pants?
*_...is Student's T Distribution [_**_07:03_**_] the closest continuous function to the binomial, distribution (as in flipping a coin a small number of times), or how does it compare..._*
traducción al español por favorr
"Confidence interval quantify our uncertainty. They also demonstrates the trade off of accuracy for precision " crash course
This lady looks and speaks like steve from gamers nexus' mom.
I’ve always thought that if a CI contained 0, it’s not significant? Am I wrong here?
CI's aren't usually considered "significant" or not, but if a CI contains 0, and your null is that the parameter or difference = 0, then yes, a CI containing 0 means that you'd also fail to reject the null.
Chelsea Parlett yes, I should have stated more clearly. The CI itself isn’t considered significant, but the accompanying p-value- when containing 0- will not be significant.
the accuracy and precision statement is swapped, it need revison
I guess i will be leaving cos of the comments😂
If we calculated 5 CI and it turn out that all CI didn't contain true population, is it considered as bad luck? Lol
But seriously is it possible? .-.
ahmad fahadh ilyas it is possible, though unlikely!
Why would you buy new music just to load up a running playlist? Just put a suitably-long selection of your existing music on there.
Pfhorrest or you could just buy the 70 songs and put them on shuffle and repeat
Where was this when I was in 9th grade 😂🤔
Put the speed to .75 shes finally talking normal
Are you kidding? She sounds stoned at 0.75. x1.25 sounds normal to me.
Did you ever end up getting your pants-
So if you're really lucky, that chocolate bar contains infinite calories.
What...!
feel like i'm in a Casino with the stupid sound f/x
*_...95% has been around for decades, so, why, 95%-I'd use Systems Theory to spec. returns, calculate the time-and-cost of processing a rate of returns, a probability of overrunning time-and-cost allotted, (got to stay in business)..._*
I miss jhon and Hank :(
The forms of colorblindness I'm aware of are all genetic. I don't think we have a simple medication that could work like the one described.
Last comment
Incase you didn’t know, crash course, the AP test was weeks ago.
Lauren Divitto I look these videos now and like "nope, never gonna think of stats again😄" lol
I just dumped my name and address
I usually like ur videos but this one lacked examples with real numbers and graphs. Hard to follow
Barf
11th comment
“It’s center”?!?!
Please proofread your blurbs. I expected more from CrashCourse...
3rd comment
This was explained so bad
Not a good explanation
I'm sorry, but the crash course series just falls apart with mathematical subjects like statistics or physics. It's too fast, too abstract, too oral ... if you're not mathematically inclined you need to see someone do example problems over and over and then over again until you get it.
Hey Adrian don't record when you're sad, take a day off. You're clearly super sad.