Slashdot Mirror
CS Students Want Advice on Helping Strugglers?
geekgirl21 asks: "My university's Computer Science Society (a student organization) wants to create a booklet about helping students get through the major's introductory courses (3 semesters of Java). The focus of this booklet is toward the people who work hard but still struggle with the courses. Specifically, we'd would like some advice about where to get the following: clearly written books/articles about the concepts of Java and OOP, how to prepare for tests, how to take notes in class, how to productively complete a coding assignment, and how to write good, documented code. Also, organization is not our group's strong point. So how do you recommend splitting up the work to complete this seemingly intimidating task? Thanks in advance about any advice you can provide."
You can always take the approach that has a very strong heritage in the Linux community, and has recently been catching on in some certain colleges: RTFM! READ THE FUCKING MANUAL!
Seriously, that is what some colleges are doing. You can be expelled for helping a struggling student; it falls under the same class as cheating or plaguarism. The idea is that they want to weed out all those who can't hack it on their own, so as to save them from wasting years of work. After all, programmers need to be able to get stuff done on their own, without asking for help every 5 minutes, right?
Well, I can understand this approach: programmers need to have some degree of independence. But everyone struggles with some things, and maybe this person has really great ideas for programs/algorithms, but just has problems putting them into computer language. It'd be a shame to waste such a person because (s)he couldn't get it right by rtfm'ing.
Also, in the real programming world, though much is done independently, no programmer is an island unto himself. No one can do everything, and the wise programmer seeks the advice of specialists in particular areas.
Personally, I think that you can only help other people understand something/solve something if you know how you do. You might solve a problem lickety-split. How? If you don't know how, you can't really help anyone. But if you can analyze your thought process and come up with exactly how you solved that problem, you can help other people think in the way you do about the problems (which is obviously a good way to think, since you solved it quickly). It might be a rigorous logical thought process, or a highly innovative one. Even if you are really good, you can benefit from this type of thing by understanding the thought processes of others who are really good, thus adding to your own thought processes.
social sciences can never use experience to verify their statemen
First most colleges offer courses in how to organize and prepare for tests. Most students should take this course or class if they can... No reason for your organization to bear the brunt of teaching someone how to study.
Second, the students will need an idea on how mathmatics and logic work, consider sending them to the math and physic tutors for the basics.
Third, how much of Java is about the fundamentals of OOP, more than likely if this is a basic programming class, the people learning it don't need to try to learn programming to start with, indeed they just need to get java to print and they need basic flow control... Trying to teach the oop and what it means at this stage of the game is a little premature... As long as they know how to get a basic program working, that's what they need to focus on at the moment... once they get the basics down, then they can move into thoery...
Fourth, where are your senior classmen, tutoring is one of the jobs of the senior classmen. So get with the tutoring department and get some of your senior classmen paid for tutoring students. A one-on-one education with someone that has been through this will help you to tailor your training in the future... Just make sure the tutors are giving feedback as to what they find students need the most help in.
/* TODO: Spawn child process, interest child in technology, have child write a new sig */
I recently read the first few pages of an introductory Java course taught at a University (too much wind, couldn't go snowboarding). It appeared to be the best introductory coding course ever written. It said, basically:
* Coding is hard. We don't expect everyone to be able to do it.
* You will only learn to code if you practice. It's like learning a musical instrument, there are no short cuts.
* You will still be able to work in IT if you can't code.
* Did we mention hard? Good.
Dave
I write a blog now, you should be afraid.
Don't help these people. Let them fail. Seriously.
1) There are already way, way too many bad programmers out there. Every profession has the standard bell curve of skilled vs. lame, but software is 10 times more skewed toward the lame end due to too many people dropping out of school to go work at a dotcom. Don't add fuel to the fire.
2) Every good programmer I know learned 90% of his skills on his own. Yes, you can learn things at school that are difficult to learn on your own, but quite frankly, these things shouldn't require special help to pick up (outside of normal coursework). If you can't figure them out on your own, your chances of applying them correctly are pretty damn low.
3) [and most important] Too many people go into Computer Science because the money's good, or because it seems like a better idea than the alternatives, or because they like playing Quake and they think that's what C.S. is all about. Then they get into their 3rd year and decide that they hate programming and switch majors. The sooner these people figure out what it is they REALLY want to do, the better, for them and us. It takes a special kind of person to actually ENJOY (not just tolerate) programming, and if you don't enjoy it you're going to regret making it your life's work.
On the other hand...
It's not like there is anything you actually learn in four years of CS education that you can't learn in 3 months with a pile of books.
