Selection Bias: Do You Really See the Whole Picture?
ฝัง
- เผยแพร่เมื่อ 3 ก.ค. 2024
- Selection bias is a common problem that can occur when analyzing data that is not fully representative of the information intended to be studied. This can lead to inaccurate conclusions and decisions that are not based on the full picture. A classic example of selection bias is the story of the statistician Abraham Wald and the missing US Air Force planes.
SUPPORT us to learn more about human beings!
/ sprouts 🐦:
DOWNLOAD video without ads and background music 🤫:
sproutsschools.com/video-less...
SIGN UP to our mailing list and never miss a new video from us 🔔:
eepurl.com/dNU4BQ
SOURCES and teaching resources 🎓:
sproutsschools.com/selection-...
VISIT our website 🌐:
www.sproutsschools.com
CONTRIBUTE by upvoting your favorite topic or suggesting new ones☑️ :
sprouts.featureupvote.com/
THANKS to our patrons
This video was made with the support of our Patrons: Alex Rodriguez, Andrea Basillio Rava, Angela, ArkiTechy, Artur, azad bel, Badrah, Bernd Gaertner, Cedric.Wang, Daniele Diniz, David Markham, Delandric Webb, Digital INnov8ors, Dr. Matthias Müller-Mellin, Duane Bemister, Eva Marie Koblin, Fatenah G Issa, Floris Devreese, Frari63, Gerry Labelle, Ginger, Harmoniac Design, ICH KANN DEUTSCH UND ES WAR EINFACH!, Izzy, Jana Heinze, Jannes Kroon, Jeffrey Cassianna, Jim Pilgrim, Joanne Doyle, John Burghardt, Jonathan Schwarz, Jorge Luis Mejia Velazquez, jun omar ebdane, Khadijah Sellers, Leonel, Linus Linderoth, Lucia Simone Winston, Marcel, María, martin, Mathis Nu, Mezes.Macko, Michael Paradis, Mindozone, Natalie O’Brien, Nick Valerio, Nicki, Okan Elibol, Oweeda Newton, Peter Bishop, Povilas Ambrasas, Raymond Fujioka, Roel Vermeulen, Scott Gregory, scripz, Sebastian Huaytan Meder, Si, Stefan Gros, Stephen Clark, Stuart Bishop, Susan Schuster, Takashi HIROSE, Thomas Aschan, Tristan Scifo, Victor Paweletz, Yassine Hamza, Yvonne Clapham and all the others.Thank you! To join them visit www.patreon.com/sprouts
ORIGINAL TITLE
The Selection Bias
COLLABORATORS
Script: Jonas Koblin
Co-Author: Ludovico Saint Amour Di Chanaz
Artist: Pascal Gaggelli
Voice: Matt Abbott
Coloring: Nalin
Editing: Peera Lertsukittipongsa
Production: Selina Bador
Sound Design: Miguel Ojeda
SOUNDTRACKS
Magical Keys - Studio Le Bus
Midnight Mischief - Jonathan Boyle
Nice Toys - Studio Le Bus
DIG DEEPER with these top videos, games and resources:
Read about Simpson’s paradox
en.wikipedia.org/wiki/Simpson...
Read this article about the different possible types of selection bias
www.scribbr.com/research-bias....
Read about why a Hospital is the most dangerous place on earth.
link.springer.com/chapter/10....
Read this article on how to avoid selection bias in geriatric research.
pubmed.ncbi.nlm.nih.gov/31211...
SOURCES
en.wikipedia.org/wiki/Selecti...
en.wikipedia.org/wiki/Abraham...
www.cdc.gov/nchs/products/dat...
CLASSROOM ACTIVITY
Visit our website to see suggested activity for classroom on this topic.
CHAPTER
00:00 Introduction
00:20 Selection Bias
00:34 Abraham Wald's story
01:38 The missing planes
02:22 The simpson paradox
02:38 The real example
03:19 Conclusion
03:33 What do you think?
04:23 Patrons credits
05:29 Ending
#sproutsschools #selectionbias #bias #psychology
To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/sprouts/. The first 200 of you will get 20% off Brilliant’s annual premium subscription.
Plane
I don't even have to do the experiment to know the outcome. You might be familiar with your close neighbours, but the farther you go the more you become a "stranger". Since people who are afraid of strangers won't even open the door for you, you will report back that all (or most) people who answered to you said they are not afraid of strangers, when in fact there might be quite a few who are.
This is a bigger issue especially with social media apps, making it easier to put people in danger/manipulate them. Thanks for the informative video!
I love how each new video you release helps me realize how dumb i am afresh. And the way your animations and sound effects drive home your rather profound arguments both effectively and comically without being obstructive is just genius.
Hey, you have a learning mindset and it can never correspond to dumbness. Thank you for your sweet words :)
Another example are polls and surveys - we only see opinions of people who agreed to be surveyed, and the response rate can be as low as 5%. So we may be ignoring opinions of people who refuse to take part in such research which may heavily skew the data we are looking after. Conversely, people who respond to the survey may have some biases that correlate with their choice of participating in this.
True
I used to work in an auto parts store. I had a customer come in with a set of brake pads to warranty and to buy a new brake rotor. Upon inspection of the pads, I noticed that one side was almost brand new and the other side was worn down to the metal. I told the customer that whatever side the brand new pads came off of needed to have the brake caliper replaced.
He thought that the side that was worn needed to be replaced. I asked him to imagine two people pushing a boulder up a hill, but only one was doing the work while the other supervised. When the boulder gets to the top of the hill, who's going to be tired, the worker or the supervisor? He said the worker, and at that, I held up the worn brake pads. He immediately understood and ordered a new caliper for the side that was brand new.
Excellent.
This is a major issue that is ignored. It doesn't fit the narrative.
I'll try and make my point clearer by stating why I no longer take part in surveys.
A very well known company asked me to take surveys for them. The surveys that I was asked to take part in where sponsored (paid for) by companies that wanted to prove something.
They always prevetted the survey with a series of questions that were designed to 'select' those that would do that. The perfect example?
Right. Framing the question or creating a bias via the options available is a widespread way to manipulate public opinion. We need more people to spot that!
This is done by people pushing the gun control narrative. Democrats are more likely to have a bias against guns, whereas Republicans are more likely to have a bias for guns. So they get Democrats to participate in their surveys by asking very specific feeder questions to ensure they get the desired audience to result in the desired responses.
We need more sprouts videos on Biases
Glad we think alike ;)
If I did 10 neighbors challenge the police would eventually show at my house and start asking ME questions..
haha
one of my favourite examples of selection bias comes from one of my old schoolteachers, he always started each class by asking everyone not present to raise their hands
Really appreciate the dedication to finding the truth in science.
Thank you Sprouts, very good video
Glad you enjoyed it :)
Super topic
Excellent point. I used statistical sampling techniques in my job for years and was aware of the bias issue in selecting the sample(s).
None of my neighbors would answer the door so I couldn't get an answer...or did I?
Oh... O_O
Awesome video by the way
awesome explanation good job
Thanks a lot 😊
Exactly why one must delve deeper into questions, and NOT rely on statistics that may be incomplete due to manipulation of outcomes.
😉👍
Woooow you are great at teaching complicated theories in simple manner 😊
Also called survivorship bias.
True!
great video!
Glad you enjoyed it
I knocked on 10 doors. Nobody opened. So my deduction is they all have to love strangers.
A protected aircraft engine will not matter if the wings carrying the EXPLOSIVE FUEL, that is also necessary for the engine to operate, is vulnerable
It doesn't explode very often.
mostly it just drips out and burns there.
"We conducted a survey among volunteers, and it turns out people really like surveys!"
Please Make a Video on SISU the finnish art of courage
tAKE notes yall this is how influencers get people to believe bullshit is from selection bias
lol knock on ten neighbors doors and become the neighborhood weirdo by asking if they’re afraid of strangers… no more Big Bang theory for you, you’ll end up getting a van if get the reference 😂😂✌️
What is the rate of deaths of non-patients at hospitals vs out of hospitals?
Part of the entire picture is that we humans want a quick and clear conclusion. We tend to regard looking into the details as being tiresome and also have difficulties accepting uncertainties or the unknown.
Interesting 🤔
This is why most governments don't fight corruption. If they do, their time in power will be marked by scandals
Mind blowing ‼️ is media narrating with selection bias or are we selection biased🤔
They get an order for "correct data", invite low income actors to participate (if some visibility of actual population is required) or go to the areas where they find the people they need.
My guess would be that me asking my neighbors if they are afraid of strangers would taint the results. Basically, it's a stranger asking if you're afraid of strangers. Creepy and a potential to meet my local law enforcement. 😳 😅
I never really believed those political pools they presented in tv and news and whatnot, only to reafirm my beliefs the day I was stopped and asked a few questions... last bit of my interview (roughly translating):
Interviewer - "what's your academic level?"
Me - "just graduated as an engineer."
Interviewer - "Ow, ok... sorry, your info is not needed, we are only asking people without a university degree."
I won't answer my door if you knock on it so don't waste your time.
If you don't answer the door, you've answered the question.
**KNOCK** **KNOCK**
I can save your family from Doulingo
@@alext8828 i think that be selection bias.. If you knock on a door while NOT knowing if someone's home, you can't make up your mind that the person IS home and therefore doesn't want to talk. Do not count this as one of ten houses and move on to the next.
Haha
I won't, trick
also the bed is a very dangerous place. Most people end their lives while in bed!
And when women were confronted with the data the refused to admit they were wrong 😆
I interviewed 300 people who played Russian Roulette. The survival rate is 100%.
I knocked on 10 of my neighbours doors.
They were all home but only four answered.
They told me that they had no problem with strangers.
So four out of four, or 100%, of my neighbours aren't afraid of strangers.
Interesting...
Shouldn't you have tried to open the door? If they left their door open, wouldn't this speak volumes as to whether they were afraid of strangers or not?
I don't understand😢
I'm too afraid of strangers to do the task you've given...either that or I'm too lazy. :P
So this is the weakness in pattern seeking, but how could you possibly know you have selection bias unless the sample or statistic is exposed through the addition of a larger context?
Is someone's best information invalidated because elsewhere in the universe the majority is the inverse of their every experience?
Valid question! Maybe by seeking for other possibilities? Questioning the reliability of any given information!
It is only true until it is falsified ;)
Went to knock at neightbors. Got shot...
O_O
What's with the eggs 🥚 🤫🤭🤣
Metaphors! ;)
maybe not a good idea to knock on the door of your neighbors especially if you live in American
Climate change *cough*
😊😊😊😊😊
😊😊😊😊😊
😊😊😊😊😊
😊😊😊😊😊
Did they? No. Most certainly not. I've yet to find someone who believes that reality is a figment of their own imagination. Perhaps its due to the fact that anyone like that wouldn't be interested in research.
so an example of selection bias would be the gender wage gap?
You forgot to mention purposeful creation of selection bias.
Also known as 👉🍒