I often read statements on this site that are like "I'm a super-motivated, highly-performing driven person and I ought to be able to take Artificial Intelligence from the Google Director of Search Technology and get credit for that", with which I agree entirely.
But I teach at a college that is good but not (by various measures) first tier. I've been teaching for more than 30 years, and I teach Calculus, and while I have many students who do well, the majority of my folks find Calc to be hard, very hard. So I perceive from the statement in the first paragraph that many HN readers may have some trouble understanding who is in these seats.
Perhaps half of the people in my class would have trouble writing down, say, what an arctangent is. An even larger issue, it seems to me, is that they do not have enough experience to know whether or not they understand something. They are not incapable dopes, but for instance if you go over homework numbers 11, 13, 15, and 17 and say "Do you understand?" they will nod. But you often find that if you ask them to do number 19, their pencil just hovers over the page.
I see a couple of notes that argue something like "Well, they were given the opportunity .." but that is too hard a line, for me, personally. They are good people but they need support or guidance, over a medium-term period.
TLDR: for first-rate performers this is great stuff. For the middle of the bell curve, not so much.
> the majority of my folks find Calc to be hard, very hard.
I'm not even a middle of the bell curve kind of student, going as far as being outright rejected from all of colleges I applied to. However, there is something I have learned over the years about learning hard topics:
First, dive in and be totally confused. It doesn't matter if you do not understand anything. Your brain will still start to form relationships from what you took in. Revisit the topic again in a week, a month, whatever seems appropriate. You will probably still have a lot of confusion, but the idea is to just refresh those aforementioned relationships. Take a period of time away from it again, and by the third time you visit the subject it will "magically" make sense.
But this does not really fit the traditional education model. If you try to revisit the topic in a couple of months, the course will probably be over. When you are directing your own learning, however, learning these hard topics really turns out to be not that difficult at all.
I don't know anything about your classes except for what's in your post, but in my experience, most teachers who have this opinion of their students are usually not aware that it may be their teaching style that results in basic math classes being "hard". That teaching style typically boils down to regurgitating what the book already provides, which isn't what students need. I've also found that teachers who use that methodology don't actually know the subject matter terribly well outside of whatever cookie-cutter textbook is being used.
I've never once seen a math teacher who got more than half the class to really properly "understand" in the way that's described. I've had teachers who went by the book, I've had teachers who ignored the book. I've had teachers who were boring, I've had teachers who were incredibly engaging. None of them really reached most of their students.
My differential equations professor had a 98% passing rate for that class over a 20-year span. He was an ex-Bell Labs and later Lockheed employee though, so he actually knew the material and how to teach.
Calculus is not basic math. And many students find it difficult. I remember Calculus 3, in particular, was hard for me. Now I look back at Green's Theorem, Lagrange multipliers, Vector fields and surface integrals and it all seems so basic. But that was NOT the case for the 17 year old me sitting in that big lecture hall.
Thanks for mentioning the ever-important votes! They mean so much to me.
Many students find It difficult because they have bad teachers, bad textbooks, or both. Also, in the future, spouting off math-related terms doesn't really add anything to the discussion. Anyone can open a book and look at the table of contents and copy some unfamiliar words down.
As a rule, whenever I accidentally up or down vote someone, I leave a message explaining the mistake and my reasoning. Doesn't happen often... but I always thought of it as a courtesy. Again... no offense was intended.
On the question of math topics not adding anything to the discussion. Well, we'll need to agree to disagree. The point was that Calculus... is NOT an easy subject to learn. And you get a sense of how esoteric it seems to an initiate by looking at the topics it encompasses.
Finally, try to relax a little man... no need to get snippy on HN. We should be able to have a discussion on a, relatively uncontroversial topic, without getting emotional.
You didn't offend me, I guess my sarcasm wasn't evident. I just find the voting system to be pointless. I also don't know how you thought I was being emotional.
As for calculus, it's incredibly easy to learn - you just need the correct book. Most students don't know that unfortunately. Inventing calculus on the other hand isn't easy, but fortunately most of that has been done for us.
