Gas station without pumps

2012 December 31

2012 in review

Filed under: Uncategorized — gasstationwithoutpumps @ 11:16
Tags: , , ,

The WordPress.com stats helper monkeys prepared a 2012 annual report for this blog.

Here are some excerpts:

This blog was viewed about 110,000 times in 2012.

In 2012, there were 420 new posts, growing the total archive of this blog to 929 posts. There were 282 pictures uploaded, taking up a total of 34 MB. That’s about 5 pictures per week.

The busiest day of the year was May 14th with 705 views. The most popular post that day was 2011 AP Exam Score Distribution

Some of your most popular posts were written before 2012. Your writing has staying power! Consider writing about those topics again.

Click here to see the complete report.

I’m a little surprised that I managed to average 1.15 posts per day—I knew I’d been spending too much time on the blog, but … I’ve been getting a lot of views (for an unaffiliated academic blog), over 8000 every month since March, despite my obsession since the middle of June with the design of the Applied Circuits course (98 posts and pages so far, and the course doesn’t even start for another couple of weeks).

I’m rather lax about posting pictures, so I was surprised to see the number up to 282 (and averaging 123kB each—I do use Photoshop Elements to crop, touch up, and downsample images before posting them).  I suspect that a lot of the “pictures” are actually circuit diagrams and gnuplot graphs, which would be much smaller if WordPress.com would allow a vector graphics format like SVG rather than insisting on raster graphics like PNG and JPG for everything.  I guess they feel they have to support obsolete browsers, like Internet Exploder, that still don’t support SVG.

Here are my top-viewed posts for the year (over 300 views):

Home page / Archives 28,751
2011 AP Exam Score Distribution 10,615
West Point Bridge Designer 2011 3,990
Installing gnuplot—a nightmare 2,836
Why no digital oscilloscope for Macbooks and iPads? 1,814
2012 AP Exam Score Distribution 1,638
How many AP courses are too many? 1,618
Bring back the mammoth! 1,424
West Point Bridge Design Contest 2012 1,320
County Fair with Pictures 925
Soda-bottle rockets 821
Why Discrete Math Is Important and The Calculus Trap 812
Teaching voice projection 711
Raspberry Pi 697
Underwater ROV contest 651
Waterproofing cameras for underwater ROVs 639
Resources for bioinformatics in AP Bio 637
Coursera Course Catalog 608
Soda-bottle rocket simulation: take 2 608
Learning to use I2C 591
Satisfying UC’s a–g requirements with home school 516
EMG and EKG works 503
Computer languages for kids 500
Speed of sound lab writeup 498
Underwater ROV 481
Magnetometer and accelerometer read simultaneously 474
Soda-bottle rocket simulation 472
Should high schools and colleges teach sentence diagramming? 470
Google Scholar vs. Scopus and SciFinder 466
DRACO: broad-spectrum antiviral drugs 398
Green beard effect 394
Adding bioinformatics to AP Bio 388
Carol Dweck’s Mindset 361
Medical Instrumentation, Chapter 6 355
NSF “clarifies” Broader Impacts 351
Better electrode placement for EKG blinky 351
Prevnar 13 approved for adults 341
Instrumentation amp lab 321
A critique of CS textbooks 315
iPad Oscilloscope 305

People coming to the home page are still the biggest single group, but overall, more people are finding my pages through search engines than by other means. It is a bit disappointing that two of my top pages are just pointers to someone else’s collection of AP exam score distributions, and another two are for middle-school students looking to cheat on the West Point bridge design contest (which apparently some schools assign as homework).

The gnuplot post is mainly used by people looking for a way to install gnuplot on Macs. I think that the gnuplot community really needs to get their act together and put up a proper binary installer for Mac OS X. They need to stop pretending that everything is fine if you can dual-boot your system or put up a 2Gbyte Linux-lookalike environment on your Mac.

There is now an ok USB oscilloscope for the Mac, though the code is still clearly beta-release quality.  I posted about it in FET threshold tests with Bitscope.

A couple of the popular posts are mainly fluff (the mammoth post and the Green Beard joke picture).

A lot of my most-viewed posts are technical ones on electronics, computer engineering, robotics, or physics, though more at a teaching or hobbyist level than professional, cutting-edge stuff.  I’m fairly happy with that, as I’m turning more into a teacher than a researcher over the past couple of years—I’ve always enjoyed creating new courses, and I’ve been putting a lot of time into that lately. My more philosophical musings have had less readership, perhaps because they are harder to find in a Google search and are less likely to be searched for.

I’m still not seeing many comments (about 2.4/post and 40% of the comments are by me).  I’d like to have more feedback from my readers, but with so many of them being one-time viewers who came via search, I can’t really expect much conversation in the comments.

I’ve seen a strong weekly periodicity in my viewership, with dips over the weekend (people must be goofing off at work when they access my blog).  The plot is particularly clear when I look at the Google Webmaster tools, which has minima on Saturdays:

As much history as Google's webmaster tools is willing to show me.  It would really help to have the clicks and impressions on a log scale, so that the clicks didn't have just a one-pixel fluctuation.  I suppose it is too much to expect Google to be willing to provide free tools that are actually well designed.

