Insightful Report on the State of AP CS in California

January 16, 2017 at 7:10 am 14 comments


Insightful new report from ACCESS-CA on who is taking AP CS in California and on the challenges (quoted below):

Despite the strong outlook for the technology economy in California, there are major challenges in meeting the growing demand for skilled technology workers and preparing Californians to participate in the workforce of the future:

The lack of computer science standards, courses, and teachers and the lack of alignment between computing pathways and workforce needs. Roughly 65% of high schools in California offer no computing classes and the state has yet to develop a statewide plan for computing education.

The lack of diversity in the computing education pipeline and within the technology sector, particularly given the rapidly-increasing diversity of California’s population. 60% of California’s student population is Latinx or African American, yet these students comprise just 16% of students taking AP CS A and 15% of the technology workforce


Entry filed under: Uncategorized. Tags: , , , , .

California developing Computer Science Standards Power law of practice in software implementation: Does this explain the “W” going away?

14 Comments Add your own

  • 1. alanone1  |  January 16, 2017 at 7:35 am

    Not just to be tiresome — but it does need to be pointed out over and over until actually generally comprehended: the notions behind the AP curriculum and testing are quite terrible in my opinion. They have very little to do with the spirit of computing, and of learning computing.

    The notions behind “coding for all” are not as hamhandedly stupid, but they also quite miss what computing should be about, not just today, but especially in the not too distant future.

    Let’s put a lot more effort into getting things better before unleashing more of the dogged, blind and heavy state educational apparati to engage in this mess.

    I am quite certain as a high schooler these days I would not go near any kind of official computing — or perhaps any at all. This is the ultimate in making a sow’s ear out of a once beautiful silk purse!

  • 2. gasstationwithoutpumps  |  January 16, 2017 at 4:47 pm

    While I agree with Alan that the current AP curriculum is rather clunky, I think that we’d be better off trying a thousand different things rather than waiting. That is, I’d like to see the state put money behind having CS in high schools without a specific curriculum. The result will be that 90% of the courses will be crap, but a mandated curriculum is likely to end up with 95% crap. And even a crap programming course may be better than no exposure at all, in getting students interested in learning the subject later on.

    Huge numbers of students in CA want to study CS in college, but the variation in prior preparation is enormous. The two extremes are students who have done a little typing in Word with not much else and students who have had the equivalent of half a computer science or computer engineering degree.

    Needless to say, the difference in preparation depends heavily on the wealth and education of the parents, with first-in-family students having almost no preparation and children of engineers and programmers having much more substantial backgrounds.

    • 3. alanone1  |  January 19, 2017 at 12:28 am

      They are more or less following your suggestion in the UK. This is leading to the kind of pop culture — perhaps “Guitar Hero”? or is it “three chords”? — approach that the -coding for all- movement seems to be doing. But I think I agree with your idea that “only” 90% will be crap rather than “95%” (and maybe less debilitating crap at that).

      I most definitely agree that no exposure is likely really bad, but I think there could be more choices than just “none” or “a random course”.

      One way to look at the nature of the beast — maybe the “problem” — is that the -synthetic- and -huge degrees of freedom- and -processes- and most especially the -language machine- nature of the medium means that -design- is where learners need to get to — and to get there without being messed up by a myriad of technical details that can swamp learning to design in favor of just learning patterns and responses as a 21st century catechism.

      I think that it takes at least two years of “doing helpful things” to get perspectives for the start of good design thinking, and that a real curriculum needs to be organized to enable this pathway.

  • 4. gflint  |  January 17, 2017 at 11:01 am

    After looking at the APCS curriculum I would not take it either, and I like programming and CS! CS and programming can be made fun and relevant and still prepare students for a college CS track. The report proves that students do not like APCS (not CS in general) and that maybe there is a large shortage of teachers for APCS (as opposed to low level CS). Using APCS as a measure of CS in the US is like using the sales of Tesla cars to measure the market for overall car sales in the US. As soon as something is labeled “AP” it eliminates a huge percentage of students purely by intimidation. Teachers are also reluctant to teach AP simply because of the attached bureaucracy. Offer a state wide low level course titled “CS and Game Making” and see what happens to CS enrollment numbers.

    • 5. Mark Guzdial  |  January 17, 2017 at 12:51 pm

      Sure, but do you have a better measure of CS education (which students get it, who doesn’t, and how they’re doing) at a state or national scale? And do we know that APCS is that far off? For example, is there a state where participation in education (in terms of number of students, number of teachers, or diversity) is ten times better in non-AP CS classes than in AP CS? If we want to know where we are wrt CS education, looking at APCS is a reasonable measure for today.

      • 6. gflint  |  January 17, 2017 at 5:26 pm

        Is there a measure for anything in education that does not have a standardized test? APCS has a standardized test therefore it is the measure of CS in the US. Those numbers are an easy target. Is there a better method? Probably the percentage of students taking CS/programming in a school that offers CS/programming. The trouble is that number would be a real pain to get. The fact that APCS is a measure, and perhaps the only measure, does not make it a good measure. A bad measure, even if it is the only measure, is still a bad measure. As far as I know my state of Montana put out zero APCS tests last year. Does this mean there is no CS in Montana? I know of several Montana school districts that are working hard at expanding their CS courses. They just do not include APCS therefore the numbers do not get into national stats.

        The whole numbers thing can get a bit tricky, especially if the government gets a hold of them and wants to prove something with them. Using APCS numbers as a picture of CS in the US is just asking for trouble.

