Log in

Previous Entry | Next Entry

Achtung graphics and database students!

banazir: Euler went blind, you know.
zengeneral: I plan to become deaf, so that I can ignore the cries of my students.

Don't make me break out this icon! ;-)

Actually, I've had a generally positive impression of students this semester. There are a few who are having difficulties with CIS 560, which is being taught as a heavy-duty relational database course again for the first time in several years (perhaps 3 or 4). Maria Zamfir-Bleyberg had this course exclusively until fall, 2000, and she taught it as a pure RDB course with a bit of SQL, a lot of relational algebra, relational calculus, and normalization. I remember how she found some students unable to grasp normalization, and had to cover a lot of what she and Anindya Banerjee call "remedial set theory".

From personal experience, I've encountered resistance to the "power-levelling" approach that instructors employed to get students through normalization and concepts of algebra and set theory when I was a grad or undergrad. Naming no names, most CS students and the stronger IS students have little trouble with it, but the ones whose discrete math foundations aren't 100% fresh seem to be struggling. This, along with a desire to improve the undergrad curriculum, is why I posted my straw poll about discrete math. Some have likened the course to a graduate course (though I took this actual characterization with a grain of salt, since IS majors hardly take any grad electives). They seem to be getting the impression that I'm "force-feeding" the material, when it's actually to get ahead of the problem and see what I need to cover remedially.

My solution has been to scale back the workload a little, but I think students are still nervous about our being 7 chapters into the textbook at a point when they would like to be 3 chapters in. Despite my reassurances, I think there's still an uneasy feeling that I'm going to try to cover 25 chapters instead of 15. Actually, anyone can cover the last dozen chapters of Silberschatz et al. on their own; it's the first 7 chapters that need to be taught 2-3 times, and that's why I've "breezed through" SQL.

Of course, if you listen to zengeneral, anything over 2 lectures spent on SQL is too much. ;-D



( 15 comments — Leave a comment )
Mar. 6th, 2006 09:57 pm (UTC)
That icon made me lol.
Mar. 6th, 2006 11:37 pm (UTC)
Me too :)
Mar. 7th, 2006 09:13 am (UTC)
Well, glad you enjoyed
There's plenty more Bitter!Elrond where that came from!

Mar. 7th, 2006 12:12 am (UTC)
Laugh it up, temporal tour guide
That was the idea. :-D

Mar. 7th, 2006 03:18 am (UTC)
one lecture on SQL would be too much.
Mar. 7th, 2006 09:23 am (UTC)
Textbook allergies and the perils of a little knowledge
Well, equally, you could say that CIS 560 is a superfluous course to a proper Math 510 and CIS 501.

That might be true (I never actually took a DB course as an undergrad and I did all my reading on my own as a grad and as a faculty member, without ever so much as auditing a DB course), but I wouldn't go that far. Where would it end? Rule the IS major superfluous? (I disagree; even the MIS major, which produces many service personnel for IT, serves an important purpose.)

I will say this, though: if one requires more thaneven one lecture beyond the relational algebra to understand SQL, it could mean a textbook allergy, which is a problem. It's like lecturing about BNF or precedence of various programming languages; who actually does? Not even in a compilers course did I ever get one lecture about BNF.

"Look it up" was my professors' favorite phrase, and we did. We learned how to look up things they didn't actually assign, either, because we grasped fairly early (2nd-3rd year) that required texts are a complete artifice to lull people into the complacent idea that a single book can encompass the subject of any course. People don't often read the rest of Alexander Pope's little poem "a little knowledge is a dangerous thing", and they should. It says, paraphrased: if you don't plan to keep reading, you shouldn't crack open the first book.

Mar. 7th, 2006 06:58 pm (UTC)
Re: Textbook allergies and the perils of a little knowledge
I never actually took a DB course as an undergrad and I did all my reading on my own as a grad and as a faculty member, without ever so much as auditing a DB course.

Ditto. All my DB knowledge is self-taught, from reading books and such to pick up what I need to know for whatever project I'm working on.

