T-Tests: A Matched Pair Made in Heaven: Crash Course Statistics #27
Вставка
- Опубліковано 14 сер 2018
- Today we're going to walk through a couple of statistical approaches to answer the question: "is coffee from the local cafe, Caf-fiend, better than that other cafe, The Blend Den?" We'll build a two sample t-test which will tell us how many standard errors away from the mean our observed difference is in our tasting experiment, and then we'll introduce a matched pair t-tests which allow us to remove variation in the experiment. All of these approaches rely on the test statistic framework we introduced last episode.
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, Erika & Alexa Saur, 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, D.A. Noe, Shawn Arnold, Ruth Perez, Malcolm Callis, Ken Penttinen, Advait Shinde, William McGraw, Andrei Krishkevich, Rachel Bright, Mayumi Maeda, Kathy & Tim Philip, Jirat, Eric Kitchen, Ian Dundore, Chris Peters
--
Want to find Crash Course elsewhere on the internet?
Facebook - / youtubecrashcourse
Twitter - / thecrashcourse
Tumblr - / thecrashcourse
Support Crash Course on Patreon: / crashcourse
CC Kids: / crashcoursekids
Thank you for sticking with a single example through the entire length of the lesson.
Hi. I remembered Crash Course from AP classes in high school. Here I am now, as a graduate student, still watching Crash Course. Thank you for your videos. They really help!
Alex throwing the coffee cup at 6:15 cracked me up- the animation is always so perfect (especially with how subtle it can be) :D
6:14 Missed an opportunity to say "coffee is not really her cup of tea" 😄
Bold of you to assume that I have friends let alone 16 of them
This course is an absolute lifesaver. I finally understand what is going on in my statistics class. :)
Can statistics be fun? You make it so. Thanks for sharing!
Teresa White nay not really
This video was colourfully boring
8:43
"T-statistics tell us how many standard errors away from the mean our observed difference is"
i literally use this as a guide to fully understand statistics inside and out. i go back and rewatch a video if i am ever having trouble. thank you!
Your explanations are amazing! Thanks a lot for the videos! You're saving my MSc 😍😅
Ugh I did this stuff last semester at University in my biology class! Took me ages to wrap my head around this stuff! As usually, you guys explain it so well! I wish this were released then 😂
I did not know there was a crash course related to statistics. This is so cool! Thanks!
For some reason I really want to drink some coffee right now
"some reason"
I scored T higher by 8 points.
Thank you so so much :))) This is a great video. I stuck with the t test for hours but now I got it
Great episode, made some of the concepts from previous ones finally click in
10:05 haha that laugh! So charismatic
I have a final tomorrow morning, this is a blessing
Explained better than it was in my Psychology class :/
same
Question: Calculating the 2-group standard error gives different results from calculating the 1-group standard error of the differences. Which one is used when? Is the 2-group one used when they're independent?
I see the Cafe Du Monde coffee with the chicory made in it on the desk. YUMMMMMMM Who knew about the coffee.
The table at 1:28 is missing a row. 16 friends mentioned and the denominators 8 at 2:58 are the evidence.
Adrean and team many thanks for the videos it is great fun
3:11
There are two ways to determine if there is a statistically significant difference between the two coffee shops:
1. We can calculate the critical t-value and if our t-statistic is greater than the CRITICAL VALUE we reject the null hypothesis
2. We can calculate the p-value from our t-statistic and if the p-value is smaller than our chosen ALPHA LEVEL we can reject the null hypothesis
please cover the Monte Carlo method and algorithm in future episodes :)
New in a new town and already with 16 new friends... interesting xD
Thanks
Crash Course is the best
Holy frick my mind just can’t comprehend the formulas for some reason they look like mush
T-Test for Coffee :D
Tim Hortons in the background!
Spoon is such a great band.
Interesting
Anyone know where she got the critical values from? I looked at the t-table for two-tailed distribution, and the values she is using do not seem to match up to a sample size of 8 and an alpha level of 0.05. Help?
The Sugarcubes sound like they belong in coffee
When I put the T-statistic into the Excel formula =NORM.S.DIST(-.321167,1) it spits out a Pvalue of 0.0006598, different from 0.005823 seen at 9:22. Is this because the probability curve in this small-sample coffee shop example is ... uh ... "shorter and fatter"(?) than a completely normal, large-sample bell curve?
Also, how would I look up the critical value for this example in a T-distribution table? What "degree of freedom" row would it be under?
I've actually heard convincing arguments saying that absence of evidence IS evidence of absence. I think the phrase would've been more accurate if it said "absence of evidence is not proof of absence," but then it wouldn't be as catchy, I guess.