        Maybe the new president will tweet about it. Right now as a CS teacher I feel like a long-tailed cat at a square dance. I just do not see anything good coming out of the next four years and extremely dubious numbers are not going to help. The only hope is if the number crunchers realize the vague nature of the numbers.

  • 7. Bonnie  |  January 18, 2017 at 8:12 am

    Can I be the dissenter and put in some positive words for APCS? My oldest kid is taking it right now and not only am I pleased by what I see him doing for the course, but he really loves it too. More importantly, now that high schools in NY are finally adopting APCS, we are starting to see kids come into our program having taken it. We place them in our CSII course, and they are amazingly well prepared, sadly probably better prepared than our students who go through our own CS1 course. We are considering whether we should move the APCS completers right on to our CSIII course. And the best part, and this is anecdotal only, but we see more women in the APCS completers than we do in our general incoming freshmen.

    • 8. alanone1  |  January 18, 2017 at 8:30 am

      Did I miss something? Isn’t this course still basically a Java course?

      • 9. Mark Guzdial  |  January 18, 2017 at 3:57 pm

        It’s a bit more than a Java course, but remember that the point of AP CS is that it looks like what most schools teach in CS1. AP CS is meant to be equivalent to a college-level introductory CS course. The College Board regularly surveys computer science departments about what they teach in CS1 and whether AP CS looks like their intro course. The AP CS is overwhelmingly approved of by CS departments across the country. Not all, of course, but the majority of CS departments teach an introductory computer science course that is like AP CS.

        Our analysis at Georgia Tech (done years ago by Allison Elliot Tew) matches Bonnie’s experience. APCS students who do well on the exam do better in CS2 than the students who take Georgia Tech’s CS1’s (any of the three).

        I agree that it’s not an accurate model of what is CS. It is well matched to undergraduate computer science in the United States.

        • 10. alanone1  |  January 18, 2017 at 11:52 pm

          OK, well you got me to go to the website and download the course description, etc. It has been updated since the last time I looked at it, but still seems quite terrible to me in almost all respects.

          And, yes, I truly hate what UCLA — a top 10 school they say — has for its first course.

          Beyond the very poor notion of content for computing, there is perhaps even a worse approach to pedagogy and what is known about how human brains learn best and the barriers of various kinds of cognitive load place on learning.

          I think most of the field in general, especially in universities, and most especially in AP, has renormalized to a “budget of unfortunate ideas” to the extent that they cannot now be seen.

          • 11. Mark Guzdial  |  January 19, 2017 at 11:57 am

            I completely agree, especially with:

            Beyond the very poor notion of content for computing, there is perhaps even a worse approach to pedagogy and what is known about how human brains learn best and the barriers of various kinds of cognitive load place on learning.

            I’m involved in a cross-university effort at Georgia Tech to change that, but it’s depressing how few levers we have. How do we convince university faculty (professors, lecturers, instructors, whatever) that the way they teach is less effective than it could be, that the way that they learned is not the best for everyone, and that we do know better ways to teach?

            • 12. alanone1  |  January 19, 2017 at 12:25 pm

              Hi Mark

              It won’t make you feel any better, but the situation in K-12, especially 7-12, for Biology is at least as bad.

              Bruce Alberts — writer of the great book “Molecular Biology of the Cell”, and for 12 years head of the NAS — has been campaigning at every level with not a huge amount of success. Here’s an editorial he wrote for Science:

              Click to access Failureofskin.pdf

              The main Bio textbook in high school and many colleges is Campbell, and it is an enormous mishmash of uncontextualized facts with essentially no science for the learners: essentially the stupendous amount of classifications and terms from classical biology multiplied by a factor four or more by the new terminologies of molecular biology. It’s the exact opposite of the “few deeply focused projects” that Bruce advocates. What they are actually doing is really crazy, but they are doing it with vigor.

              And one can imagine why people who don’t know what they are doing might opt for “coverage”. This leads to 1500 page textbooks in so many fields today, and superfestooned with pictures. But generally mishmashes. This is certainly true of the several “introductory texts” for first course in programming, both in C++ and in Java.

              And the experts are also a problem. We’d like to learn from the best, but they often “know too much”. One of the few slightly better Bio texts has many things in it that are good to know about science and Biology, but these completely miss the problems of cognitive load and the latencies required for getting fluent enough to get to the next fluency level. These books cannot be read “as books” by beginners. This forces them to have to learn in class, and this is not a good place to learn most things.

              Given our human finite internal resources for learning things, the learners need much thinner materials to work with, and teachers who know when they should shut up.

  • 13. gasstationwithoutpumps  |  January 19, 2017 at 12:05 am

    “I agree that it’s not an accurate model of what is CS. It is well matched to undergraduate computer science in the United States.”

    That is the sad part of the state of CS education in our undergraduate programs.

    Incidentally, did you look at how AP CS students did compared with matched controls who took your local CS1 courses (say, controlling for same SES). I suspect that the AP courses aren’t any better at teaching CS1, just more selective in who is allowed in (based mainly on the wealth of the school).

    • 14. Mark Guzdial  |  January 19, 2017 at 11:52 am

      It’s a great suggestion, but no — we had no access to SES. I’ll bet you’re right.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Trackback this post  |  Subscribe to the comments via RSS Feed

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 9,045 other followers


Recent Posts

Blog Stats

  • 2,024,541 hits
January 2017

CS Teaching Tips

%d bloggers like this: