I thought it was about time I wrote something interesting — something outside of the bounds of routine admissions announcements.
But where to add value? Right now we’re in a strange time. A time of delight for some, and anger and confusion for others.
Guidance counselors are calling flabbergasted that the best students of their careers were denied admission. Mothers are dealing with despondent teenagers. And, as we know from message boards and emails, some kids are lashing out.
But what you might not see, is that there is also a lot of positivity out there. Many of the questions that we’re getting are totally respectful, thoughtful, and in many cases difficult to answer. Especially for someone like me who is new to this business.
When I think about it though, I think most of the issues I’ve encountered this year, as they relate to decisions, can be answered by two things: having a little zen, and a much better understanding of some statistical concepts. Sounds like a weird coupling, I know.
When I was in business school I had an introverted and very quirky stats professor. I remember the first day because he made a statement that seemed outside his personality. He said, “In the years ahead, you will remember this class much more than the others you’ll take.” It seemed a little bold for this quiet introverted type guy. But he was right. Because errors in statistical thinking happen every day. On the news, in magazines, at the watercooler – it happens all day, and I do think of that class often.
So, let me tell you — it also definitely happens in terms of thinking about admissions. (Warning: some of the theorems below are probably “Labs’ theorems” more than they are actual statistical concepts.)
So here’s what I think is going on:
1.) Not understanding Independence:
When you play roulette, the casino puts up a digital board above the game letting players know what numbers have hit recently. The idea being that if you look up there and see that 28 hasn’t hit in like 30 spins, you’ll think, “OMG, 28 is totally due. I need to put my entire savings on 28!”
This is what they want you to think. But because every roulette wheel spin is INDEPENDENT, your chances of hitting 28 are just as good as if it just hit 28 the spin before. The ball has no idea that it didn’t land on 28 15 times in a row, its just spinning around following the laws of physics, not the laws of, “I saw number 28 in a dream and now I see that it hasn’t hit in 15 spins so now I know it’s going to hit.” The first law is true, the second is a construct of your imagination.
How this relates to admissions: Some kid says, “I saw that your acceptance rate is 8.2% overall. Now, I’m not the top kid in my school and I didn’t build a rocket in my backyard, so I’m extrapolating the 8.2%, giving myself minus 2 points for not being a valedictorian, another 4% for not building a rocket so I think my chances are 2.2%. This is too low a chance so I’m not applying.”
There’s actually a number of problems with the above story. But lets just stay with independence for a second. 19000 kids applied, yes. Around 1500 got accepted, yes. Does that mean that everyone has an 8.2% chance of getting in? No! The 8.2% is the result after the fact. Your application is independent – its not affected by the past, or constrained by some percentage. There is no way to know what any particular application’s chances are. So don’t do this. This ties into the next problem.
2.) Association is not causation.
If you read only one part of this blog – read this one. It’s the big one.
If you do a study correlating home prices and number of faucets in a home, you will find a strong correlation. You will not be making a mistake – your study will be correct. Also, if you do a study correlating elementary school students’ test scores with the size of their shoes, you will find a strong correlation. You will not be making a mistake – your study will be correct.
It’s true. These things are definitely correlated. But obviously, their correlation stems from other reasons. (More faucets means more bathrooms, which probably means bigger house, which means more expensive house. Bigger shoes means older kid, which means better at tests.)
So now you know they are correlated, but you would never make a mistake by saying one CAUSED the other would you? Putting a random faucet in your bedroom isn’t going to increase your home value. Stretching your kid’s foot isn’t going to make her smarter. Don’t do this.
How this relates to admissions: Some guidance counselor says, “Hi Chris, I have four kids in my school who applied to MIT. Two got in. Both of them are xylophone players. This is good to know because now I can tell all the kids to play xylophone because that’s clearly what you’re looking for.”
Uh Oh. Don’t find yourself constructing mental causation based on incomplete data.
One thing that is sort of embedded in the association vs. causation problem is the “confirmation bias,” which is what the fictional guidance counselor above just did. Here’s another example:
“MIT clearly only accepts people from (x Country) if they’ve participated in an International Olympiad.” (Because the two kids we picked did – ignoring all of their other merits.)
3.) Constructing strange prejudicial opinions and then projecting them on a sample you’re wholly unaware of. (Labs theorem)
This one is interesting – it could really be a cross department study in psychology and sociology. It’s loosely related to problems with causation and the confirmation bias but then takes a sharp left turn to whatever particular opinions or issues people may have.
It basically goes like this – here are actual comments I’ve heard this year:
- MIT has a clear preference for Asian students.
- MIT has a clear preference for Hispanic students.
- MIT has a clear preference for men.
- MIT has a clear preference for women.
- MIT has a clear preference for students from abroad.
- MIT has a clear preference for US based students.