I suspect most of [EDIT: many?] HN readers breezed through calculus using nothing but free online resources. With zero attitude, I say that I am truly impressed. I wish I was that amazing of a student.
I followed the more typical educational path of taking calculus in college. Some of it was easy, some of it was hard, but I seriously doubt that I would have been self-motivated enough to go through the entire sequence of calculus course material on my own. The pressure induced by being part of an in-person class, for a grade, and not for free, made me feel like I had to learn the material, and avail myself of whatever resources I needed, be it asking questions of the professor, going to tutoring sessions, reading extra books, whatever, to make sure that I learned it.
I totally wish I was so self-motivated that I could learn absolutely anything at any breadth and at any depth without any external reason to do so. But on my own, there are limits to what I make myself study. Which isn't to say that I'm stupid on my own, but rather, I appreciated the in-person college experience mainly for applying pressure on me to learn things that I wouldn't have otherwise. (And as a result, I did become more inclined to study broader and deeper on my own.)
In my cursory experience with MOOCs thus far, I don't see them applying that same pressure. Maybe they can; I'm not sure. But unless they do, I think that a lot of people (especially outside of HN, etc.) will find value in actually going to school.
Yes. If you draw a bell curve of all first year students, say at schools where the entering avg Math SAT is 500 or above, then the shaded area of those who struggle with Calculus would, in my estimation, encompass .9 or more of the area below that curve.
This is despite hundreds of years of effort by well-meaning, perceptive, and hard-working teachers and students to refine the presentation to make the subject as clear as possible. Because most teachers and most texts are not the dipshits of the world. In short, Calculus is hard.
The thing about massive courses is that they are massive. When you get big, you cannot help but get average. In my experience average students (say at the schools I mentioned above) need support, including reasonably intensive personal attention from an experienced and knowledgable teacher. I don't believe the typical person in my course can pick it up on their own. (Not to dismiss the potential; I think there is a lot of exciting stuff happening that can help people learn. But I've seen nothing that would make attention unneeded.)
I do perceive that many HN readers could just pick Calc up, and frankly so could I, which is a good reason to read HN. People here pique my interest every day. But often in the discussions following articles like the one here missing is an understanding that you (speaking to a typical reader) are not average.
I majored in math and had pretty good ACT and SAT scores but I still find Calculus and related math very difficult.
My unscientific observation is that some people find the analytical/continuous parts of math difficult but can still do really well with stuff like abstract algebra, combinatorics, linear algebra, and so forth. Other people (i.e. physics & engineering majors) are the exact opposite.
Then of course there are a select few who are good at both.
My brother is a math teacher and was more of a physics guy and his claim is that people who struggle with calc generally had poor instruction in trigonometry in their youth. I'm not sure how true this could be but I definitely hated my trig instructor the most out of any teacher I've ever had.
I simply have to disagree. I've only met two teachers I didn't think were dipshits and that's over the course of a B.S. in bioengineering, a masters in CS, and a trip through an MD/PhD program. These are all at top ranked schools, too. These people may have been smart, but they certainly weren't quality instructors. In my opinion, teaching should be something people do when they retire for fun. The two non-dipshits were doing exactly that.
Students should absolutely be able to pick it up on their own, but most classes use crappy textbooks and the teachers recite from the crappy textbooks. Stewart's Calculus is used by a ton of schools and that book is utter garbage. The guy is using it as a way to print money, which is fine, but you'd think after 10 versions, he'd eventually get things right.
I'd say our posts are completely opposite in thinking. You think most texts and instructors aren't dipshits, I think they are. You think Calculus is hard to learn, I think it's easy.
I think that on HN some people do not understand what the academic experience is like for the majority of students and you've provided confirmation of that.
I know what the academic experience is like for the majority of students, it was me my first year of college. I realized that almost all teachers are dogshit after that and only attended class when it was required, studying on my own. Most people don't come to that realization until it's too late, though.
For the 44 to 24 percent of students who passed the course, this was not a failure at all. I don't see why they should be denied access to education because the likelihood of failing the course is higher. Online courses should be offered cheaper than regular on-site courses to take the increased risk into account, but that doesn't mean they should not be offered at all.
But if your customers don't understand the product then all bets are off and you have a moral (and, frankly, self-serving) responsibility to explain the product to them.
The way it sounds to me, SJSU and Udacity and are going to take the time to study the data and figure out why it is that so many students failed. If, and this seems pretty likely to me, it turns out that the students were confused about the product (the courses) then I would imagine that steps will be taken to clarify things so that people can make choices that will lead to the best-possible outcomes.
Many price points, many products doesn't work if people don't understand the products.
In the education market, customers do not know what they are purchasing by definition. This is why there is so much crap in the industry -- companies have an incentive to spend money on marketing rather than content.
The latest round of MOOCs are not really monetizing educational content (which is and has been online elsewhere, often for free). They are packaging institutional prestige, the same dynamic Facebook used to drive student adoption. Khan Academy is the exception.
A lot of the material on coursera isn't that easily available for free. The courses usually take a new and integrated aproach that you would see in a university course, created and presented by an expert instructor. I don't care for their prestige, but probably coursera is using the instutional prestige as a selection criterion, and it works, mostly.
I also think Udacity is more in between Khan Academy and Coursera, in that Udacity tries to find new ways of online instruction.
And I think showing that they fail people actually increases the value of the course. I'm always wary of non-free online courses provided by universities that have an extremely high pass rate.
The problem that most current US and UK 'educators' just don't seem to understand is that the customer analogy does not fit into the learning situation. Why? For the simple reason that it takes more than just having money to actually learn something.
Also, just because some people are unable or unwilling to absorb and understand some knowledge does not make that knowledge and the way it is presented invalid or useless.
So every time a course like this is canceled because "half the customers are rejecting the product", another nail is driven into the coffin of education.
I wonder how the Open University has dealt with that problem, after all they have been running distance learning courses for large number of students for over 40 years:
Institutions are beholden to political pressures. While idealistically their goal is to provide the most raw education, failure makes them look bad. They don't operate in a vacuum. Just like the rest of society, they are influenced by good people that want the most education for the most people and by complete shitbags that want to take the resources they consume. Institutions, unfortunately, have to provide education in a way that is defensible from criticism. The mere perception of failure is an avenue for attack even though everyone knows that most MOOC students will be casual and not fully committed.
I'm distinctly of two minds about this. On one hand I believe interacting with other people is essential to learning. That studying together, or talking about a text, or hacking on a project together is where the best sort of critical thinking germinates. In that sense I am happy to see some evidence in favor of my hunch.
On the other hand, the course material that online classes are making available for free is tremendously valuable. I've never finished an online course but I have begun to regularly use online course materials as a supplemental resource to books I'm reading and to projects I'm working on.
I'm concerned that enthusiasm for publishing course materials online will dry up if evidence shows it's not a silver bullet for the world's higher education problems. In other words, I'm concerned that expectations are far too high and that educators and institutions are going to lose interest in a fantastic innovation just because it doesn't solve all our problems.
Online courses may not actively discourage collaboration, but they certainly do not force it in the same way that is possible in a traditional classroom. Don't get me wrong, I am a huge supporter of online courses. But a teacher staring you down, asking you a question, or encouraging you to work on a project in a small group is much more attention-grabbing than many online teaching mechanisms available.
I just finished an online neuroanatomy class at SJSU this past week. This just goes to show how dumb the kids have become in California's school systems. I went back to school this past year to change careers into health care. I took a year of pre-reqs at a local community college, and then this class as a pre-req for my program.
It is shockingly absurd how easy and dumbed down they make classes in California's lower echelon schools. They basically hand you everything you need to know on every test on a silver platter: you don't need to read the book; they give you a study guide and EXPLICITLY tell you what will and will not be on the test; they sometimes will outright give you the exam questions, and students STILL screw that up. I took an abnormal psychology class where the instructor gave EVERYONE the exam questions, and students STILL did not show up for the exam or got questions wrong. I was like, HOW THE HELL IS THIS EVEN HAPPENING???!?!
