Before I started tutoring I believed that anyone can learn first year math and science if only they put in the time and effort. Before I went to grad school I believed that I can learn all the advanced math and theoretical physics topics I was interested in. Neither belief survived experimental testing.
I am interested in this—what exactly happened? Feel free to reply in private if you wish. Was it the case of:
(a) Morale breaking (this is not meant to be judgmental, this happens very often in graduate school, and certainly happened to me). Morale management is really hard.
(b) You felt lots of people in your program were much faster/smarter than you? (Also very common..)
(c) You felt you could learn [topic], but it would take unreasonably long (e.g. not 4-6 years it takes to get a thesis out)?
(d) You felt that literally you just could not get something, regardless of time investment? Could you give an example of such a topic?
All of the above, but the root cause is limited aptitude. I know you don’t believe that, but you probably will the day you hit your own limit.
Music is a good example. You can aspire to play the hardest and most exquisite music pieces with the best, and compose new masterpieces, but without the talent you will not progress much farther than “twinkle twinkle little star” (i’m exaggerating a bit).
Or sports. Not everyone who wants to makes it to the major leagues.
In math and sciences I have frequently observed a really motivated person learning something with extreme effort, doing the exercises, then coming to the next session with half the newly learned skills gone, and having to start nearly from scratch. As a result, the effort which is linear for many is exponential for them. Or worse.
I was in a similar situation. A couple of grad courses were easy, some harder, and one or two nearly impossible for me. I was able to do well enough on them, but it was hell. There would be no way for me to get to the level where I could do research in the area. Yet some other students just kept going, mastering the new material at the same rate as the old. (And others were forced to drop the course or the program long before.)
Think of, say, a high jumper. You can see one barely clearing 2.20 on the technique alone, with no hope of going higher, And you can see someone else doing the same height in a much more sloppy way, clearly able to do a lot better with a better technique. Brains are not much different from muscles. The limits are there, if not clearly visible.
Re your question (d), I have never tried to put enough time myself to test it, but I have tutored an aspired programmer who gave up after realizing he cannot think in the way required (see also 99.5% of programming job candidates fail the FizzBuzz test, though this seems a stretch).
(see also 99.5% of programming job candidates fail the FizzBuzz test, though this seems a stretch).
This is misleading. Bad programmers spend more time interviewing before being hired, thus the pool of job interview candidates is biased towards bad programmers.
Even if a bad programmer did 200 times as many interviews as a good programmer, that would mean that about half the programmers can’t do FizzBuzz, which is still unsettling.
If your idea of a “bad programmer” is someone who studied programming, but had unimpressive results, then yes, the idea that half the programmers can’t do FizzBuzz is unsettling.
However, the set of “bad programmers” also includes crazy people who believe they understand programming without any good reasons; overconfident people who used Excel for a few months and now believe they know everything there is about using computers; etc. It is not so difficult to believe that these people are as numerous as the real programmers.
In other words, instead of a less skilled programmer, imagine a non-programmer with an extreme case of Dunning–Kruger effect.
By the way, I wonder how much this effect is culture-dependent. There seems to be something in the American culture that supports overconfidence, at least in job interviews.
By “programmer” in this context I meant ‘someone who applies for a programming job and makes it to the interview stage’. Which unless they outright lied on their CV means they probably have some kind of certification. In another article I read that more than half of comp sci graduates can’t do FizzBuzz.
In a halfway decent world, granting a comp sci degree to someone who can’t do FizzBuzz would be punishable as fraud.
99.5% of programming job candidates fail the FizzBuzz test
This cannot be right. I have a variant of this (using an excel spreed sheet) on a technical interview for data analysts, which is a pretty low level position (the average candidate has an associates degree and “some knowledge of excel”). 60%-70% of the applicants, with no claimed programming experience, can make an excel sheet do the fizz buzz thing.