One of the more unfortunate gaps in my knowledge. I wish there were a way to take all CS courses, lots of electives, and still finish undergrad in four years! (I mean, without dying from course overload)
Mar. 7th, 2006 03:53 am (UTC)
Hah love the icon! And I have many teachers who would love to be able to turn off their ears for a while.
Mar. 7th, 2006 09:51 am (UTC)
Gene pool's running low on chlorine!
Haha, I love your icon, too - did you design it?

Thanks for a laugh. Please feel free to take my icon if you like, or pass it along to any teachers.

Mar. 8th, 2006 12:33 am (UTC)
Re: Gene pool's running low on chlorine!
Nope, it came from a person who didn't know who originally created it...go ahead and use it too if you'd like.
Mar. 7th, 2006 06:51 am (UTC)

Maybe I'm weird...

But I totally hated my database class. :(

But I'm an undergrad.
Mar. 7th, 2006 06:51 am (UTC)
PS your icon made me lawl too.

Mar. 7th, 2006 09:46 am (UTC)
SQL and undergrad database education: the power of plasticity
"Mikey, you're not supposed to like it." Seriously, databases is not a fancy, nor a pretty, nor an entirely pleasant subject. I avoided it for a long time in the late 1980s and early 1990s because it seemed to be about automating things in a way that emphasized scalability over efficiency and brute force over elegance.

Actually, that's not the case at all, as I learned much later, but relational algebra and SQL are like Othello: a minute to learn, a lifetime to master. Dr. Mehdi Harandi, one of my professors in grad school and later a Ph.D. committee member, once said of DB grad researchers that "they know all about the substance of relationl algebra and how to do extremely esoteric and complex things, but ask them how to do something very basic and simple and they hit mental blocks".

I personally find SQL counterintuitive, even with a fairly broad and deep background in logic programming, declarative and procedural semantics, and set theory. The way I view it is that SQL is to databases what Prolog is to theorem proving and what Jedi training is to Force-sensitive Rebels: the more you learned the conventional way, even using the new tools, the more you have to unlearn. The imperative programming paradigm makes us "fussy": we worry about how something gets done, how it is executed, and we can't stop, even if someone tells us that a query engine or an inference engine is taking care of things for us. Many of us who came in late are allergic to the homonculi of declarative semantics just as (or perhaps even more than) we can be resistant to the applicative semantics of functional programming. I get a lot of people who pick up a functional langauge such as Objective CAML or SML as their third or fourth programming language, and it's like watching a girl or boy who starts gymnastics or figure skating at 12, chess at 15, or foreign languages at 18: most of the neural plasticity is gone, no matter the physical vitality and flexibility.

My point is: rejoice that you learned things as an undergrad, even painful things, especially painful things! Even if you hated them, you have a better chance of making them second nature and "powering through" to more interesting and profound concepts. Accept and exploit that advantage. There's a reason why Heralds of Valdemar are chosen as younglings, why Jedi are normally only taken as infants, why Klingon kids practice with batlethmey from the age of three. It represents the power of plasticity.

Mar. 7th, 2006 07:00 pm (UTC)
Re: SQL and undergrad database education: the power of plasticity
The way I view it is that SQL is to databases what Prolog is to theorem proving

GASPPPPPPP. Actually, it's true. I love Prolog, but I've been using OTTER for a course this term, and I like how I don't have to make assumptions such as "where do I place my clauses", etc.
Mar. 14th, 2006 01:09 am (UTC)
Re: SQL and undergrad database education: the power of plasticity
Hmmm ...

This subject comes up in the usenet group rec.music.makers.piano occasionally (usually in the context of adult vs. child learners). I'm sitting on the fence because as an adult learner of piano, I have been able to keep pace with my teacher's child learners because I have as much time to dedicate to practice as they do. As this relates to CS classes, the same principle may apply; it is easier to learn something (assuming you are already equipped with the appropriate tools to learn it) when you have more time/less distractions.
( 15 comments — Leave a comment )

Latest Month

December 2008

KSU Genetic and Evolutionary Computation (GEC) Lab



Science, Technology, Engineering, Math (STEM) Communities

Fresh Pages


Powered by LiveJournal.com
Designed by Naoto Kishi