I don't find this system of learning at fault. I find the students at fault. I think they should fail these kids as a costly lesson: study or lose out on $2000 bones. Bunch of nit wits.
> I took an abnormal psychology class where the instructor gave EVERYONE the exam questions, and students STILL did not show up for the exam or got questions wrong.
This may not be what happened in your situation, but giving students every possible exam question – with the caveat that there's a ton of them and they have open-ended answers – is actually a really smart thing to do. Anyone who wants to pass the exam will spend an inordinate amount of time crafting the perfect answers (either by themselves or in groups, which is fine) and then memorizing those. You could even say it gamifies the learning process, as every day of study you're keen to get through as many questions as possible. It's almost like being asked to write a 100 mini-essays. You end up learning more, not less.
You're close, but think about incentives rather than offense. Think about it more from the perspective of the lecturer. What are your incentives here? An easy class will net you a high pass rate, good reviews from your students, easier office hours, less to prepare for lectures. A hard class will net you a low pass rate, bad reviews, irate students (and maybe parents) beating down your door, and takes a long time to prepare.
The incentive structure for lecturers is weighed very heavily towards easy classes. The only way to counteract this is for the entire institution to take pride in being difficult and create a culture where students value that difficulty as much as the institution does. That takes a heck of a lot of effort and prestige: Harvard, Oxbridge, Stanford, these are the sort of schools that can make such things happen. SJSU? Not so much. This is not to say SJSU isn't a good school. It actually is a good school with good staff. It's just very hard to stop things going out of alignment.
I recently interviewed many students who were taking online classes as opposed to in person classes.
I was surprised to learn that most chose to do them online so that they can cheat on the tests. Free from instructor supervision all tests were open book and thus were much easier than taking them in person.
While many of us have high ideals for online education, I've learned that the majority of students get a degree for the diploma first and the education second.
I don't see how this is different than normal schools. Many people I know who went to traditional school don't know anything coming out and yes, they mostly did it for the degree and little more.
I think that it really depends on the class. I know that the finals for certain MOOC courses I took were very difficult indeed, and it was absolutely impossible "google" the answers. Even with the open book and notes, the tests were challenging enough that these cheating mechanisms were pretty worthless.
Also note that there is a wide diversity in the course quality. Many of them are super easy and watered down to the point that it is meaningless, while others are quite challenging. While I wouldn't expect an interviewer to know which class falls into which category, just realize that many of these certs are virtually meaningless regardless of the results.
I've pondered creating a site for ranking the MOOCs that are available now, to serve both the students who take these courses and the employers who are facing these students.
Either way, the complaint of "diploma first" feels a bit off-hand. If people are finishing college without knowing how to program fizzbuzz, what is the difference between the good and the bad? Clearly it isn't the quality of the class: it is the quality of the person regardless his or her background.
"Usually, two of every five students earn a grade below C and must retake the course or change career plans. So last spring, Ellen Junn, the provost, visited Anant Agarwal, an M.I.T. professor who taught a free online version of the circuits class, to ask whether San Jose State could become a living lab for his course, the first offering from edX, an online collaboration of Harvard and the Massachusetts Institute of Technology.
...
The results were striking: 91 percent of those in the blended section passed, compared with 59 percent in the traditional class."
It is extremely easy to cheat on a lot of online courses. One major flaw is many instructors use a test-bank for all of their exam questions, which makes verbatim Googling for the answer a piece of cake.
There are examples to the contrary though. I just finished "Introduction to Systems Biology" from coursera, and the quizzes and exams where tough. They really required you to have understood the topic and googling didn't help me a lot.
I dont get why universities require closed book tests. Are professors so lazy/overworked that they can't come up with questions that can't be easily googled or referred from a text book?
Failure rates that high are normal at many European universities. The idea there often is "let many people in, but let them fail out if they don't make the cut".
Not everyone is suited to some of the rigours of higher education, and when you make it freely available, more people are going to try it out, and more people are going to fail. I wouldn't call a 58% failure rate a bad thing.
My understanding is that the failure they're referring to is among those paying the $150 to take the course for credit. I agree with your point (presumably many taking the course, free or otherwise, aren't college students) but this would be an important distinction.
As some who really enjoys online courses, this is really disappointing news. I know it takes a huge amount of dedication to finish an online class (especially a free one where the cost of dropping out is zero). Hopefully Udacity et al can figure out the pedagogy before the Georgia Tech online CS masters begins (http://www.omscs.gatech.edu).
I bet there will be many similar cases before the behavior of students taking MOOCs changes. I just hope it doesn't dissuade the effort. I'm sure the long-term benefit of the effort is worth it.
The behavior of students won't change quickly, maybe not at all. The outside expectations might have to be lowered. Maybe a diagnostic exam should be offered.
From what I understand, the admission criteria for the GT OMS is that one is accepted conditionally until he/she passes 2 courses with a B or better, and then is fully admitted.
A bigger red flag would be if almost all of the students passed the final. I'm not an advocate for deliberately failing students: the exam should be designed so that an A shows true mastery and understanding.
Many of my university's teachers took pride in their low pass rates. Not because they failed so many students, but because they passed a minority that had truly mastered the material.
The job of a university teacher is to help students learn - if a student fails the class, then the teacher has failed to do their job. Sure, it's possible that some of the blame for that failure may lie with the student, but nonetheless, the teacher still failed to achieve what they set out to do. A teacher failing a student is like a programmer shipping buggy software - it may be the case that, because of circumstances beyond your control (incomplete specs or dysfunctional management; unmotivated students or weak syllabuses), you couldn't do your best work in a particular instance, and you don't necessarily need to feel guilty about that; but you shouldn't be proud of it.
The problem with this point of view is that X classes are taken in parallel. So you start getting competition among the managers/teachers for your time, as each one understands that "spend all your time on just my class, my suggested resources, and as much time with just me as you need" will lead to the optimal amount of help they can give. They can't do any more as ultimately it's the student's job to learn and the transfer of tacit knowledge is incredibly difficult -- you can lead a horse to water etc. Some teachers give up on doing anything important around the crunch times, knowing they can't win (typically humanities teachers at a primarily technical or engineering or project-based school), some teachers just lessen the load as much as they're comfortable with, others take an "I'm the most important and difficult" attitude and the only one who will get an A will be the one who got an F in at least one other class.
I wish schools would restructure to be serial with context-dependent branching, much like learning on the job tends to be, that way you can spend Y weeks focusing intently on a small spectrum of topics, then move on. Some classes are 3 hours every week for 12ish weeks, if you serialized that you could be done in 12 days. Plus the student and the professor would be in harmony, each having the other's undivided attention.
And some of my university teachers simply made the exams so difficult and long, the scores had little correlation with actual mastery of material (Jim Barby, I'm looking at your midterm).
According to me, examination system for online courses would be better like GRE system. Student should be allowed to take course from anywhere but he should go to the authorized center for exam.
I wouldn't be surprised. I attended the neighboring SF State and it's bad over there too. Many people fail out of the CS 101 course (including aspiring CS majors), and the Calculus 1 course offerings all the time. What I find more troubling however, is that we have a significant portion of the student body failing and retaking remedial Algebra I and II classes (to the point where sections for those two classes alone outnumber pretty much all other math classes -- despite Calc and Stats courses being required for everybody), and that many have a similarly hard time meeting even the basic English requirements to graduate.
This is a serious problem that goes well beyond the supposed student 'laziness' theory that's so popular here on HN. Clearly our lower levels of education are screwing students up. Most of the students that get into college are seriously ill-prepared, and universities have taken it upon themselves to compensate for this by offering several remedial courses and lowering grading standards.
It may be tempting to suggest that universities just stop doing this, but we've got to remember that our society values that degree way too much to ever let them get away with that (just imagine the revolt if all the colleges across the country started drastically lowering their acceptance rates). This sort of mindless degree-worship is silly on it's own, but it allows universities to grab money from otherwise under-qualified students, and who's gonna make a peep about that?
