Khan Academy: Computer Science
khanacademy.org
khanacademy.org
I really hate taking shots at Khan Academy, because I love what they do and his other videos have really helped a lot of my students in chemistry, economics, etc.
However, he's moving through the material MUCH more quickly than I've ever been able to do as a teacher. And I get nearly 5 hours a week of face time with my students.
For example, his third lecture is on for loops. My students, on the other hand, complete literally 80 assignments before beginning for loops. And they still have some trouble.
Admittedly I'm using Java, and Python is simpler, but I think he's VASTLY underestimating how slowly you need to move through seemly obvious concepts for the average total beginner.
80 assignments? What are these assignments? Are you teaching them how to use recursion before looping in Java? I get that Java has some ceremony, but not 70 assignments worth.
The first day I learned to program (with BASIC) I learned looping. I didn't know the details of it (like there was an index variable I could access), but it was as easy as anything else.
Seriously, what are they learning before looping?
See for yourself: http://cs.leanderisd.org/current/cs1/assignments.html
Anyway, the "Swing" assignments are there in the sequence for four reasons:
1) To break up the monotony a little bit. 2) To give them more practice copying details accurately. 3) To make them say, "That was kind of cool." 4) To explain why we're not doing "real" programs (a.k.a. GUI) yet. "It takes that much code just to display a blank window on the screen? Guess I can live with the command prompt a while longer."
I'm a bit disappointed to be honest.
I am very glad I already have been coding for years because our teacher and the planned curriculum seems to confuse people even more.
I think I will refer some people to your assignments, if that's okay? Hopefully someone can learn more from this :-)
I have 75 kids in three sections, and I'd have 100 kids in four sections if they'd open up a fourth section. And all of my students except maybe the bottom 5-10% are moving confidently through the material at their own pace.
I'll bet your school either has 1) a lot fewer students 2) a much more hand-picked set of kids, or 3) is leaving a much higher percentage of kids behind, who are either lost or cheating.
My school is a small (~130 per grade), but renowned international school in Shanghai, China. Your first two guesses are very much spot-on, but I'd have to disagree with the third. Those of us that are ahead help out with the one's who are struggling and in the end, I'd like to think that it all works out well.
With that in mind, however, your numbers are indeed more impressive. Because our class size is so small, we're constantly involved in face-to-face interaction with our instructor and with our peers. I doubt our pace would translate into the same success in your school. Is your course designed for the AP exam?
I do wonder if your most able students might appreciate more opportunity to "play" and be creative, but in all honesty, they just need to stick with it into the Project Euler stuff at the end.
Is this a whole-year course?
Yes, this is a whole year (36 weeks).
The material you're seeing is only the intro material, though. I teach three other courses, including video game development. The faster students are encouraged to skip assignments if they're coding something they already understand, and they get bonus points if they do something creative with an existing assignment.
I've tried to optimize for the normal case: the weak-to-average student. The great ones however, don't need much help from me to learn a ton very quickly.
I've always thought programming classes should start with reading and understanding a program that has some loops, assignments, etc. Maybe a program that draws some shapes or parses/prints some strings. Then have students modify it some. Then afterward, start writing your own stuff from scratch.
I think you remember things much easier if you write them out (and have to make them compile/execute) yourself. If I read some code in a language I haven't encountered before, I find I don't absorb any of the little details of syntax.
Moreover, I think it makes sense to expose learners to the fun of programming early. And to me that means seeing something that you wrote running.
Perhaps you don't write well until you've read a lot - but nobody expects you to write well as a novice.
I've done a lot of pondering about how to teach programming because I tried to learn several times myself, both on my own and in classes.
I bounced off each time, until finally sticking. I always wonder what made it different that last time, and if it applies to others.
===
Have you considered starting with something with less conceptual overhead than java-in-a-shell? There's so much you need to understand before you can even get to writing code that way. You need to learn how to start and navigate a shell on your computer (different on each type, and depends on your system configuration). You need to learn about paths, and java versions, and sdks, and what 'compiling' even means, and the concept of source code vs a binary.
This doesn't seem like much to people in the field, but I remember it being really intimidating.
In contrast, I really love codecademy.org:
You're writing code within seconds. You can access it on any computer and it always works the same way. Which is to say, it works like any other web site. You start seeing things happen right away, and you don't need to understand all the infrastructure initially.
===
I feel like getting people to that first high where they see their code generate output is really important. If you can get them quickly to feeling a little bit of that power that comes from software, that starts the fire, and they'll have more motivation to figure out terminals and paths and verisons and OSes and everything else.
We install the JDK, do a basic DOS primer, type in our first program (in Notepad), and compile and run it on the command line in a single 90-minute period, like the 3rd day of class.
Getting to the "bang" as quickly as possible is HUGELY important.
I know that when I teach myself anything it's best to focus on one new concept at a time. Scheme's light weight allows all but the CS concepts to fall away from the new coder's consideration.
And why not code in a low-friction way even when not being taught? I wasn't intending to defend FP, but I think it's silly to think of assignment as a fundamental part of CS or the art of coding. It's just one tool.
On the other hand, you learn about laziness, higher-order functions, deterministic and logic programming and how to implement the aforementioned, on top of a bunch of stuff you would learn in a Java class like recursion, OOP and networking. All this in the first four chapters; the fifth one introduces a low-level register based language but my class never got to it :( It is pretty cool.
Ultimately, SICP just has drastically different (and I think better) priorities and topics than any course taught in Java or Python. I also suspect that a good SICP course moves even faster than the material here, simply because of the sheer density of big ideas in the book and the simplicity of Scheme.
I was initially far more interested in how I could change the colours of things on screen and get the speaker to make funny noises. Most of what I wrote were text based adventure games so I didn't really need to loop very often.
When I did need to loop I just implemented this using a conditional IF with a GOTO that jumped back to the start of the "block" (I kept paper notes as to which line numbers each subroutine started on) if required or continued to the next line of code.
Later when I started using MS Visual Basic and figured out how to use GOSUB I would do loops essentially with recursion that would call the same subroutine again if required.
It wasn't until I decided to learn C and php that I learnt about real loop structures. At this point I assumed that everything I had done before was stupid and wrong, it was not until I took formal CS classes where I was introduced to recursion and assemly language (JMP and JE) that I realised that pherhaps I hadn't been quite so naieve before.
You really, really have no idea how difficult precise logical thinking is for an average person.
have you looked at: etoys scratch starlogo alice or scheme ?
At some point soon, I'm planning on moving to something where they can distribute what they're doing, either as JavaScript, or pygame, or something. I'm hoping to be able to literally translate a scratch program to the n+1 environment so that they can see the equivalence of the looping constructs and the event handlers and everything else.
I can't imagine anything that would have pleased me more than being able to show off my own web site when I was that age. Of course, making silly web sites for myself is how I got into programming in the first place, so I'm probably biased ;)
* Runs everywhere. It's the BASIC of this age * It's not that bad. * They can do stuff their friends can see
I'd rather do python, since it's that much cleaner, less curly braces and semicolons and better loops. But, the results aren't as public, and it's a harder environment to set up. And given that I have to work with the school's IT, I think I'm going to be lucky that we can load chrome on the system.
So, clarity for teaching, or ubiquity. Still deciding.
I will say that Stacy Armstrong's "A+ Computer Science" curriculum is quite good and well worth the money.
I've taught Pascal and C++ before, and dabbled in Scheme, too. In my opinion, Scheme is lower-friction than Java, but precise logical thinking is hard in any language, and precise logical recursive thinking is even harder for the average student, which negates any syntax-related benefits.
My kids are very successful in Java, and so I'll stick with what's working until the College Board changes the language of the AP exam on me.
in logo this is how you draw a square forward 100 Right 90 Forward 100 Right 90 Forward 100 Right 90 Forward 100 Right 90
Then in the second lesson you introduce loops repeat 4 Forward 100 Right 90 End
then you ask kids to draw hexagons octagons etc. and they will never forget what a for loop does.
Edit changed repeat 8 to repeat 4. Thank you.
I don't think it's optimal to make someone type
class Cross extends TurtleGraphicsWindow { public void myTurtleCmds() ...
Of course, the College Board, in its infinite wisdom, has decreed that the AP Computer Science tests be given in Java. I suppose we should be grateful that they no longer do it in C++.
I’ve heard of a CS curriculum that starts by teaching Scheme and ends with a segment on “OK, now you know how to program in Scheme, here is how to do it in Java”. I don’t know if that can fit in one high-school class, though.
This isn't quite right. The teaching language was ABC, a different project that Guido worked on. Python was his own language, created for the Amoeba project (an OS research project), inspired and influenced by, but not the same as, ABC, which was not object oriented. Guido says that the purpose of python for the Amoeba project was to: "bridge the gap between C and the shell"
http://en.wikipedia.org/wiki/ABC_%28programming_language%29
http://python-history.blogspot.com/2009/01/personal-history-...
However, if you have additional information, I'd love to hear about it. I've heard the "Python was originally a teaching language" several times, but as far as I can tell, that's not actually true.
I don't think there was ever any of the boilerplate required to make a Java program actually run. When I took the exam, I'd never compiled a single line of Java in my life (yep, there are far better ways to get things done).
(No sooner than that though, or they'd have to have announced it already; major changes in public education need a LOT of lead time.)
However, I'll disagree that Python would be that much easier than Java for a beginner's course. It's lower-friction, to be sure, but the VAST majority of student difficulties are with precise logical thinking, not with the syntax.
Decomposing a problem into discrete, correctly-ordered steps is the core difficulty. This is unrelated to syntax.
There is now a teaching certificate in just "Computer Science", although it wasn't offered in Texas for a few years while they were switching from one to the other.
Other states vary widely, I'm sure.
It seems that you would at least mention loops early on, to tell them that repetition is something a computer is very good at.
What if the Kahn Academy is teaching the concept in a more clear, concise way than you are? Maybe Java isn't the best language to teach these concepts? Maybe your lessons could be more effective? I don't know what your particular situation is, but I encourage you to play around with your lessons to see if you can teach it as fast as the Kahn Academy. Then again, they could have it totally wrong.
I've taught literally one thousand teenagers to program, face to face. I'd be shocked if Salman Khan had taught even a couple dozen students face to face.
Edit: I accidentally submitted this too early, and then replied instead of editing. See my reply for the rest of this thought.
Then, have all 200 kids attempt to code FizzBuzz on their own, with no assistance. I think probably 60-70% of my students could get it working. I suspect less than 20% of Khan's would have understood/retained enough to get it done.
You are in the 20%, so it's hard for you to see the difference. Salman Khan is in the 1%, and he hasn't been in the trenches long enough to see the difference.
That's what I'm saying.
That said, I have always wondered how effective video is as a teaching tool for programming? I guess when I think about it, it's the same delivery tool as an in-person lecture, but in some ways, more effective since you can pause, rewind, etc.
See the comments on this article about how to do a Fibonacci recursive example (which seems to be well done; the teacher takes it as slow as humanly possible): http://www.khanacademy.org/video/recursive-fibonacci-example...
The top comments are from people not knowing the difference between the equality and assignment operators:
"Why do I get an error with a single equal, but not with a double? This is what hung me up for a while."
...I don't know the history of this particular user, it's possible he/she skipped a bunch of the basic lessons and decided to learn recursion...but I wonder if there's a better way to somehow make sure people get the basics down, in an online teaching environment.
Teletypes did have a one key, probably because they were all-uppercase, so there wasn't a lowercase ell available.
It seems Khan Academy has a great model for instruction and Code Academy has a great model for practice; it seems reasonable to posit that a combination of the two would be superior to either one on their own.
I would further posit that the blend between the two models of instruction and practice would be something that might be tailored to the learning style of each individual student. Some people learn well by understanding before doing, others by doing first then understanding.
It's a lot better than some of in-person lectures I've taken, but maybe that's just me. I've been encouraging my peers to use online videos as a source to learn stuff they have a hard time grasping in class for quite a while. They don't listen, but that's another matter...
Anyways, I'm glad Khan Academy is coming out with this. I was freaking out when I couldn't solve some project Eular questions, and pretty much lost all of confidence in my programming abilities for a while. Online tutorials helped me a lot, and I think I've become a better programmer because of that.
One small nitpick: I haven't seen the videos yet, but it looks like they'll be in python. I'd rather it be C or C++. From what little I know of the language, it's pretty easy to pick up, and doesn't put pressure on the programmer to take care of details like memory management etc. I know for a fact that many algorithms that I've implementer in C++ that take over 100 LOC, I've found solutions online that have the same algorithm in 20-30 LOC. Used to irritate me a lot. :P
As experts, syntax doesn't annoy us. We see past the curly brackets (or the parentheses), because we can quickly see the internal AST that derive from them. Beginners can't.
Now, it would be wonderful to teach them a systems language that forces them to manage their own memory. I think a good candidate would be to keep most of the semantics of C, but simplifying its syntax (and find a solution to simplify includes).
There is a crucial point that should not be ignored. The first two languages someone learns should be easy. Monstrosities like C++ give the poor beginner the impression that learning any language is hard to learn, and that would discourage such learning in the future.
I disagree 100%, if you're talking about a college curriculum. The intro class at my university [1] used C, and I'm incredibly grateful. If I hadn't been forced to grok pointers and malloc, I'm not sure I ever would have (I haven't done any C programming since then). But even though I'm not writing C directly, knowing how things work at a lower level has helped me understand all sorts of concepts in higher-level languages.
[1] http://cs50.tv
I don't loathe C nearly as much as C++. I understand that learning about low level is useful. Pointers are indeed a very important concept (though not so difficult once you understand that variables themselves are already a form of indirection). So is Manual memory management.
But if you are tempted to put C++ in the same basket, don't. C++ is a very very different language from C. Sure, it looks like it just has some additional features here and there, but those features make C++ practices very different from C ones.
I've read way too many people online suggest Project Euler as a way to "teach yourself programming". This is almost a sadistic suggestion. Project Euler is way too hard to teach yourself programming.
Evidence: nearly 180,000 people have solved problem #1. Only 76,000 have solved problem #10. Losing 60% of your audience by problem #10 is pretty bad question design for learning.
Anecdote: I've been programming for over 25 years. I was near the top of my class in a top-10 computer science university program. And I've solved 30 Project Euler problems in the last 3-1/2 years. But I still can't solve problem # 18. I've worked on it off and on since December 2007. Only recently, quite by accident I learned that it is probably a "dynamic programming" question, which is something I've never done before.
Never lose confidence in your coding ability because Project Euler stumps you now and again.
I've dug up a lot of papers on effectiveness of teaching strategies in physics in particular (since I'm a physics student), and it seems in physics that your time spent working practice problems and making predictions about experiments is more important than actual lectures.
An example:
http://www.sciencemag.org/content/332/6031/862.abstract
If anything, I think the Khan Academy is best suited for a mixed approach where students watch lecture videos, then attend a problem session or help session to work examples, get help on exercises, and so on. The Khan Academy's exercise framework is an important step in this direction, and I hope they continue to improve it.
So, between three of us, we saw that simple line of code in three different ways. I wonder what other people see in more complicated code.
Not correctly understanding assignment is one of the biggest failure points for beginners learning programming. The '=' operator has little relation to '=' in any form the students have encountered before. It means something like "evaluate the shit on the right hand side, and store it in a variable with the name on the left hand side". That meaning is not really even close to the meaning of the english word "equals".
You can personally call it whatever you want, but if you want to be a good teacher to beginning programmers, then you don't call both = and == the same thing. "x gets 5", "set x to 5" or "x is now 5" are all acceptable, but "x equals 5" is not acceptable for teaching purposes.
x <- 5;
I agree with your main point, though, that failure to understand assignment is one of the biggest failure points for beginners. What programmers need to be able to do is trace through the execution of a program in their head, thinking "once this statement executes, the state will change in so-and-such way."
If I had to teach introductory programming again, I would focus more on having people fiddle around with variables in a REPL, just to get a feel for that sort of thing.
"David Gries recalled a time when he was lecturing with Dijkstra in the audience when he read "x=a" as "x equals a". This is of course a short hand for "assign a to x" and not a statement that x really is logically equal to a. From the back of the room Dijkstra shouted "becomes" and the room fell silent. Gries thought about it for a moment and said, "Thank you Edsger, for correcting me. If I make the same mistake again, stop me." Everything went fine until twenty minutes later he made the same mistake when the shout from the back of the room came again - "becomes". Gries once again thanked Dijkstra and claims to never have made the mistake again."
The first language I was taught was Object Pascal, and the assignment operator ":=" was pronounced "becomes equal to".
The only thing I don't like about video is that I can't do search for it. Sometime they show you a piece of code and you don't remember exactly where, you can't easily go to that part of the video.
http://scratch.mit.edu/ http://info.scratch.mit.edu/sites/infoscratch.media.mit.edu/...
Then they work on doing something interesting while I answer questions and ask "What happens if...".
To me the biggest hurdle in programming is not the programming itself but the sys-admin effort to setup a development environment.
Compare that to the Firefox download page, where it auto-detects everything it needs to know about your computer and gives you one simple, unambiguous giant "download me" button that just works. If this was a conversion funnel, python would be losing most of their money (i.e. potential users) at this step.
The old cruel IRC days of RTFM made me a better, more independent problem solver.
Click on Python.
I'm not sure how many of the Khan Academy tutorials you could get through...
I'd say stick with something a bit simpler, especially given that it's Khan Academy, not Stanford, so it's likely to attract far more casual lay-users who just want to know "a little bit about programming." The casual nature of the engagement means there is probably a very low threshold for quitting if the program is resulting in more confusion/frustration than "aha moments."
I think it's great they're doing it, tho! More people should learn "a little bit about programming."
In fact, it may be what holds us back from genuinely better paradigms.
I realize that this is meant to be an "intro to CS" class and not a web dev class, but a product can teach me how to create a simple web app that connects to a popular API, retrieves data and displays the data on a web page is the product that will get my money.
For Python and Ruby, it's best to have a handle on the languages first before throwing in Web stuff. That's the case for two reasons: first, particularly with Python, the web might not be the student's ultimate objective. But more importantly, you want to start in manageable chunks and throwing in web stuff can muddle things early on.
Which is part of why PHP is so popular, especially among people with more of a web background and less of a programming background. I agree that it's better to have a handle on the langage first, then throwing in web stuff, but if you take that tact, many people will instead use the language that starts with Web stuff, then adds in programming.
You can see him talk in person in this TED talk: http://www.ted.com/talks/salman_khan_let_s_use_video_to_rein...
I know that's not as cool as contributing to Khan Academy, and makes it harder to get an audience, but it's something you can do at any time, without asking permission from anyone.
Khan does not like all this meetings and talk stuff. He want people to do the same thing he did. Just start and talk later.
Does anyone have an opinion on whether Ruby or another language is better suited for teaching basic programming concepts?
Initially, the goal should be to learn primitive data structure concepts and usage(i.e. ints, char, strings, pointers, etc.). After that, they should learn more complex structures, like linked lists. These should be taught without referencing the built-in standard library structures. After that, simple sorting algorithms can be taught(including recursion), as well as OO concepts.
Teaching all of that in a GC language is nice, because it removes a lot of the memory management that starting out programmers always get caught up with. It lets the student concentrate more on the concepts than on tricky memory issues. That can be taught latter.
They found that students who were taught Python instead of Java scored higher on their tests. Languages like Python and Ruby make it far easier for people to jump straight into the code. If I'm learning about linked lists why do I need this main method? Ruby and Python have lots of abstractions which make it easier for someone learning to be able to focus on the problem at hand.