It's not proof. I just said that. It WOULD be evidence that there are no black swans, which is an appropriate conclusion until a point in time when black swans ARE observed.
Also not everyone will witness everything that we as a collective species knows to be true or to exist. We have a community of people doing research, making studies, taking notes, doing observations, and sharing their data with others. The most useful of these are reproducible experiments or studies. When multiple different groups are observing the same thing, we are more confident in those results. We never exactly reach true certainty, just theorems that have yet to be proven incorrect.
So, when one group or person observes a black swan, that's evidence that they exist. When multiple different people or groups have observed black swans, the evidence becomes more and more convincing that there are, indeed, black swans. Even those of us who don't directly observe them can trust the collective society of scientists (for example) who have all agreed on this conclusion.
I've never been to China, so for all "I" know, it doesn't exist, but tons of people say it's there and say they've even been there or came from there, and I've seen photos and videos and satellite footage showing it's existence, so I have tons of evidence that it really does exist.
If people are looking for something and not observing it in any capacity, that absence of evidence of existence is absolutely evidence of absence of existence. We have no reason to believe something exists that has not been observed. People can claim whatever they want, but until others can observe said claims in controlled experiments, and those results can be reproduced, it cannot be substantiated.
Why are you basically repeating what I'm saying and then claiming I'm wrong?
Dan Scannell I wasn't very nice... but I still don't agree
Can I use pair T Test to evaluate the differences in attitudes before and after an intervention of the same class. The problem is that I did not ask them to write their names or the identifying sign
Final is tomorrow 🤦♂️🤦♂️
Jack Rea I got 1 hour left trying to get that 5
same tho f to pay respects to our grades
The Lovin' Spoonful?
What do you think we are going to cover next?
My first day of my statistics course in college the professor proved through statistics that all firearms should be outlawed. The second day of the course our professor proved through statistics that states that have shall issue gun permits are the states with the lowest amount of crimes
Aren't firearms outlawed unless you have a gun permit? I get the point, but maybe he was messing with you?
AngelHQ Short answer .... NO. Many states don't want you to carry them around on your person without a permit but to own one you just need to pass a background check. Then there are states like New Jersey and cities like Chicago .... very restrictive. OTOH, Vermont and Utah have almost no restrictions.
Basically you need to know your local laws and be careful when you travel.
grazie raga
I seem to remember from statistics class that the standard deviation of a small sample size (less than 25) was meaningless. Am I wrong?
Needs a cat cafe named cat-ffienated.
5:18 do you mean difference of 0.3 or SMALLER?
Hi, will you please add Turkish subtitles to all videos?
Why do you have 10 digits calculated in your standard deviation? You don't have enough data for the third digit to matter and the means only differ by the second digit anyway.
10:09
With coffee that similar, you might as well focus on price or service instead
Aaaaah but i have to calculate the p value with a table :s
I don't understand everything. Help
Much more confusing than their other videos on stats. Maybe I was just bored.
Caffeinated coffee scoring lower! I'd like to see the T test on that one.
Why is it that I can find all of that information for free but need to pay 9 grand per year to have someone teach me the exact same thing in a slightly less entertaining way?
for the first time I think I understand.. more coffee examples please
8:43 how many std err away from the mean the T dist is = T score
"standard deviation, or SD, measures the amount of variability or dispersion for a subject set of data from the mean, while the standard error of the mean, or SEM, measures how far the sample mean of the data is likely to be from the true population mean. The SEM is always smaller than the SD."
8:26 how did she get the average variation?
I'm wondering this too...
Actually I've just calculated a number very close to it... it takes ages and I don't want to double check it, but basically:
Step 1: subtract the average difference from each difference value
Step 2: square the answers to get rid of any negative signs
Step 3: sum the answers
Step 4: calculate variance by dividing by n-1 (15)
Step 5: Sqrt the variance to get the standard deviation
That's the .22....
I like ur white shirt AH
Over... caffeinate? I'm sorry, I don't understand this term. Why not accuse us of giving our friends too much fresh air to breathe?
Yeah... they really lost me at that metaphor. Over-caffeinate? How does that even work? What does it mean?
Yal the best
6 significant digit figures ? really ?
spoon
heh
I don't understand ANY of this lol
Bakfast!
First hehe
Couldn’t follow it 👎🏽👎🏽👎🏽
👎👎👎
You failed to capture my interest with your metaphor
when it comes to Coffee I would go much deeper in a measuring, I would learn where they're giving their beans from and the method they use to brew it. There are so many factors when it comes to coffee, that metaphor does not work