I think it's pretty common for schools to intentionally use some lower level courses as washout courses to avoid wasting resources on someone who won't be able to cut it at the upper division. Remember that STEM degrees cost the university significantly more as well. Calc II, E&M, and the first algorithms course all have very high failure rates.
I completed half of an MBA program in the classroom (1+ yrs), then moved out of state and had to switch to an online program at another university, both of which were AACSB accredited.
I must say, the online program was much more challenging, from almost all areas like personal responsibility, participation and academic rigor. The only area that could use improvement is testing, where some professors would prefer multiple choice questions, but this is a problem in any program.
Regarding the San Jose classes, I think a lot of undergraduate courses essentially run on autopilot so if a student isn't interested, of course they aren't going to learn anything.
One could say that 22-44% is actually a much higher pass rate than most MOOCs, and this was a success. It's just that the school is stuck in an old way of thinking. What's wrong with using on-line courses to see where people stand out? Perhaps let them avoid Fs on their report card if they drop out soon enough. (This is true for many normal classes too) Designed right, the feedback mechanisms should empower the middle tier students too.
AP Computer Science is supposed to be a college-level Intro to Computer Science for high school students. In 2012, 26,103 AP CS students took the exam. 63.6% received a "3" or above, meaning colleges accepting AP scores probably awarded credit.
It also means 36.4% probably didn't receive credit.
A "3" sits on the bubble. Since AP scores are "recommendations of proficiency", not final exam grades in the traditional sense, students with "3's" are often advised to reconsider taking the course again. It depends on the school of course, but nobody complains. That's how it works.
Adding the 15.6% of students with a "3" to the 36.4% means 52% of students might not be "proficient" by definition.
The San Jose State numbers aren't shocking, they're about right given the total population. Plus, looking at the courses and the material, I'm willing to bet 40% drop the traditional class before the final anyway. Are they included in SJSU's comparative statistic?
I would like to see how many students represent the 44 to 24 percent.
If it's a couple hundred I wouldn't consider it failure.
Plus has anybody considered the fact that some people (like me) consider exams just way for professors/everybody to keep stats on how the class is doing and me taking the exam doesn't help me at all with learning what i wanted to learn from the course.
I have taken a lot of online courses because i was interested in the subject did the small tests sprinkled in because i found the vaguely useful never took the exams.
I like the small details/intricacies being pointed out to me in the tests but I'm not going to bother remembering it all just the knowledge of it being there is enough I'll look up the details when i need it.
I'm not sure if people realize (in general) how difficult online courses are in general, compared to in person classes. Not everyone is an aspiring hacker who is smart and disciplined and posts on HN. Online classes have long been shown to be more difficult than their in person counterparts.
I personally found the quote to be in poor taste to a certain extent. He's talking about experimenting with people's lives. Yes, the students agreed, at least in some sense, to be a part of the experiment, but it is still a rather callous response in my mind.
I basically came here to say this, as that comment stuck out from the article for me. I was thinking, that it must be nice to experiment using students' tuition money. I'm pretty sure the student didn't expect to be part of an experiment when they signed up.
But I teach at a college that is good but not (by various measures) first tier. I've been teaching for more than 30 years, and I teach Calculus, and while I have many students who do well, the majority of my folks find Calc to be hard, very hard. So I perceive from the statement in the first paragraph that many HN readers may have some trouble understanding who is in these seats.
Perhaps half of the people in my class would have trouble writing down, say, what an arctangent is. An even larger issue, it seems to me, is that they do not have enough experience to know whether or not they understand something. They are not incapable dopes, but for instance if you go over homework numbers 11, 13, 15, and 17 and say "Do you understand?" they will nod. But you often find that if you ask them to do number 19, their pencil just hovers over the page.
I see a couple of notes that argue something like "Well, they were given the opportunity .." but that is too hard a line, for me, personally. They are good people but they need support or guidance, over a medium-term period.
TLDR: for first-rate performers this is great stuff. For the middle of the bell curve, not so much.