Making a List, Checking it Twice

A while back the College Board dropped one of the Advanced Placement Computer Science exams – the so-called AB exam. (About 14 months ago and I wrote about dropping the APCS AB exam here) Ever since then there has been a lot of discussion about the AP computer science curriculum in specific and the high school computer science curriculum in general. The College Board is working on a ne AP CS exam. The current A exam is still going to be around and the word is that it is in for some minor changes. One of these is a possible change in name to “AP Computer Science: Programming.” Now cynics and critics (of whom I have been known to be one) might wonder out loud why not “AP Computer Science: Java” but that’s not likely to happen. In theory the language could change one of these days. But there is also a totally new AP CS course in the works. This one is likely to be called “AP Computer Science: Principles” though I don’t believe any of these names are set in stone just yet.

What is this new course going to be like? Well that is hard to say as the Commission the College Board appointed to work in it is just getting started. But if seems like “there’s interest in providing teachers with choices of what language they teach. ” Mark Guzdial who is on the Commission lays out some of the issues involved in a recent blog post called “Nudging Computing Education” In this post he says that an unlimited set of choices is unlikely. I have no doubt that he is right. He also talks about some of the pros and cons of some languages - C++/Java and Scheme.

Now all of these languages have their supporters and detractors. Personally I think anyone going into a career (academic or industrial) should learn C++ at some point. I am not a fan of it (or Java for that matter) as a first language. Likewise I know people who are having great success teaching computer science with Scheme but also know students who all but dropped the idea of taking additional classes because of bad experiences with that language. Truth is – no language is perfect. But most of us have our favorites. So now a pitch for my recommendation for a language to make the short list for the new AP CS Principles course- Visual Basic. Come on you knew that was coming. 🙂

On the pro side Visual Basic has an easy to learn syntax. many students find the semi-colons and curly braces of C-style languages and the complex nesting of parentheses of Scheme confusing and frustrating. That seems, in my experience at least, to be less of a problem with Visual Basic. Or other versions of Basic such as Small Basic for example. Then there is the matter of graphical user interface programming. Sure it is not by itself that important a concept but it can be part and parcel to students experiencing success and creating something that looks and feels “real” and relevant. Getting students interested and letting them have fun should be an important part of a first programming experience. With Visual Basic this is simple, fast and easy. Done right an instructor can keep students focused on code and concepts while the students enjoy creating software they can show friends. Plus there are features in the IDE like Intellisense that make exploration and self-learning easy. I think that is a plus.

Now critics will say that it is a proprietary language and only runs on Windows. True. But that still means it runs on most computers that most schools have. Cost is not really a barrier because the Express edition of Visual Basic is a free download. Students can even get the professional edition for home use for free through the DreamSpark program. And even if a teacher doesn’t want to use it that is not a good reason, in my opinion, from excluding it from a list of options for other teachers to choose from.

People might argue for a C-style first language because the APCS: Programming course is in Java. I have two replies to that. One is that after several years of starting with Visual Basic and following up with C++ (as C-style a language as it gets) I found that students with a good foundation of the concepts had little if any trouble picking up C++ from as inadequate a teacher as me. The other possibility is to include C# on the list of accepted languages.

Why? Well because C# has the same ease of creating GUIs that VB has. The same powerful and helpful IDE that VB has. And the colons and curly braces others seem to love and cherish. So that is two languages I’m suggesting. I suspect that there will be some strong support of Python as well.

Python is a dynamic language that is seeing increasing use in education. So the question for the commission is how many languages for their list.

I would argue that C++/Java/C# are all close enough that to include one means you can probably include all three. Given how many people are using Visual Basic for non-AP courses and the desire to see this new AP course reach many more schools than the present exam I think it should be included as well. Even if you don’t buy all of my earlier arguments. Python and Scheme? Probably or some people will scream pretty loudly. So does the list look like:

  • C++
  • C#
  • Java
  • Python
  • Scheme
  • Visual Basic

We could do worse. The one monkey wrench in this list is Scheme. While the other languages are procedural/object oriented languages Scheme is a functional language. Things work a little differently (some would say a lot differently) with functional languages. I’m glad I don’t have to make that decision or worse yet write the test that includes Scheme with the other languages. Not that that isn’t a worthwhile goal.

So what do you think? Any language you would add or subtract? Anything your seriously object to? Leave a comment or better yet hunt me down at NECC and give me a piece of your mind.

Technorati Tags: ,,,
Comments (9)

  1. Leigh Ann says:

    How can we choose a language without knowing what the programming goals of the course are?

    Lately I’ve become more and more convinced that we (as CS educators) take the wrong approach to selecting an introductory language.  We argue the merits of the language as a whole, with a small bit about its pedagogical importance.

    I think there needs to be some serious research here as to the aspects of an introductory language and how it affects students educationally.

    Anyway – without knowing what parts of "programming" we are teaching – aren’t we putting the cart before the horse?

  2. AlfredTh says:

    It’s funny you should bring that up. I was thinking the same thing the other day. I thought of this post first but didn’t get down to writing it right away. I realized that when ever someone asks about CS curriculum the talk too often turns to tools rather than concepts. I am as guilty of that as any one but how often have you heard someone just jump in with tool talk right away? Pretty offen I’d imagine. I feel another post coming on. 🙂

  3. Ian Bogost says:

    Visual Basic is a fascinating programming environment. As much as C has been a sort of lingua franca for programming over the past 20 years, few seem willing to admit that a huge number of computer programs used in business and creative contexts have been written in VB. And often, those programs are written by people who would not identify as "computer scientists."

    It seems like such a state of affairs does at least suggest taking VB seriously as a platform. I find the suggestion in this post compelling, or at least provocative.

    Perhaps if the CS world were more concerned with getting people to enjoy and use programming in their personal and professional lives, and less concerned with maintaining its major numbers and other meaningless statistics, we would all benefit.

    APCS aside, VB.NET can be used for Windows apps and for web apps. Microsoft is missing an opportunity by limiting their tools to their own OS. Object Basic environments like Real Basic and Blitz have long offered x-platform compilation (Win/Mac/Linux).

    A nitpick: while VB.NET Express is free, it still requires registration with Microsoft to activate. Some of us consider the act of giving our information to corporations an actual cost, not merely a trifle.

  4. Leigh Ann says:

    And why settle on one language?  If this course is supposed to be about principles of computer science why wouldnt you compare and contrast different language paradigms?

    How is Java/C++ different from Scheme and should you be able to write a simple instruction in each?  Isnt that a principle of CS?

    If we did get rid of APCS AB (the harder exam) – lets NOT add an APCS A "lite" with some extra ethics.  Lets give this exam some meat to make it worthy of college credit.  Why not make a list of the "Principles" of CS that would make this worthy of credit at major institutions.  I’ll be even with programming de-emphasized there are lots of other concepts out there begging to be covered.

    * Computability and the Halting Problem

    * Machine Learning (there are some great GUI interfaces like Weka that require NO programming but allow you to build complex data models)

    * Human Computer Interaction – what are some of the basic principles here? Could easily mock up interfaces and talk about design and interactivity

    * Discrete Mathematics – How does RSA work? What is encryption and why is it not REALLY safe?

    * Security and Privacy – I really think students need a much better understanding of this domain

    * Robotics – What are the current SOFTWARE problems with robotics?

    * Graphics – What is the difference between Vector art and traditional pixel graphics

    * Internet and Web technologies – what is a content management system?  What is a mash-up? What tools exist for creating them?  What is XML and why is it so powerful?  What does it mean to have a database behind a web site?

    These are our big ideas.  All of them are covered in depth in college level CS courses – why not change our idea of "101" to match other disciplines in a survey? (ie Bio or physics – you get lots of big ideas before diving into the specifics)

  5. Ian Bogost says:

    Leigh Ann, it seems like you’re right that all of these ideas are principles of computer science. Well, maybe not robotics, which seems like an application area (why not commerce? why not art?).

    Here’s a question I often wonder about:why is it that computer science is so undecided about what it’s core principles are? It seems unusual.