As someone taking a data structures course in post-secondary school where the facilitation language is Java, I've struggled more with the Java development environment than I have with actually understanding the data structures and the implementations. Even the Java specific implementations are simple once you can get your package and classes playing nice and compiled to bytecode.
This is not a slam on Java, just an observation from personal experience that some interpreted languages like Ruby and Python tend to make it easier to do casual programming exercises or tasks than those needing any compiling prior to execution, such as C, C++, and Java.
The biggest issue I've faced with my courses where Java is used to teach the material is having to deal with Java IDEs and boilerplate code required to test tiny bits of code and the resulting behaviour. Interpreted languages or those with REPLs tend to not have this problem from what I've observed.
Note: I know about BeanShell, but I've had poor results with it. Maybe my environment is at fault, but it seems to miss functionality as basic as command history, which makes editing commands very tedious.
Pretty much. What really needs to be taught is that programming languages are tools. There are pros and cons to every language, so the best language for the concepts being taught should be used. A student that's come out of a good CS curriculum won't be a great programmer immediately after graduating. However, they'll have all the tools that they need to learn how to become a great programmer.
However, if I was just hacking around on my own, I'd probably go with JavaScript just because of how easy it is to make something cool and show it off.
Computer science is a little different, because people have been learning CS online since well before Khan Academy.
2. Click subtitles.
3. Click the subtitle widget.
4. Click "subtitle homepage".
5. Hover over "download subtitles".
6. Click txt to download a plain-text transcript.
I'm pretty sure most of the Khan Academy videos have subtitles (at least English ones).
Disclaimer: I work for universalsubtitles.org, the site that handles all the Khan Academy subtitles.
I'm gunning to compete with my own platform, but he's in as a non-profit which means a for-profit doesn't stand a chance, especially with the timescales the sector operates on.
The only way to create something to compete and light that fire under their ass would be a "generous donation". I think I could do it though, within a year tops, but unfortunately, I need to eat...
http://www.aleks.com/about_aleks
...but I'm not sure. The Khan Academy exercise framework is also open-source, so presumably you can contribute that way.
Try this link: http://www.carnegielearning.com/higher-ed-solutions/cognitiv...
As yet, there isn't a way to replicate that in software. One day it will be, but until then, khan academy and those like it can't even come close to having a good teacher.
That said, I can easily see how it could take the place of a bad teacher.
EDIT: I do see now that the CM software you linked to does have a human component. That's an interesting aspect and I'm going to have to take a careful look at this.
Not sure what you're using to judge "too long to roll out fully," but we're running pilots in tons of classrooms right now where students are using Khan Academy to move at their own pace. We have 1.75MM interactive problems done per school day. We're using our own data more and more (http://bjk5.com/post/10171483254/a-bingo-split-testing-now-o...) to make the process more adaptive and engaging to individual students, and we've added hundreds of interactive math exercises (along w/ tons of new features: http://bjk5.com/post/8826207372/khan-academy-internship-summ...) in way less than a year.
We're nowhere close to where we need to be, and the possibilities for adaptive learning with all of our data make us all salivate for more time. Regardless, I can definitively tell you that every day around here is energized like crazy. We don't lack the fire.
Second, "he's in as a non-profit which means a for-profit doesn't stand a chance" is completely nonsensical. If that were true there would be no such thing as a for-profit. For one thing, at any sort of scale, the difference between a "non-profit" and "profit" can be as little as 10%, which is approximately the average profit made by Fortune 500 companies, within a percent or two.
Besides, I think that in the future of distributed, decoupled learning, there's room for many successful curricula. It will eventually be normal to get your history from X and math/physics from Y and English from Z, or even mix and match math modules. If you think you can compete, go for it.
http://www.khanacademy.org/video/npr-story-on-ka-los-altos-p...
Even the MIT "intro to programming and computer science" class (which uses Python) doesn't touch OOP until lecture 14, after dynamic programming, divide and conquer, bit of runtime complexity, recursion and more...