Midterm Self-Assessment

As we roll in to the middle of the semester, I'm starting to get a clearer picture of the hard data (grades) from what I have observed via anecdotal evidence. There's a clear Gaussian distribution of students of what I would call the average 70%, with 15% on either tail, making up the very good and very poor students. So far, I'm very pleased with this for a number of reasons.

The pressures to curve the class are non-existent. This is pretty important to me because that's a direct correlation to whether I have the class on the right trajectory. I feel like the average engineering student should be getting between an 80 and a 90 in a class, with the below average in the 70s and the above average in the 90s. If the entire class is in the 90s, then I'm not making the material difficult enough for the very good students and that's a disservice to them. If everyone is failing during the midterm, not only is there pressure to curve the class upward but it also drops morale, which can erode student confidence and evaluation scores.

Most students are pleased with the class, based on informal polling. I ask for feedback anonymously (or not if the student doesn't mind) after every large assignment. A handful of students use this to vent with things like "this assignment sucks" and "I didn't learn any of this crap". But the majority of the responses have been positive in the form of constructive criticism. Comments like "I think this aspect wasn't explained clearly and the notes were equally unclear. Can you give a better example next time?" are things that I can directly use during the next course to make improvements. Also, after these assignments, I do let students know about some of the changes for subsequent semesters so they know their comments haven't fallen on deaf ears. I'm thinking this could be a key part to get good evaluations which always looks better than bad evaluations.

I feel like I've covered more material and required more work from the students and they've responded positively. I've added some changes from previous semesters, beefed up the writing requirements, and had more lectures. Students have grumbled a little bit (who doesn't love more work) but most have grudgingly admitted it's been for their benefit. This also means I have a heavier time commitment for this class than previous semesters under different instructors but I think it's been positive for both the students and myself.

Now, with that said, there's definitely room for improvement. I've been waiting until the last minute to make up assignments and lab manuals, which is not a good trend to start. Also, there are a few more topics I should have covered during the lecture part of the course which would have benefited the students and probably saved me a few headaches. I've been late on getting information to the TAs but I don't require grading from the TAs so I don't think they're totally mad at me yet.

Overall, I'm fairly pleased so far. I only hope that I can keep this up for the rest of the semester and have the students be successful in the course.

First Classroom Lecture Report

Well, I've survived the first day! Actually, it wasn't so bad. I started out with a PowerPoint that covered the course objectives and syllabus. There was a lot of information regarding TAs, labs, lab rooms, etc, that didn't want to skip and so even though I dislike PowerPoint, I went along with it.

I tried to reiterate the relevant stuff about 3 times so hopefully some of it sank in. I got a few laughs at my jokes that seemed at least mostly genuine. Probably the best part of the PowerPoint portion of class was when I reviewed the Academic Honesty Policy with them. I told them this is something profs babble about on the first day and then most students don't think about it again. But for me, as a prof, if I suspect someone has cheated, it's about 20-30 hours of insane paperwork and meetings that I would gladly go through walking barefoot on hot coals if it meant dragging down a cheater. The class got very silent about that for a few minutes as I paused to let that sink in.

One student was brave enough to ask if I was serious and I responded by smiling gleefully and saying "Yes". I think they got the message.

The only other things to report on are that the students (being seniors) are a rowdy bunch and the room was over 85 degrees, which is too hot to teach. After the PowerPoint, I had to roll up my sleeves and loosen my tie for the Chalk'n'Talk section (GMP's phrase, not mine). The few undergrads that I have met prior to the class which are department helpers mentioned to me after that they thought the class was very good. So some initial praise was nice.

I only covered about 40% of the material that I wanted to cover but that's ok because I don't have set material to cover for the course (since it's labs/projects). That also means I have lecture 2 done so I'm ahead of the game. :-D

All-in-all, it was fine. My nervousness and sleepless nights seemed all for naught, although I didn't sleep again last night and I might have a fever so that could be the reason why I'm not sleeping...

The Economics of Academia (Part 1)

Today, over at Engineer Blogs, I discuss a reader's question on the economics of academia. Essentially, the reader asks why can't we have 20:1 class ratios and lecturers being paid $100/hr and not worry about bringing in research money.

It's not an easy question with a short answer so expect a few posts on this topic. I start out by comparing those constraints to a typical community college model.

Lesson Plans Down the Drain

Today, I thought I would be writing to say "Yay, I got a few lesson plans done so I won't be totally scrambling in my first few weeks". It turns out, last night, I found out my first Lab Module for my class, and first two lectures are completely wrecked. Plus, all my subsequent labs/lectures (which build on the first one), are all screwed too. I need to back up for second to give some context.

Part of the Lab Course I'm teaching will deal with data acquisition and signal processing. Currently, many of the students use LabVIEW but hate the interface. So, as part of the course, I was going to introduce them to Data Acquisition in Matlab, something I've been doing for the better part of 4 years (and talked about my love of Matlab here).

I wrote a whole new lab based on using Matlab as the data acquisition and data manipulation environment. I made fancy pictures, charts, snippets of code, etc. Then, last night, as I was working on my lectures to compliment the lab, DrWife had a great idea. We were giving NanoGEARS a bath and she was playing with some rattling toys. As she was shaking them, the pitch (obviously) changed when she was loosely holding it versus tightly holding it or when it was underwater. DrWife suggested that would be a great classroom demonstration. I could tie in damping (via water) or adding stiffness (holding tighter). And, because it was audible, it's much more interactive than just showing sine waves to the students.

I could use Matlab to acquire the signals, analyze the signals in the Fourier Domain. Add filtering, etc. So last night, as I was trying to acquire a simple audio signal with Matlab [using instrhwinfo('winsound') ], all hell broke loose. It turns out, in the new x64 version of Matlab (R2011a) only CompactDAQ devices are supported!?!?!? WTF? So all that legacy code that I've built up over 4 years is entirely down the shitter. So none of my sound examples will work using the standard code (now called legacy interface). And basically any DAQ card you purchase from NI, Agilent, or any other vendor won't work. CompactDAQ devices are such a small part of the possible hardware spectrum for data acquisition that this is a joke.

And the beauty of all of this is the computer lab at SnowU just fucking upgraded to x64 Matlab R2011a. So none of those computers can use Matlab to acquire signals with the data acquisition toolbox. Basically, I've planned 5 labs around using a used-to-be-mainstay piece of software that now is completely worthless. Now I have to scramble to make new lectures and a new lab manual centered around LabVIEW (another worthless software).

I've emailed, I've called, I've posted on the forums but I can't get any response as to when Matlab's going to fix their problem. And it is a fucking problem. Researchers want the new x64 Matlab because it supports 8+ GB of RAM, meaning much higher computation rates and memory. But for researchers dealing with instrumentation, like me, I also first need to acquire the fucking signals!

I would love to rant more but I have lectures and labs to redo. Fuck!

Academic Breakdown

Today, over at Engineer Blogs, I discuss why professors live and die based on their funding and give some rough indication of the amount of funding needed to sustain a research group.