If it was true that 99.5% of candidates fail the FizzBuzz test, then someone who passes it is better than 99.5% of the candidates who get to the interview stage, and should be hired immediately for any computer software job they try out for (unless you believe more than 100 people on the average get interviewed before anyone is hired) . The experience in the job market, of people who can pass the test, does not bear this out.
unless you believe more than 100 people on the average get interviewed before anyone is hired
This is accurate for the top companies- as of 2011, Google interviewed over 300 people for each spot filled. Many of these people were plausibly interviewed multiple times, or for multiple positions.
The job market isn’t just Google. Is it really true that anyone who can program FizzBuzz will immediately get snapped up by the first place they apply to, if they are not applying to someplace like Google which receives such large numbers of applications? I find it hard to believe that the average accounting company or bank that needs programmers has to do 100 interviews on the average every time it hires one person.
(Furthermore, multiply by how many competent programmers they go through. If they hire on the average 1 out of every 4 competent programmers who applies, that makes it 400 interviews for each new hire.)
You seem to be confusing applicants with people who are given interviews. Typically less than half of applicants even make it to the interview stage- sometimes much, much less than half.
There’s also enough evidence out there to say that this level of applicants is common. Starbucks had over a hundred applicants for each position it offered recently; Proctor and Gamble had around 500. This guy also says it’s common for programmers.
You seem to be confusing applicants with people who are given interviews
No, I’m not. From shminux’s link:
The “Fizz-Buzz test” is an interview question designed to help filter out the 99.5% of programming job candidates who can’t seem to program their way out of a wet paper bag
The quote does not claim there has been no filtering done before the interview stage. If you read the original source it explicitly states that it is considering all aplicants, not only those who make it to the interview stage: “We get between 100 and 200 [resumes] per opening.”
You are confusing two different sources, the one that mentions FizzBuzz and the one in your link. Although both sources use the number 200, they are using it to refer to different things. It is the former (which uses it to refer to interviewees) which I object to, not the latter (which uses it to refer to resumes), except insofar as the latter is used to try to prove the former.
No I’m not. The Fizzbuzz article cited above is a wiki article. It is not based on original research, and draws from other articles. You will find the article I linked to linked to in a quote at the top of the first article in the ‘articles’ section of the wiki article; it is indeed the original source for the claim.
The wiki article uses as a source for the FizzBuzz statement the article at http://tickletux.wordpress.com/2007/01/24/using-fizzbuzz-to-find-developers-who-grok-coding/ . The wiki does not use as a source the article you just gave me a link to, which is http://www.joelonsoftware.com/items/2005/01/27.html and contains the “We get between 100 and 200 [resumes] per opening” quote. What you describe is neither the source for the statement, nor the first link in the articles section, but the second link in the article that is the first link in the articles section. It is a stretch to claim that this is the wiki’s source when the statement directly contains a source which is not the article you point to.
Furthermore, if you follow through the chain of articles, you find that because writers are playing a game of telephone with articles, the separate claims that people 1) cannot solve FizzBuzz (at a rate of 50% over computer science graduates) and 2) cannot program (at a rate of 99.5% over resumes) have been morphed into the Frankenstein-like claim that 99.5% cannot solve FizzBuzz as an interview question, which is not what either source says and which spuriously combines the two and changes from the plausible resume to the implausible interviewee. That combined statement is the one that I said doesn’t fit a basic sanity check. And it doesn’t.
Let’s simplify for the moment and assume that all software developers in the world could be ranked in absolute order of skill, and that you had a magical screening process that found the “best” person from any field.
Now, when you get those 200 resumes, and hire the best person from the top 200, does that mean you’re hiring the top 0.5%?
“Maybe.”
No. You’re not. Think about what happens to the other 199 that you didn’t hire.
They go look for another job.
That means, in this horribly simplified universe, that the entire world could consist of 1,000,000 programmers, of whom the worst 199 keep applying for every job and never getting them, but the best 999,801 always get jobs as soon as they apply for one. So every time a job is listed the 199 losers apply, as usual, and one guy from the pool of 999,801 applies, and he gets the job, of course, because he’s the best, and now, in this contrived example, every employer thinks they’re getting the top 0.5% when they’re actually getting the top 99.9801%.