As much history as Google’s webmaster tools is willing to show me. It would really help to have the clicks and impressions on a log scale, so that the clicks didn’t have just a one-pixel fluctuation. I suppose it is too much to expect Google to be willing to provide free tools that are actually well designed for geeks like me.  My overall click-through rate of 4% is decent, I suppose,

My university web pages have a higher number of clicks, but 57% of them for the soda-bottle rocket launcher plans and the mead recipe, neither of which are related to my professional responsibilities.  The next few are for handouts from tech writing course I taught in 2003—I’ve not really been putting much of general interest on my University web pages lately.

Quantcast.com estimates my 2012 page views for my blog as about 103,000 (a bit less than WordPress.com number, which is more directly measured), but Quantcast also estimates the number of visits and people as about 55,000 and 43,000 respectively, so people are viewing about two pages per visit.  Somewhat surprisingly, given my age and how many of my posts are about educating my son, my demographics skews towards younger adults without kids.  Less surprisingly, given the technical content and academic tone, my demographics skew heavily towards people with graduate degrees.

Not visible to the outside world, I also have 169 draft posts (usually just a pointer to a web page I planned to comment on, or a couple of lines of notes) and 71 tabs in Firefox for web pages I planned to blog about but haven’t even gotten into draft posts yet.  So I have material for a couple hundred more posts, if I can remember what it was I wanted to say.  If I ever go through the lot of them, I’ll probably throw out ¾ of them as being no longer relevant or interesting—they weren’t compelling enough at the time to make me write the post immediately.

Overall, I’m fairly pleased with my blog this year, and hope that next year will continue to be successful—maybe I can get the draft-post backlog whittled down a little (though to-do lists always seem to grow rather than shrink, no matter how many things I do).

2010 October 4

Secret class web pages

Filed under: Uncategorized — gasstationwithoutpumps @ 05:41
Tags: , , ,

I get irritated by people who have web pages for their courses, but make everything in their class hidden behind a password, even the syllabus for the course!  Some faculty seem to think that it is an imposition to put anything on web, and that everything they do in a course is so secret and precious that no one should know anything about it.  This, of course, means that everything they teach has to be a dead-end which no other course can build on, since there is no way to tell what they have taught.

Open class web pages have many advantages:

  • It is easier for students to choose courses, knowing roughly what is in the course and what the workload is, without having to rely on the often grossly inaccurate reports of fellow students.  The tiny descriptions in the catalog are way too vague and often very out of date.
  • It is easier for advisers to recommend courses to students.  Once again, the catalog descriptions are often inadequate, and faculty and staff advisers have to guess whether a particular course which sounds promising actually covers the material that a student needs.  There may be several courses which appear, in the catalog, to cover similar material. Open class web pages with more detailed syllabi and week-to-week schedules (or more detailed course notes) can provide the information needed to decide which courses have the depth and breadth or specific topics to best meet the student’s needs.
  • It is easier to share the teaching load among the faculty.  Teaching is often improved if teachers get a break from teaching the same course over and over, either through sabbaticals or by rotating teaching duties.  It is very helpful for faculty taking on a course they have not taught before (or have not taught recently) to have fairly detailed descriptions of what has been taught in the course in the past, including how much time has been spent on each topic and what homework assignments have been given.  Although this can sometimes be handled privately between teachers, having open course web pages makes it easier for faculty to decide which courses they are already well-prepared to teach and which will require a lot of extra study and preparation, which is good to know before curriculum and leave planning is done.
  • It is easier to plan courses. In science and engineering courses, there are nearly always prerequisite courses that are supposed to prepare students. Faculty need to check exactly what is taught in the prerequisites for any course they teach, so that they don’t bore the students by repeating material or lose them by assuming they know things that they’ve never been taught.  Once again, the tiny descriptions in course catalogs are nearly useless for this.
  • It is easier to plan curricula. In engineering fields, we generally have to tweak the curriculum every couple of years and overhaul it every decade.  These changes may be to correct for drift that has occurred in the classes that are taught, to adapt to changes in priorities in the field, to take advantage of new expertise on the faculty, to compensate for changes in prior student preparation, or to meet new budgetary constraints.  In interdisciplinary fields like bioinformatics, where the required courses may be taught by seven or more different departments, keeping track of the changes in the courses is difficult.  Small changes that seem unimportant to the host department, or that are easily corrected in follow-on courses in that department may be deadly to other programs that do not include the follow-on courses. It is not safe to rely on students to tell us that some course we require no longer teaches the material that we required the course for, since they are often unaware of the structure of the curriculum, seeing it as wholly unrelated, arbitrary barriers to their getting their degree (see my post Just scoring points).

There are some good reasons for having parts of a class web page be private:

  • Faculty may have permission to distribute copyright-protected or proprietary material to a class, without permission to distribute it to the world.
  • Student on-line discussions may have more involvement if the students know that the readers are restricted to fellow class members. (On the other hand, students put more care into work if they know that parents, future teachers, and employers could look at it, so there are good reasons for having some student work be for a wider audience.)
  • Faculty may be developing their own detailed teaching materials which they are not willing to have appropriated by plagiarists.  Preliminary drafts of textbooks are often distributed free by faculty to students in their classes as the textbooks are being developed, but it is not reasonable to expect these drafts to be provided to the world.

The default web setup for courses in our School of Engineering is for there to be a public web page for the syllabus and other general material, and a password-protected directory for material that has to be restricted to just the class. This approach allows for both public and private portions of a class web site, with fairly simple control by the faculty over which material goes where.

Some faculty do an excellent job of maintaining course web pages that are useful to the students, other faculty, and to the university as a whole.  If only I could convince more faculty of the value of doing this!

%d bloggers like this: