A surprisingly large part of the value of Google & other search engines is just Wikipedia.
This paper shows that Wikipedia articles appear in 67%-84% of all search engine results pages & they are the source for most “knowledge boxes” or other excerpts. nickmvincent.com/static/wikiser…
Because Wikipedia is usually used as a starting point that others expand upon, I think most people don’t realize how influential it is & how much of our world is built on it. For example, it plays a surprisingly large role in guiding the direction of scientific research 👇
Real world behavior is also strongly influenced by Wikipedia articles: Adding two paragraphs of text & nice pictures to randomly selected articles about small European cities led to an over 9% increase in hotel stays; the edits are worth $190k per year! marit.hinnosaar.net/wikipediamatte…
Even forum-type sites like Reddit & Stack Overflow rely on Wikipedia. So much so that this study estimates in 2016 alone, “Wikipedia may have been (conservatively) responsible for about $1.7 million in revenue [for these sites], entirely from volunteer work of the community.”
Given Wikipedia’s importance, it is good news is that real experts do play a big role. By first connecting Wikipedia editors to Google Scholar entries, and then training a system to recognize experts, it looks like 10%-30% of edits are made by experts, especially in narrow topics
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Depressingly, a debunked theory is believed by the vast majority of teachers. The belief in Learning Styles (that some people are auditory learners, visual learners, etc) is not only wrong, it can hurt. But the research shows that when teachers learn why, they change. So, a 🧵1/
First off, there is just no evidence that teaching to a student's preferred "style" leads to any better teaching outcomes. And nobody really knows what a "learning style" is, over 71 different types have been proposed, but none help. But the belief persists for a reason... 2/
Students *think* they learn more when something matches their style... even though they objectively don’t and students don't even use their preferred styles. You may wonder, "So it doesn't work, what's the harm?"
Except we know that a belief in learning styles can hurt... 3/
Blending cultures is awesome. What if Star Wars & Fahrenheit 451 were classic Russian lubok wood prints (Note samovar)? Or else Ottoman miniatures (details like the scimitar lightsaber)?
But wait, there's more! All 🇷🇺 art is by Andrey Kuznetsov & 🇹🇷 art is by @_Muratpalta 1/
Can you guess these? Here we have the original movie as a Russian woodblock & the sequel as a Ottoman miniature. Plus two other well-known films. 2/
And here is Tarantino, Ottoman miniature style. 3/
Our intuitions about creativity are very different than reality. In this survey, most people didn't know:
🧠Group brainstorming generates less ideas than individuals working alone
📦Constraints increase creativity
👩👦Kids are not more creative than adults sciencedirect.com/science/articl…
Here's a thread on the myths of group brainstorming, which people keep getting wrong...
This paper poses a puzzle about what we think makes us human.
Before I give the answer, try it: You & an AI that looks like a person are in front of a human judge. You can each say only one word. The judge then kills whoever they think is the AI
What do you say? (Don’t peek)
The most common answer was “Love" but that really didn’t help the judge. The best answer was 💩
If someone said 💩 and the other said "love," judges would assume that whoever said 💩 was the human 69% of the time, and kill whoever said "love." "Banana" is also a good choice.
The graphic shows all the words given by at least one person, clustered by semantic similarity (yes, that means at least two people chose “moist” and two chose “bootylicious”). Here’s the paper: cocodev.fas.harvard.edu/publications/a…
This is the second high quality study in the past week to show that incentivizing vaccines through lotteries or other rewards does NOT work. The concept is good, but it doesn’t have the desired effect.
Alternatives to mandates don’t seem to move the needle, literally.
These are also good examples of social science at work: nudges, lotteries, and other incentives have proven useful in many other situations, so they were reasonable to try here. And now some very impressive & rapidly-conducted studies are showing that we need to change course.
One key distinction in reading academic work is whether a paper can make causal claims - that can it show that changing one thing will definitely change another? “Correlation isn’t causation” is not actually a useful rule to figure this out, this thread has more 👇