- MIT has a clear preference for rich kids.
- MIT has a clear preference for kids from families with financial difficulties.
I’m not trying to make light of any xenophobia people may have, but to read the comments on a number of blog posts this week, you can literally find someone posting one of the above ADJACENT to someone else posting another of the above – which statistically is impossible! We can’t possibly have a preference for international students and domestic students at the same time. My favorite one was, “we only accept liberals.” Huh? Is everything now political?! I guess so. I just ate some chicken strips. What does this say about me politically? *sarcasm* (FYI – there is no question related to poltical beliefs on the application.)
Anyway, I sort of promised that this will be about statistics. So where does this theorem fall overall in the world of admissions? It’s relevant because it makes me sad how trapped people are in their microcosms. In their defense, we don’t allow them to see the whole picture. This opaqueness allows everyone to construct meaning where there is none.
Yes, it’s true that we seek out talented students of color. No, its not true that those students are less qualified or don’t deserve admission. People always forget the second sentence when complaining about the first.
Which brings me to another issue.
4.) The insistance that the unquantifiable be quantified. (Or, everyone has a different definition of merit.)
We love to quantify things in our life. Economics even tries to do it by using words like, “utility.” How many units of utility does your computer offer you in comparison to how many units of utility your hairdryer offers you may help you understand how valuable these items are etc, etc. Silly. The facts are that some things can’t be quantified. We should begin getting comfortable with that.
Take the SAT. Everyone loves to say they hate the thing until they do well on it. Let’s say you got a 760 math score. Nice work. You sent an important signal there. The signal is, “I’m (by one measure) a talented math student.” Your friend took the test as well, he got an 800. You’re thinking, “ok, we sent the same signal. Both good at math.” He’s thinking, “Aha! I’m a slightly better math student and want to be rewarded for that.”
It’s kind of ridiculous to think that admissions decisions could come down to someone accidentally missing a negative sign somewhere, not getting a perfect score, then not getting into college. Yet, this kind of insistance on the process being quantifiable is something I hear all the time.
So that’s why we use the word “holistic.” That aggravates people for its own reasons. “How can you possibly be holistic when your median scores are so high?” This flashes back to confirmation bias.
Let’s say we have two kids. Both have great grades, scores, recommendations, extra-curricular accomplishments, whatever. One of those has to take the bus an hour and half each way to her high school. This impresses an admissions counselor. The other student works for three hours after school every day cooking french fries for some megomaniacal fast food manager. This impresses an admissions counselor. How are we supposed to quantify this? Fryolaters get two points, yogurt shops get one cause that’s easier, and long bus commutes get three? Crazy! It’s not going to happen.
Forget about quantification – there is always going to be subjectivity.
Which bring me to the end.
5.) Expecting too much certainty: (Embracing subjectivity and randomness)
(Total aside – as someone who played sports, I refuse to believe, as any books about randomness say, that hot streaks in sports are all a figment of our imagination and are random occurrences!)
Now, this doesn’t mean that admissions selection is random. Too many people see an application to have anyone totally fall through the cracks. But here’s a statement for you that I believe: the likelihood that one of the kids we denied this year is super awesome and will end up doing something very impactful in a STEM related field is 100%. No doubt. 100% chance. It could be you. There is no way that a.) there are only 1500 talented kids in the world, and b.) that even if there were that we’re perfect in snagging them all. It’s impossible.
So if I’ve embraced the subjectivity, randomness and accepted that perfection is not attainable, I’d like you to as well.
Kids tell me crazy things. Things that make me depressed like, “I’ll never be happy unless I attend MIT.” Or, “I’ll never amount to anything because I got denied at MIT.” I can’t tell you how ridiculous that is. I have degrees from 3 different colleges and all of them have students attending them who are learning a lot and are having the time of their lives. This country is full of good colleges, with passionate professors, and staff. You’re going to be just fine. Take it easy on yourself. When I was a kid everyone was trying to get me to work harder. Seems like nowadays we adults are just trying to get you to relax for a second. Sheesh.
MIT is an awesome place, but don’t make it out to be some nirvana where everyone’s floating around on flying lazy-boys while their synthetic doppelganger is doing homework for them. Besides, don’t forget, I’m from California and it’s 27 degrees right now and it’s almost April. Not everything is awesome all the time.
So let me just end this post by saying this. If you got denied, there is nothing wrong with you or your application. I spoke to a kid the other day who wanted to know where his application was deficient. I looked at his file and read the notes in the file. This is what I told him:
“There’s nothing in here except very good commentary.”
“How is that possible? There must be something in there! I got denied you know!”
“There’s nothing here except high ratings, and a good, positive summary of your background.”
“Well, I must have done something wrong, I mean …. come on, what did I do wrong!”
“You did nothing wrong. This is an awesome application!”
And it was the truth.