  6. AlfredTh says:

    Ian, One option is to down load the ISO image for installing all of the Express editions. This is designed for ease of install in classroom/lab situations and the copies installed from these ISO images do not have to be registered to activate. Little known secret. 🙂

  7. Spencer Brody says:

    There are a lot of great points here.  I like the idea of using multiple languages in this new course to expose students to a variety of languages and so there can be some discussion about the pros and cons of each language.  I actually think that Java should NOT be offered as an option for this course, so long as the APCS "programming" remains in Java.  I think Java is a great language and since it is so popular right now, it makes a lot of sense to teach it (and especially Object Oriented Programming) in high school, and that makes sense to be done in the "programming" course.  But for a "principles" course, there are many other languages that are much better to teach the foundations of computer science in, and students should be exposed to multiple languages to see how the different concepts and techniques show up in different languages.  Do you know which course is going to be presented as the "intro" course?  In my mind it would make more sense to start with the "principles" course and introduce students to programming with some simpler to pick up languages, then the next year – after the students understand the principles – introduce them to Java and OOP and get into more complicated programming techniques.  In that situation, I’d say start with VB for half a year, since its real easy to learn and its GUI interface allows students to create interesting programs they can show off to their friends fairly quickly.  After that, I’d say move to Python – its still pretty easy to pick up, but its also more powerful than VB and it’s growing rapidly in popularity right now and showing up in college intro courses more and more (along with Scheme, but despite my love of Scheme I think it may be a bit scary to highschoolers in their first programming course).  I’m actually not a huge fan of teaching C++ in high school… I agree that it’s a very important language that any programmer should know, but it has so many pitfalls and complexities when you are first learning about it that I fear would be too frustrating to someone just learning programming and would be likely to turn the student off to CS in general.  I’ve never used C#, so I can’t say much about it… but from what I’ve heard its very similar to Java, so I’m not sure what would be gained by exposing students to it.  In my ideal, the only languages offered for the "principles" course would be VB, Python and Scheme, requiring that 2 of the 3 be taught, and recommending VB and Python.

    I still say, however, that the hard part isn’t picking the languages, but deciding the curriculum of concepts that need to be taught and figuring out how to make sure they get taught well.  By concepts I don’t mean just things like graphics or security (though those are of course important) but also things like software engineering – teaching students how to think about code, how to design and structure a program so as to be clear, modular, efficient, etc.  I see very often people learning the syntax and knowing how to make programs that do what they want, but without understanding the value of abstractions and writing modular, reusable code.  I think this is something that is much harder to teach and measure, however, as well as to find teachers at the high school level who can really understand these concepts well enough to convey them to students.  Defining the curriculum in a way that is clear and covers all the necessary important topics is going to be really important and also really difficult, considering how abstract some of these notions are and considering Ian’s point that CS isn’t very unified in its idea about what the core principles and concepts of computer science even are…

  8. I’m a high school student taking APCS in my junior year. I don’t know what the curriculum will cover yet even though I begged the teacher to at least give me last years IB course but he didn’t. I actually pointed him to the CB notice that explains the discontinuing of APCSAB.

    When you say first language, do you mean the real _first_ one? If I taught an APCS class, it would have a class before it too, an actual introductory class. It would start with javascript because it runs everywhere. It’s easy and effortless and it looks pretty. It also has many computer science concepts like classes/prototypes or strict types/loose types. We’d move into Java or C# after that. I pointed my schools CS teacher to DreamSpart too, he uses Macs and is oblivious to everything else. I think C# is actually better because it can handle those graphics that teens want so much better than Java.

    I’ll be able to tell you guys what I think later into the start of the school year once I start that class. I know some programming so I’m not there for that, I’m there for the concepts which have eluded me in books and other places.

    I really want to learn about this but the CB didn’t exactly make it any easier.

    Have a good one.

  9. Myra Deister says:

    There has been many interesting points made.  The only caution I have is that at least at my high school, I have no funding.  In the Visual Basic class, I have been using Visual Basic.Net Express, but I don’t have a textbook because there are no funds to purchase one.  Most of what I do in my computer science classes is what I have managed to put together either free or I have financed it.  I would love to do more with robotics, but I haven’t found a way to finance it.  

Skip to main content