Interviewing America’s best and brightest

A friend of mine has been looking to hire smart people from across a range of disciplines. Here are his tales from the trenches:

In theory the economy is horrible and people want jobs. In practice it seems that no one I interview has any real knowledge. I just interviewed a woman getting a PhD in electrical engineering and bioinformatics from [top school, but not quite at the Stanford level] and she couldn’t tell me the probability of flipping heads three times in a row with a fair coin. I just interviewed a 4.0 physics major from [top state school] and he didn’t know how to convert miles to kilometers, and had no idea how to convert cubic kilometers in cubic meters.

For the record this is depressingly common. I’ve probably done ~700 interviews in the last five years of grads from top-notch technical programs. Perhaps ten percent were able to do basic mental math, reason about probability, etc.

Tyler Cowen says that we’ve run out of low-hanging fruit (free land, big technology improvements, uneducated young people). In “Universities and Economic Growth”, I speculated that the lack of improvement in our universities, essentially unchanged since the year 1088, was a drag on economic growth. My friend’s experience with interviewing folks fresh out of 16-22 years of our best schooling, proves, I think, Tyler Cowen wrong on the subject of at least one of his three kinds of fruit tree. It is true that we’ve taken young people and given them high school diplomas and college degrees of various kinds. But that does not mean that we’ve educated them. So there may be room yet for significant economic growth driven by improved education.

22 thoughts on “Interviewing America’s best and brightest

  1. No, the problem is that you can’t teach people how to think because there’s no formula for that. All you can do is put ’em on the spot from an early age and demand it, and that’s unfair, because the wrong ones might fail. Furthermore, if we required thinking to graduate from college, we’d have fewer college grads, and we need more of ’em, not less!

    The nation that piles up the most empty boxes that say “college graduate” on them is the one that wins the future. It’s obvious.

    Solving problems is grunt work. Nobody needs to do that. Regular old thinking is obsolete. We need Critical Thinking instead. It’s much safer because it starts with the right answer from the get-go.

    Competition and thinking and stuff all involve uncertainty about outcomes, and no civilized nation can tolerate that. It would be an attack on the educated classes, or at least on those children of the educated classes who are as useless as their parents. Only a radical Luddite maniac would evaluate engineers without taking their violin lessons into account.

  2. For what sort of job is your friend soliciting applications? The recurring theme on this blog is that smart people are fleeing/should flee engineering and science for medicine and law, finance, and government work.

    Ergo, if your friend is interviewing people for something else–and _especially_ if he’s looking for scientists or engineers (as the EE PhD might suggest) — he’s not getting smart people because the smart people are all in med school, at Goldman, or working for Uncle Sam.

  3. My first real job out of the air force was at a new CRT plant. in theory those with better math skills were assigned as mixers, and basic algebra was required for the job. Everyoone had basic high school geometry, everyone was familiar with the metric system. When we were asked to find the approximate capacity in liters of various square and cylendrical tanks I was the only one out of 30 or so to use a tape measure and calculator, others were using 3 and 10 liter buckets to measure thousand liter tanks.

  4. J: Good question. I was trying to keep my friend (and the clueless folks with fancy degrees) anonymous. I will say that the pay is excellent and that new workers may find that “numbers are a lot more interesting when you put dollar signs in front of them” (my friend Max who quit his job teaching statistics at a top university in order to run a hedge fund).

  5. so he’s only looking for young people?

    I know plenty of middle aged people looking for work, who would have no trouble with that interview, or a harder one, with great credentials – and no one is hiring them, or even answering their letters.

    If you want specific references, email me.

  6. (Peemature phone post). We teach formulas, but not real world applications, not how to use the formulas, and especially not how to create a formula.

  7. dw: I think he is primarily looking for young people (he himself is pretty young). There is a huge prejudice in this country against older workers, as I’m sure you’re aware. It doesn’t help that we typically have employer-paid health insurance and the cost of insuring a 60-year-old is astronomical.

    (This is one reason that I stopped following the 2008 presidential election as soon as John McCain was nominated; I didn’t think too many voters would elect someone who was 90% dead.)

  8. Some of us who are smart enough to figure basic probabilities (seriously– the coin flip example has to be hyperbole, no?) and realize that you need to cube the 1000 (in the denominator, in your example) when doing the conversion, but not sufficiently intelligent to grok Eigenfunctions or moments in metroid spaces work in low finance. Generally speaking, we like it here.

  9. I’m young AND I know basic probability and unit conversion, hook me up!

    In all seriousness, I’m surrounded by these interviewees every day, so I’m not entirely surprised. I’m an undergrad in engineering. Many of my friends in engineering largely rely on solutions manuals (easily found online for nearly all textbooks) and rote memorization to get through courses that cover basic unit conversion and probability. I think lack of motivation and horrible teaching quality are key factors that lead to this.

  10. I find with that the one in ten is about right, whether Americans for folks with Visas from offshore.

    To bad he is looking for young, I was going to recommend he hire someone from the military who did tech. The military did all the work of filtering out the 9 in 10 and has spent their money educating the thinkers.

    Too bad the service doesn’t hand out degrees.

  11. Phil,

    I read anecdotes like these all the time, many from your blog. But how come I’ve never met any of these people in real life–engineering school graduates who can’t do high school math & physics?

  12. psb: Come to our helicopter ground school! Almost any weekend! (And I do think that they can do high school math and physics, which means that they know what kind of problem to expect and it will look just like the problems in the textbook; that’s how they got an A in the classes. But they can’t use the knowledge in other contexts, e.g., my friend’s job interviews or our flight school’s helicopter ground school.)

  13. This is probably a tangent, and I could be mistaken, but there seems to be a bias against learning/remembering/applying ‘NOT my homework’ material in the U.S. As a programmer in the software industry, pretty much every time I’m seen with a book on graduate-level probability/statistics or abstract algebra, my colleagues (including science PhDs) ask me how this knowledge (or the quest for it) is helpful outside pure research.

    It seems that people are mostly interested in only solving little homework problems they are assigned at work. And the market (apparently, a lot of hiring managers with PhDs in non-technical fields) seems to agree with this mentality. However, not only people seem to be uninterested in ‘this is not my homework’ stuff, they have apparently forgotten what does get taught (in the cited case, for example, the engineering problems where math is applicable to industrial software work).

    (Of course, all of this could be different in other disciplines, or at hard-core software or engineering companies.)

  14. I think thats whats so hard about medicine- the same thing presents differently in different people. Even elementary histology demonstrates this…smooth muscle samples take dye differently, have different striations, etc. The book will have one photo, but a real world sample often bears slight resemblence.

    Very rarely in college do I have to remember something for more than 3 weeks to get an A in a class. I found that after 16 weeks had passed since taking an elementrary anatomy class I scored signifficantly worse on the exams. This scares the hell out of me cause its probably good for doctors to be able to put this stuff in their heads and keep it there, but college doesn’t prepare you well for it. Sure chem and physics concepts build on each other, but in rare instances where you need concepts from other courses students whine that the question was unfair.

    We recently had to be able to find the area of a cylinder for a “physics for health care professionals” course so that we could estimate the atmospheric pressure applied to a human on earth. Students were stunned that we had to know stuff we’d all had in high school.

    I wish I knew how to fix this. If anyone does, lemme know cause my mcats are coming up.

  15. As an addendum, I should point out that in solving a lot of physics problems, conditions in textbook are so oversimplified that it is silly to apply them to the real world. Example: In that physics class we were told to determine whether or not a newly wed couple should purchase a water bed of certain dimensions given that their bedroom floor could hold a certain number of newtons. We’d never seen such a problem before and it had little to do with what we’d been talking about in lecture. After 3 minutes of pondering, the teacher asked how we might go about solving the problem. the room went silent for 2 minutes until the teacher called on me.