I'd disagree. Going through those x years of higher education teaches you quite a bit. You may not remember half the stuff that prof tells you, but it is the other things that you learn that are _way_ more important.
Like how to teach others. You may be really strong in some aspect of a course that you are taking, so some of the other students ask you how it works/how to do it. The easiest way is to just give them the answer. The problem with that is that those people keep coming back looking for the easy answer, but if you spend 5-10 minutes initially explaining how/why something works, that student will (hopefully) understand how to do it, and won't come back. When you first start doing things like, you'll probably stumble a bit, and it'll take a while to explain things, but once you get the hang of it, then your teaching becomes much quicker, and much more effective. (a very strong asset for you to have)
Another thing that you will learn is how to ask questions, as above, you may just start out asking for the answer directly, then you start asking why, then (hopefully) after a while, you will be able to succinctly be able to describe a situation, the problem you are having, and what you have tried in a short order of time. (again a very good asset for you)
You learn how to deal with prick collegues, superiors, and aquaintances.
A lot of people will think that these are things that you can pick up in high school, or that you already know these things (they are pretty obvious), and yes that is true. But the concentration of these factors, coupled with the pressures of higher education really drive home these qualities, and it should be just second nature to you. So that it is just a reflex, not really something you have to think about.
yea, you could probably learn some or even all of the technical details in 3 months with a stack of books, 2 pencils (37 erasers!), and a couple computers...but you'd miss the whole point of University.
The basic sleazeware produced in a drunken fury by a bunch of UCBerkeley grad students was still the core of BIND. --PV
I have tutored any number of impressionable young girls in High School algebra and calculus--and I've come to some conclusions.
1) Choosing the right textbook isn't just important, it's damn important.
2) I could have been replaced by a tape recoder with these phrases in a loop: Well, what do you think the answer is? I know that you don't think that will work--but you can't know until you try. See, you are a very smart girl when it comes to mathematics, I hope you consider it as your college major.
3) A dry textbook is worse than a lively teacher and a lively textbook is better than a dry teacher. It's about motivation, not intelligence. In my experience, confidence in your ability to tackle problems (not necessarily confidence in your solutions, though) is the surest mark of genius. And motivation, unlike intelligence, is under your control.
First of all, it's not clear that these are necessarily computer science majors. Many majors in my school, for instance, require their students to take the intro CS courses but nothing further.
Second, you should probably learn some patience -- though this stuff may seem easy to you and me, just because someone doesn't "get it" immediately doesn't mean that they won't, eventually, and go on to be successful. I'm reminded of one of the hardest classes I've ever taken, an undergraduate course on functional programming and advanced programming techniques... I hated that class at the time, but now I do that kind of thing for my research, and it's totally comfortable for me. Sometimes it's just a matter of getting over that hump, and sometimes the inability to do that may be more a problem with the teaching method than the student.
First, I don't agree with those who said "sink or swim", but I don't completely disagree either. I'd say the same about any discipline. I sucked at biology (I am really bad at memorizing stuff) so I got out of that. I have seen many people really struggle, and some of them made it, some of them didn't. Some of those who didn't might have if they'd had help. It's up to them to ask, it's up to you to offer, but frankly if they are willing to put in the effort, nobody loses from helping them out.
I can't really help with the OOP stuff, because when I took CSc our introductory stuff was on procedural programming using Pascal (I started a four-year CSc program in 1992). I would say anybody who hasn't spent their teenagerhood dicking around with computers and teaching themselves rudimentary programming will have some difficulties in any case, but there's nothing like a first-year course to get you up to speed! There are plenty of online tutorials, some of them might even be useful, and I have heard good things about the Eckels book mentioned by others here.
To prepare for tests: same as any other course! Do your assignments, ask questions in class, and stay on top of things. In a well-structured course the labs will follow the lectures fairly closely but there's generally no hope of the lab giving you first-hand experience at all of the theory you will be studying. The best thing to do with unpracticed theory is to ask questions. It's the only way I remember theory I don't use immediately--think of questions and ask them. If you can think of an intelligent question, and it's answered well, I personally have a much better chance at remembering the context.
Also, when you're actually taking the test, do everything they say: read it over first, check the number of pages, have a bottle of water or something, take a handkerchief if you have a cold (heh), do the stuff you're confident with first to build momentum and just do it, double-check, but don't double-think, your answers.
Taking notes is something I never mastered. But if you read the text in advance, if there is one, you only need to note the differences. But I found most profs handed out overheads or printouts or whatever. And in first-year CSc, I don't think you'll need much note-taking anyway--you need to do the lab work more than anything else.
How to productively complete a programming assignment? Do the absolute requirements and unless it's a multimedia course don't bother with fancies. It'll be marked by senior undergrads or grad students who will be bored with 100 assignments demonstrating recursion to solve a Fibonacci sequence. By the time you get to larger projects, hopefully you'll have a good idea of how to productively complete assignments, but at first it'll be your first skate on a frozen pond.
To write good, documented code, I would think any first-year program at a half-decent university would teach this. But, if they don't, I'd recommend starting by commenting every single line of code you write. It's better to start with too much than too little--you're less likely to get dinged for it and that way you'll get used to commenting code as part of the process. You might want to use tools like JavaDoc or Doxygen, but never EVER forget to document the code itself, because if somebody wants you to update legacy code later on, that might be all that's left.
Good luck with your booklet! :)
No, its the equivalent of saying "some people simply cannot learn to drive without someone showing them how and riding along with them at first" which if you ask me if perfectly reasonable.
I wasn't suggesting that it should be alright for people to not be able to learn from a book at all, but rather that there are quite a few people that cannot learn from only a book. The value of a good teacher cannot be underestimated.
The job of the person at the front of the classroom is not to digest knowledge for students, because it can't be done.
No, their job is to figure out how a student best learns, and translate the material to be learned into a form digestable by the student. Whether or not the student actually digests the material is a function of how much he/she wants to learn it.
The learning process involves exactly one person, the student. The only one who can teach you is you.
The learning process invovles two people, the student and the teacher. You need both a motivated student, and a good teacher in order to learn.
...Was to have a mandatory 1 credit course for the seniors in the Computer Science degreee program, that required students to provide tutoring to others. Twice a week for two hours I sat with people that were having problems. Everything from teaching them to debug their coding assignments (not doing them for them), to explaining concepts covered in class.
Frequently, the problem was just a small missing piece in their understanding. It was great to see the light come on in their expressions when they got it.
I enjoyed the experience a great deal, and now 25 years later, I still find it very rewarding and mutually educational.
Can You Say Linux? I Knew That You Could.
Why is this course so popular? First of all they choose the best professors and lecturers to teach it. They are able to explain things clearly and they are very personable/funny/approachable.
Second, the programming assignments are well designed. I always knew WHY I was doing a particular assignment. I have since helped people from other universities with their CS assignments and I can say that assignment design is a big deal. Some profs just throw some random problem at the kids without thinking about the purpose of it. They don't provide clear goals. They don't provide any libraries to use. The documentation provided with assignment needs to be clear and complete.
The most important reason that these classes were both popular and productive was the section leaders. These were CS students in their senior or junior year that taught section once a week. The position was prestigious and it paid well. They got some of the best CS students that way. It is describe briefly in this article. (Search for CS198 if you don't want to read the whole thing.) The sections were very productive. You would write code to do some particular task and the section leader would help/critique. There was almost always a section leader "on call" in a cube in the lab that you could go to for help on an assignment. More than one near the due dates. They could find a bug in 2 seconds flat. But they would just show it to you, they would "help" you see it yourself. But the most important thing they did was "interactive grading". They would go through a printout of your code line by line and write comments both good and bad while grading your program. Then they would have a 20 minute one one one session with the student and explain what was good and what was bad about thier coding. Whether the program actually functioned or not was of secondary importance, especially early in the quarter. They made sure that your code made sense, was well documented, that you had variable names that made sense, etc.
Many schools just give grades based on functionality and never even look at the code produced. This sink or swim attitude might bring to light a few naturally good coders. The mentoring aspects of the program at I just described produce a lot of great coders.
In summary:
1. Choose good instructors.
2. Give assingments with a purpose. Document them.
3. Give the students detailed feedback and lots of support early on whether they think they need it or not.
Lasers Controlled Games!
Java is not an appropriate introductory language. There are basic concepts you should learn before you start to use java. You should learn to use simple data structures, and learn procedural programming and design before you get confused by the syntax of object orientation. You should be learning how the computer works at a low level before you are expected to understand high level concepts and where it is apropriate to use them. If you start the students out with java, which may seem simple to people who already know how to design programs, the students won't understand why and why not to do certain things. They will be forced to learn through trial and error what they should have been taught to begin with.
Instead of putting your effort into helping people who have never programmed before to use java, spend your time convincing the school that the first two semesters should be basic programming with some C and some assembler (not as classes that teach the languages, but using the languages as necissary tools to demonstrate the concepts. In early courses, syntax should only be taught when it is necissary to implement a concept), and a data structures class where C pointers and structures should be introduced moving on to java towards the end. Guaranteed, after the students learn basic concepts well, they'll have no trouble picking up java. They'll also write better code for their entire career.