Taking a quote from somewhere else as a reply always risks the possibility that it doesn’t quite fit what it is being used as a reply to.
I was pointing out that the described competence level implies that a competent programmer must be in the top 0.5% of the candidates for the job, not the top 0.5% of all programmers in the world. Of course your quote is in reference to the latter, not the former, and is therefore off point. In fact, your quote says that the former is indeed true, but the latter should not be confused with it.
(Furthermore, the original FizzBuzz reference claims that only 1 out of 200 people can solve FizzBuzz as an interview question, not as something required with each resume. Only hiring 1 out of 200 candidates who submit resumes is a heck of a lot more plausible than only hiring 1 out of 200 candidates who get to the interview stage.)
Taking a quote from somewhere else as a reply always risks the possibility that it doesn’t quite fit what it is being used as a reply to.
The quote might not fit perfectly, but the insight does.
I was pointing out that the described competence level implies that a competent programmer must be in the top 0.5% of the candidates for the job, not the top 0.5% of all programmers in the world.
And the point of the quote is that this really doesn’t say as much as you think. Hence why “99.5% of candidates fail the FizzBuzz test” isn’t as implausible as on first glance.
Sorry if I was being presumptuous. I was going by your advice to people with low math aptitude to learn more math, here and on SSC. If I confused you with someone else or grossly misrepresented your views, please disregard.
I hit c) for the category theory course in my masters. I managed the first half, more or less, but it felt like it was ramping up exponentially; there were too many new layers of concepts all of which were defined in terms of the previous one, and every new layer meant a percentage slowdown in my ability to work with that concept.
During undergrad I’d been at about the 30th percentile, but only the best half of undergrads go on to do a masters (at least at that particular institution). In retrospect it shouldn’t have been a surprise that I was towards the bottom of the class, but it was.
Before I started tutoring I believed that anyone can learn first year math and science if only they put in the time and effort.
Weirdly, I had the exact opposite conversion via tutoring, where “anyone” = “college students.”
EDIT: I should clarify, I was a mediocre tutor. However, the head of the tutoring center was incredible. He regularly had people who were failing college algebra and science for non-majors and turned them into chemistry majors. In the sessions where he tried to mentor me, my students were obviously learning more than when I was by myself.
A lower bound on your first claim: Most everyone accepts that there exist people with severe intellectual disability. For many causes, the degree of impairment may range smoothly from severe into the “normal” range, where the bright lines are imposed by functional requirements like living independently or managing health care, and not by any well-defined abstract mental capability.
Before I started tutoring I believed that anyone can learn first year math and science if only they put in the time and effort. Before I went to grad school I believed that I can learn all the advanced math and theoretical physics topics I was interested in. Neither belief survived experimental testing.
I am interested in this—what exactly happened? Feel free to reply in private if you wish. Was it the case of:
(a) Morale breaking (this is not meant to be judgmental, this happens very often in graduate school, and certainly happened to me). Morale management is really hard.
(b) You felt lots of people in your program were much faster/smarter than you? (Also very common..)
(c) You felt you could learn [topic], but it would take unreasonably long (e.g. not 4-6 years it takes to get a thesis out)?
(d) You felt that literally you just could not get something, regardless of time investment? Could you give an example of such a topic?
All of the above, but the root cause is limited aptitude. I know you don’t believe that, but you probably will the day you hit your own limit.
Music is a good example. You can aspire to play the hardest and most exquisite music pieces with the best, and compose new masterpieces, but without the talent you will not progress much farther than “twinkle twinkle little star” (i’m exaggerating a bit).
Or sports. Not everyone who wants to makes it to the major leagues.
In math and sciences I have frequently observed a really motivated person learning something with extreme effort, doing the exercises, then coming to the next session with half the newly learned skills gone, and having to start nearly from scratch. As a result, the effort which is linear for many is exponential for them. Or worse.
I was in a similar situation. A couple of grad courses were easy, some harder, and one or two nearly impossible for me. I was able to do well enough on them, but it was hell. There would be no way for me to get to the level where I could do research in the area. Yet some other students just kept going, mastering the new material at the same rate as the old. (And others were forced to drop the course or the program long before.)
Think of, say, a high jumper. You can see one barely clearing 2.20 on the technique alone, with no hope of going higher, And you can see someone else doing the same height in a much more sloppy way, clearly able to do a lot better with a better technique. Brains are not much different from muscles. The limits are there, if not clearly visible.
Re your question (d), I have never tried to put enough time myself to test it, but I have tutored an aspired programmer who gave up after realizing he cannot think in the way required (see also 99.5% of programming job candidates fail the FizzBuzz test, though this seems a stretch).
This is misleading. Bad programmers spend more time interviewing before being hired, thus the pool of job interview candidates is biased towards bad programmers.
It isn’t cited + it seems awfully high → the number is probably exaggerated at some level of intentionality
Even if a bad programmer did 200 times as many interviews as a good programmer, that would mean that about half the programmers can’t do FizzBuzz, which is still unsettling.
If your idea of a “bad programmer” is someone who studied programming, but had unimpressive results, then yes, the idea that half the programmers can’t do FizzBuzz is unsettling.
However, the set of “bad programmers” also includes crazy people who believe they understand programming without any good reasons; overconfident people who used Excel for a few months and now believe they know everything there is about using computers; etc. It is not so difficult to believe that these people are as numerous as the real programmers.
In other words, instead of a less skilled programmer, imagine a non-programmer with an extreme case of Dunning–Kruger effect.
By the way, I wonder how much this effect is culture-dependent. There seems to be something in the American culture that supports overconfidence, at least in job interviews.
By “programmer” in this context I meant ‘someone who applies for a programming job and makes it to the interview stage’. Which unless they outright lied on their CV means they probably have some kind of certification. In another article I read that more than half of comp sci graduates can’t do FizzBuzz.
In a halfway decent world, granting a comp sci degree to someone who can’t do FizzBuzz would be punishable as fraud.
This cannot be right. I have a variant of this (using an excel spreed sheet) on a technical interview for data analysts, which is a pretty low level position (the average candidate has an associates degree and “some knowledge of excel”). 60%-70% of the applicants, with no claimed programming experience, can make an excel sheet do the fizz buzz thing.
There are multiple possible interpretations:
For some reason your candidates have the neccessary ability. Possibly due to pre-selection, job profile, your area, whatever.
Your excel setup leads itself for easier realization of the FizzBuzz test.
Not having programming experience may actually help here as there is no standard solution where you can get stuck.
If it was true that 99.5% of candidates fail the FizzBuzz test, then someone who passes it is better than 99.5% of the candidates who get to the interview stage, and should be hired immediately for any computer software job they try out for (unless you believe more than 100 people on the average get interviewed before anyone is hired) . The experience in the job market, of people who can pass the test, does not bear this out.
This is accurate for the top companies- as of 2011, Google interviewed over 300 people for each spot filled. Many of these people were plausibly interviewed multiple times, or for multiple positions.
The job market isn’t just Google. Is it really true that anyone who can program FizzBuzz will immediately get snapped up by the first place they apply to, if they are not applying to someplace like Google which receives such large numbers of applications? I find it hard to believe that the average accounting company or bank that needs programmers has to do 100 interviews on the average every time it hires one person.
(Furthermore, multiply by how many competent programmers they go through. If they hire on the average 1 out of every 4 competent programmers who applies, that makes it 400 interviews for each new hire.)
You seem to be confusing applicants with people who are given interviews. Typically less than half of applicants even make it to the interview stage- sometimes much, much less than half.
There’s also enough evidence out there to say that this level of applicants is common. Starbucks had over a hundred applicants for each position it offered recently; Proctor and Gamble had around 500. This guy also says it’s common for programmers.
No, I’m not. From shminux’s link:
The quote does not claim there has been no filtering done before the interview stage. If you read the original source it explicitly states that it is considering all aplicants, not only those who make it to the interview stage: “We get between 100 and 200 [resumes] per opening.”
You are confusing two different sources, the one that mentions FizzBuzz and the one in your link. Although both sources use the number 200, they are using it to refer to different things. It is the former (which uses it to refer to interviewees) which I object to, not the latter (which uses it to refer to resumes), except insofar as the latter is used to try to prove the former.
No I’m not. The Fizzbuzz article cited above is a wiki article. It is not based on original research, and draws from other articles. You will find the article I linked to linked to in a quote at the top of the first article in the ‘articles’ section of the wiki article; it is indeed the original source for the claim.
The wiki article uses as a source for the FizzBuzz statement the article at http://tickletux.wordpress.com/2007/01/24/using-fizzbuzz-to-find-developers-who-grok-coding/ . The wiki does not use as a source the article you just gave me a link to, which is http://www.joelonsoftware.com/items/2005/01/27.html and contains the “We get between 100 and 200 [resumes] per opening” quote. What you describe is neither the source for the statement, nor the first link in the articles section, but the second link in the article that is the first link in the articles section. It is a stretch to claim that this is the wiki’s source when the statement directly contains a source which is not the article you point to.
Furthermore, if you follow through the chain of articles, you find that because writers are playing a game of telephone with articles, the separate claims that people 1) cannot solve FizzBuzz (at a rate of 50% over computer science graduates) and 2) cannot program (at a rate of 99.5% over resumes) have been morphed into the Frankenstein-like claim that 99.5% cannot solve FizzBuzz as an interview question, which is not what either source says and which spuriously combines the two and changes from the plausible resume to the implausible interviewee. That combined statement is the one that I said doesn’t fit a basic sanity check. And it doesn’t.
What you’re missing is the following insight:
Taken from here.
Taking a quote from somewhere else as a reply always risks the possibility that it doesn’t quite fit what it is being used as a reply to.
I was pointing out that the described competence level implies that a competent programmer must be in the top 0.5% of the candidates for the job, not the top 0.5% of all programmers in the world. Of course your quote is in reference to the latter, not the former, and is therefore off point. In fact, your quote says that the former is indeed true, but the latter should not be confused with it.
(Furthermore, the original FizzBuzz reference claims that only 1 out of 200 people can solve FizzBuzz as an interview question, not as something required with each resume. Only hiring 1 out of 200 candidates who submit resumes is a heck of a lot more plausible than only hiring 1 out of 200 candidates who get to the interview stage.)
The quote might not fit perfectly, but the insight does.
And the point of the quote is that this really doesn’t say as much as you think. Hence why “99.5% of candidates fail the FizzBuzz test” isn’t as implausible as on first glance.
You should reread what I actually said.
Sorry if I was being presumptuous. I was going by your advice to people with low math aptitude to learn more math, here and on SSC. If I confused you with someone else or grossly misrepresented your views, please disregard.
I was talking to one specific person.
I hit c) for the category theory course in my masters. I managed the first half, more or less, but it felt like it was ramping up exponentially; there were too many new layers of concepts all of which were defined in terms of the previous one, and every new layer meant a percentage slowdown in my ability to work with that concept.
During undergrad I’d been at about the 30th percentile, but only the best half of undergrads go on to do a masters (at least at that particular institution). In retrospect it shouldn’t have been a surprise that I was towards the bottom of the class, but it was.
Weirdly, I had the exact opposite conversion via tutoring, where “anyone” = “college students.”
EDIT: I should clarify, I was a mediocre tutor. However, the head of the tutoring center was incredible. He regularly had people who were failing college algebra and science for non-majors and turned them into chemistry majors. In the sessions where he tried to mentor me, my students were obviously learning more than when I was by myself.
A lower bound on your first claim: Most everyone accepts that there exist people with severe intellectual disability. For many causes, the degree of impairment may range smoothly from severe into the “normal” range, where the bright lines are imposed by functional requirements like living independently or managing health care, and not by any well-defined abstract mental capability.
Could you provide examples of advanced math that you were unable to learn? Why do you think you failed?
derp.