Mon, 31 Jan 2011
Is My Teaching Getting Bad?
The results of the final exam of the UNIX - Programming and System Administration I course make me wonder whether I am getting gradually dumb or bad at teaching.
I used to think that as I teach, I am getting to know more and more about the subject, as well as have less errors and more clean formulations in my presentations. My lectures are now available from the video archive, so the students can rewatch me presenting the topics they have not understood the first time again, if they wish. I have switched the presentation format from old and boring black-and-white DVI slides to new, colored, shiny and graphically rich LaTeX-beamer format (thanks to Jiří Boček from servistech). Also the exam format and even some of the questions are still the same as they have been since eight years ago.
So what am I doing wrong?
18 replies for this story:
Martiner wrote:
As a former (successful) attendee of this course I believe that you have to find better students :) You are talking about ratio, but the base is growing, isnt it? If the count of successful students is the same (or is decreasing slowly), I believe there is nothing wrong with you.
Yenya wrote: Re: Martiner
Well, quite the opposite is true - the total number of students attending the course is slowly decreasing.
Kurochenko Andriy wrote:
As for me, I liked your presentation and I appreciate lectures in the video archive. The new slides are colorful and contain more information, but are inappropriate for printing in contrast with the old ones (I also like them more because they look like man pages :-) ). As for decreasing percentage of successful students. Maybe C and kernel functionality are less popular among the students every year. Who knows?
sev wrote:
Please don't start Y-axis at 45% - this is confusing visualization. Ad topic: Do you feel like the requirements to successfully pass the course are higher?
Peter Kruty wrote: Some thougths
I find this really interesting. Maybe we are loosing the real geeks used to terminals and man pages and new generation is growing, which still is interested in Unix, but also are used to Web 2.0 simplification, to the point information and easy to find core information (a la wikipedia)? I have not seen new slides, but I guess that overall approach of the seminar is still same and I remember it as comprehensive, detailed and precise. Maybe new student have issues with digesting it? I'm wondering how they score in other courses in general? How are the scored of students changing in general. Interesting, I will definitely like to talk that through :).
Petr Bartel wrote:
As a former student, I used to be fully satisfied with a form of lectures, with your presentation skills and materials were great too. Perhaps for someone who didn't pass some C/C++ introduction it could be somewhat tough. I was very surprised with your being able to know all these structures, pointers and so on by heart. Maybe it could be great to get hands on. There is some new subject PB173 where it is possible. Or maybe insist on doing supplementary tasks in slides for some points.
Milan Zamazal wrote:
Interesting. Perhaps the course is basically the same while world changes. People work with computers in very different ways than they used to 8 years ago. There are several aspects of it: a. When we started to learn Unix we read the single "wise" book several times and tried to understand everything. Some years later there still used to be extensive HOWTOs available. Today? Most often googling for answers to questions in newsgroups, i.e. quick retrieval of isolated information without deeper understanding. b. We all started with C. How many programers use C regularly today? c. It was easy to get exposed to powerful things when the choice was limited. There are so many choices today, that it's difficult to discover and start using the right tools. d. The same applies to system administration. Things used to be done using simple, basic, powerful and well documented tools. Today many things are hidden behind undocumented interfaces without understanding what really happens (I'd often like to know, but it's difficult to find out). e. You couldn't use a Unix computer without some essential knowledge of the operating system some years ago. This is no longer true (I've got several GNU/Linux users who don't know what a file is). So simply: Your course probably becomes a specialist topic; while students answered questions related to their everyday use of computers years ago (easy), they have to handle abstract subjects separated from their computer experience today (difficult).
avakar wrote:
I think the decrease in the quality of students is not limited to your lectures. The number of accepted first year students is increasing every year (it's about three times as high as it was when I got enrolled), but of course the number of talented students stays the same. Also, ever since IB000 was taken over by doctor Hlineny, the lecture no longer serves to separate chaff from the wheat, allowing some of the talent-impaired student to get to as far as the second year. So while I can't comment on the quality of lectures, as I never bothered to attend them, I don't believe that your lectures are becoming less comprehensible, rather the students are becoming more soft-brained. For the record, I attended PV090 and I loved it (most of the people from our group dropped out, so there were only two of us, yay!). P.S. Maybe I'm wrong, maybe good students stopped attending your classes when they found out what your stance on The One True Programming Language (C++) is ;)
avakar wrote:
By the way, how do I structure comments into paragraphs here?
gRis>> wrote:
I believe there is some negative correlation with programming language popularity - everyone does Java these days, C is too old and lacks the high level comfort and C++ is a pile of incoherent mess. Also, actually using Unix on day-to-day basis now neither requires nor develops substantial terminal skills.
Vlastik Krejčíř wrote:
Well, I think (and hope :-)) you teach as good as I remember. However, I can agree with Milan Zamazal's post - the world has changed. How many students started with 8-bit computers 20 years ago? The level of abstraction has increased. My experience is they aren't sure what is the difference between GET and POST...
Joey wrote:
I am interested in UNIX, I wanted to learn new things and I declare myself as geek. So I didn't hesitate to waste 3rd attempt to pass colloquium. Luckily I did, even though I hadn't study anything since 2nd attempt (and I've made +7 points progress). I accept your test as challenge, but I think it can't give you good reference about students knowledge. There are a lot of questions which are rather difficult with quite confusing answers, and with many of them I had the feeling, that I am on the right way, but I'd like to talk about it, since there were many aspects I had to keep in mind. But the test evaluation is from this point of view quite strict (+2/-1 means 3 points difference). I can realize myself, making another silly mistake in last test, resulting in not passing the course, but I think there won't be any relation with my real knowledge, skills, etc.
Me ... once again wrote: statistics
Is the stats about 47% success rate correct? Because when I look into 'Given Credists Stats' it says that 60% of people failed?
Yenya wrote: Re: 47 %
My graph does not include the "-" grade (i.e. people who did not try to pass the exam), while the credits statistics in IS MU counts them as failed.
Vasek Stodulka wrote:
It is simple - in 2002 a) everyone was installing some Linux distro and b) everyone, who was installing linux in 2002 had to know man pages, text configurations, basic commands and more then half students was downloading and compiling kernels and other stuff - so much tools was in form of source only those days. When something did not work, we simply repaired it in kernel (or i. e. mplayer) source, recompile that and had fun - I remember patching kernel for my TV card and remote controller, patching mplayer not to change resolution when switching fullscreen in SDL and making some crazy SiS chipset to work. Nowadays nobody does this stuff I think, because software is much more complicated, everything is available as binary packages and last but not least Windows is much more usable then in old times, so a lot of present students (as I heared) do not use Linux at all. And one more thing - back around 2002 there was Us doing the exam. :-)
Pavel S. wrote:
Hello, I have read such an interesting article today [in czech] http://zdrojak.root.cz/clanky/konec-starych-programatorskych-casu/ and I guess the author is right. The world has changed a lot and it's still changing. I attended your first UNIX course last year (succesfully passed) and I guess it was one of the most interesting and best-taught courses I have seen at FI. On the other hand, as several people remarked, sometimes I have found the topics a bit theoretical and not really usable for maintaining my own Linux computer. However, during next year that information saved me a lot of time several times. To sum it up, your teaching is OK (still the same?) but both the students and practical usability of the course topics are different.
Petr Šigut wrote:
Já myslím že mám daleko jednodušší odpověď:) Dříve jste recykloval nějaké otázky z předchozích let, takže se někde na FIMUNY ty otázky pomalu skladovaly a studenti ty opakující se uměli. Jak jsem dělal zkoušku já tak jste přišel s tím že jste udělal úplně novou sadu a zvládli to snad jen 3 studenti... Jinak mi se přednášky líbily a byly zábavné, ta jednička mi nešla protože C moc neumím a nechtělo se mi učit co dělá která funkce...
Yenya wrote: Possible answer: Population
Yesterday's presentation of Radek Pelanek had an interesting fact: in 1996, the population of 18 years old Czechs has been about 200,000, and about 50-100 students enrolled to the first semester. Nowadays it is 120,000 possible candidates, and 400-500 students entering the first semester. And the situation will get even worse (to about 80,000 possible candidates).
Reply to this story:
Tue, 04 Jan 2011
Outgoing Mail
I archive every e-mail message I have sent since April 1999. Today during making backups of this data I have decided to count how many messages I have sent each year:
| Year | Messages | Bytes | Bytes/message |
|---|---|---|---|
| 1999 | 2110 | 5082832 | 2408 |
| 2000 | 3211 | 9393052 | 2925 |
| 2001 | 4693 | 30317777 | 6460 |
| 2002 | 7122 | 22466804 | 3154 |
| 2003 | 7669 | 25649473 | 3344 |
| 2004 | 5664 | 19496614 | 3442 |
| 2005 | 5311 | 22807724 | 4294 |
| 2006 | 5375 | 22106651 | 4112 |
| 2007 | 4828 | 33458711 | 6930 |
| 2008 | 4575 | 28287830 | 6183 |
| 2009 | 4715 | 27140012 | 5756 |
| 2010 | 4727 | 35894517 | 7593 |
The variance is relatively high, but apparently e-mail messages are getting bigger with time (no surprise here), and I have sent less messages than I used to send five to eight years ago.
4 replies for this story:
Karel Zak wrote:
You're a bot :-) I sent 7314 e-mails in last 12 years.
Yenya wrote: Re: Karel Zak
No, I am not :-). It is just some people are used to communicating over e.g. IRC (such as in RHAT, as I have been told), and some use e-mail as a primary means of communication. I, for example, barely use Jabber, and do not use other messaging tools at all (ICQ, Facebook, IRC, etc.).
Bobby wrote:
Hi Yenya, do you count your private communication and the work one together? I am not able to sum up all my e-mails as I use many different addresses for different purposes -- I am rather curious about statistic of my communication. However (and without any data), I am pretty sure that the proportion of my personal e-mails is diminishing almost to zero in a couple of last years...
Yenya wrote: Re: Bobby
Both private and work (I use a single system for both types).
Reply to this story:
Mon, 03 Jan 2011
PF 2011
I wish happy year 2011 to everyone who reads this blog.
The image is from Pink Granite Coast in Brittany, France. I plan to post a follow-up series about the places I have visited last year, including this one.