    I explained that it was a density problem similar to some things I’d done in chemistry, and since we know the density of water is 1, and its volume, we can calculate its mass and multiple by gravity to get newtons, and then simply compare it to the rated value. The prof was pleased, but students wondered how I knew the density of water as it wasn’t mentioned at all. The prof explained that this would normally need to be found in a chart. I added that I didn’t understand why the floor would be rated in newtons or even Kg for the room. What if the pressure of the bed was applied at 4 tiny foot posts? What if you stacked the bed on its side? Surely the floor should be rated in weight per area…which is apparently the general definition of pressure. I also pointed out that the weight of the people needed to be factored in, as well as the forces from acceleration due to sex, wrestling, or whatever activities are expected to take place on the bed. Again- not part of the problem 🙁 The prof said she absolutely agreed, but the book just wants us to understand density problems. I was sad that in the real world I can’t actually determine if I should purchase a water bed even tho I got an A on that problem.

    I asked an aeronautics engineer what the tolerance is on useful load limits for a cessna 152. If I’m 10lbs over will the wings tear off? Will the axles bend? What about 50 or 80lbs? She said only that there is no tolerance, its an absolute maximum limit, and she doesn’t want to know how the plane will fail if I push it. I have basically no idea how to take my knowledge of forces and vectors and general mechanics and apply it to this problem. I’d certainly like to. Where do I need to go to school to learn this?

  16. Your friend’s experience correlates with mine. I ask every candidate the same questions, example: I draw three pole-zero plots on the board and ask the candidate to pick the one with minimum phase (if you don’t know what it means, look it up on wikipedia, good writeup). If the candidate gets it right, we get into a discussion of what minimum phase means for equalization. These people have Master’s degrees in electrical engineering from US schools and bachelors degrees from Chinese or US colleges who were previously phone screened. 75% of candidates get this particular question wrong! I guess I am setting my sights too high. maybe I should start with coin flip probability.

  17. This reminds me of Feynman’s writing about his trip to Brazil and his encounter with students who knew exactly what the definition of triboluminescence was, but no actual interest or curiosity in such an interesting phenomenon.

  18. I thought the thing that stumped Feynman’s Brazilians was not triboluminescence, but what an index meant in the context of refraction. They knew it rote, but not how that translated into real-world phenomena. In any case, that was 40 years or more ago, and we are in a post space age. My sister is a teacher, and when she was in her teaching program learning all the latest and greatest instructional methods, I’d ask her (and I paraphrase): the kids that learned math in the 40’s and 50’s ran the space program, and most of your favorite authors are dead; what exactly have we gained in the last 40 years in the way of pedagogical progress?

  19. I thought I should point out that a) I’m a new physical science Ph.D. b) I can solve -those- problems.

    Also, I work as a research scientist in the federal government. (I’m working on problems there that I’m not as sure about finding immediate solutions.)

    I wonder what kinds of problems these folks your friend interviewed can solve?

    Also, I think the culture is such that people tend to stay in school if they can’t find work. If certain subject areas (say probability and stats) are boring, I can seem them brain dumping it. A huge number of people are in post graduate education because they want to get a job, not because they care about their subject area.

  20. As universities (including graduate programs) increase in size, so the best people get concentrated at fewer and fewr of them (as well as moving between specialties according to fashion and incentives).

    Theer never is much talent – but sometimes it is spread around, and sometimes it is mopped up

    (and sometimes it is not allowed to rise to the top: )

  21. bgc: Thanks for the links. The last paper was very interesting. I’m not sure it is a bad thing for employers that universities select for conscientiousness. It is pretty painful trying to supervise a young person who thinks of himself as a genius but who isn’t conscientious. The supposed manager ends up being that person’s secretary, basically. Especially in aviation. You want someone who will do every item on the checklist, not someone who can come up with a clever way to rewrite the checklist.

Comments are closed.