Small consolation

Guilt-Free Grade Assignments

One of the hardest things to do, I’ve found, is figure out someone’s final grade.

Putting “D-” on one lab report is relatively painless… double check to make sure that yeah, it really is that bad, but it’s only one grade out of many.

Entering “C” into the official records takes a little more work. It’s particularly hard for the ones that I know have tried consistently all semester. “C” could mean that the student will have to repeat the class. It could prompt them to change majors or careers. It could affect their scholarship or financial aid. Then again, turning all “D”s into “C”s and all “C”s into “B”s isn’t doing them any favors, either, because it overestimates how prepared/apt they are for the next class. So I have to check my sympathetic wish to nudge them up, just a hair, they’ve worked so hard for that “C” and is there really THAT much difference between their work and that of the lower-end “C” students?

Ignoring the names, though, it’s an interesting problem in classification. I’ve got 119 samples, who’ve been tested in various ways with the intention of sorting them into 5 categories, which just happen to have alphabetic labels. Now all I have to do is work out the best way to classify the samples, starting with some pre-determined borders for the categories. There are two types of measurements available, which can be broadly described as measuring independent understanding (“exam” type measurements) and as measuring consistent work to gain understanding (which includes homework, lab work, and attendance grades). The easiest of these to work with are the averaged exam grades and the work-over-time-except-lab. Plotting the exam average against the other average doesn’t take into account the lab grade, however. Should that go into the exam average (because it includes some lab quizzes) or into the other average (because it’s mostly work over time)? Let’s average the lab into both grades. (Losing some orthogonality, but I don’t feel like shoving the data into Matlab for PCA.)

This actually works out pretty nicely:

Classification of 119 samples using two parameters

There’s a fairly clear line between most of the categories. Which means…. that it’s quite easy to draw straight lines between those categories as grade boundaries. This also agrees well with the grades based on the overall average (collapsing the two dimensions here into one dimension, so no surprise)… but it makes it a little easier to show a student that “yeah, you really do fall right into the middle of the ___ pack, so better luck next semester”.

Now to finish up grading the senior class… this is much harder to do with 7 samples than with 119!

Analytical: The Central Science

Fresenius, in the announcement of his journal:

It requires little knowledge to realize that all the major developments in the field of chemistry are connected in some way with the developments in the methods of analytical chemistry.  The development of suitable methods of mineral analysis resulted in the establishment of stoichiometric laws; improvement in the methods of inorganic analysis made possible the accurate determination of atomic weights; after the foundation of organic analysis a rapid development of organic chemistry occurred, while spectrography immediately led to the discovery of new elements.  The development of analytical chemistry, therefore, always preceded the development of general chemistry.  The development of analytical chemistry, however, is important to other sciences and professions. …. Together, therefore, analytical methods represent an important contribution and a worthy scientific treasure for mankind.

How (Not) to Cheat

If the first page of your take-home exam says “Do not use Wikipedia or other websites”, then do not copy paragraphs directly from Wikipedia to answer questions. Wikipedia’s English grammar isn’t perfect, but it may be better than yours. The sudden change – and surprising depth of knowledge – may arouse suspicion. Or crystallize suspicion into certainty, if you’ve been reproducing figures from the book with surprisingly accurate detail, or using formulas in chapters we never covered in class.

And certainty sends the professor to the chair, and then to the dean.

Closing the virtual office

A couple of my students (all the undergrads, as it happens) added me as friends on Facebook. No particular problem there – even if I had any interest in posting my wild and crazy exploits there for all the world (and my students) to see, I’d already become “friends” with the department chair on it.

No, the real trouble seems to be the “chat” feature. If I’m logged into Facebook, then one of my contacts can send me an instant message through the website. I can reply, if I like, and this simulates interactive chat. Nice idea, but…. the “chat” questions I’ve gotten from my students have all been homework help.

I hadn’t expected it to turn into “virtual office hours” which would never end.

I’m happy to answer questions by email, because (1) email doesn’t usually expect an instant response, (2) it’s easy to forward a reply to the entire class, and (3) they all have equal opportunity to email me. Answering questions by Facebook is a little different. For one thing, it cuts into the rare hours I’m not already thinking about this class. Second, questions are too easy to ask and too hard to answer “for everyone” on it. And this particular student isn’t really in need of that much assistance; he mostly needs a bit of self confidence in realizing he’s got the right answer.

The real annoyance of it is that it hijacks a chance to catch up with friends from college and grad school into a tutorial session. And one in which I walk the delicate line between too much info (giving away the answer) and too little (which frustrates us both). So from now on I either reject requests to be “friends” or make it clear that I’m not going to be giving homework help that way?

Probably the former… it’s easier to be firm on that.

The Work of the Fig Cartel

Whatever happened to Apple Newtons? Why are they so hard to find?

Strangest Warning Ever

“No part of this web site may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying or recording or by any information storage and retrieval system, without written permission from an officer of AWMA, Inc.”

It’s on their website. If you can read it, it’s already been transmitted. Too bad they didn’t think to blanket the world with written permission forms before setting up their site…

Feline Behavior

I’ve been hosting a cat for just about 2 months exactly. “Hosting”, since he is emphatically not mine – he’s my college roommate’s cat, staying with me this fall while Doug and family are up in New England for an internship. (Borrowed house + cat who gets nervous = bad combination, I’m told. Come to think of it, my place is sort of “borrowed” too… hmmmm….)

Anyway, it’s been my first experience of being around a cat for more than a few hours at a time. Quite interesting. Especially when it comes to digestion. Tip: Cats may love milk, but milk must actively resent being eaten by cats. Tip 2: It’s the lactose that’s the problem, since cats can’t digest it; there is a special lactose-mostly-removed milk they sell in very small cartons.

Everyone knows that cats say “meow”, or “mew”, or “meeooooowww”, or whatever spelling. I believe I’ve figured out what they’re actually saying: “Mew” just means “me”. They only say “mew” because no other personal pronoun is meaningful to a cat. “Meeeeee”, he says as he winds around my feet in the kitchen… There might be food! who should get some? “Meeeeee!” Who wants attention while I’m watching TV? “Meeeee!” Who needs in when I’m hiding away in the cat-free office to write lecture notes? “Meeeee!”

Getting rid of “home”

Finally saying good-bye to my first car this weekend. It’s about time, since I found a replacement for it a good six weeks ago and it’s been sitting in the parking lot ever since.

(Oddly, it cost more to insure as a second car than it did when it was my only one. Weird.)

I’d had it for nearly 7 years, all through grad school and my postdoc up in the cold frozen North. The cold frozen North had not been kind to it, but it’d been pretty reliable even so…. Until I moved it down to the warm sunny southlands. Was it angry about leaving “home”? The shocks sounded bad, it clearly needed work… and then I found the first patch of rust on the body. Straight into the shop! Ooh, that’s rather a lot of money right now. And there’d still be that rust. You’re – you’re not serious, that’s more than the car’s supposed to be worth…

Bad Grades

I gave my students their first big exam on Monday. Finishing up grading it tonight, so this is both a reflection and procrastination…

The exam itself was 5 pages. For an hour-long exam. Was I insane? Yes, I’d test-driven it the day before and finished up in about 25 minutes… but I’ve got a PhD in this stuff and I knew exactly what I was thinking. In the future, keep it shorter.

The first part of the exam was very straight-forward: Define 8 terms, straight out of the textbook. There were a couple that gave them trouble, such as dynamic range. Yes, they had trouble defining dynamic range. That’s one of those “core concepts” in instrumental; one of the criteria by which different instruments are evaluated. (It’s the range of concentration over which an instrument gives a linear response, between the limit of detection and the limit of linearity.)

We’d been over this in class. It’s in the textbook. Only two out of the 10 students got even half credit.

Another question dealt with propagation of error – comparing the uncertainty that we calculate (estimate, really) for using pipettes in lab. The equation to use was on the exam – not identified as such, but as long as they had some idea of what they were doing with that, they should be able to pick out the formula. The question said, “use propagation of error”.

Again, the question told them how to do the problem. ALL they had to do was pick out the formula and plug in the numbers.

Not one of them got it right.

Was it my wording? Was it their background? I’ve known/suspected that most of them have fairly limited backgrounds in analytical to begin with. Perhaps I should not have asked them about that.

And yet… this class (this subject) is all about working precisely and intelligently to achieve the best, most accurate, most precise, results possible. Understanding how uncertainty accumulates is (IMHO) a core part of that.

I dropped the question, the grades were bad enough from the rest…

Was it that they didn’t understand what I was asking? After those first “definition” questions, most of the problems were given some kind of context. They were measuring traces of zinc in copper ore, or testing fluorescence instruments. Did they have trouble picking out the useful information? Did they have trouble understanding the question I was really asking?

(Some of my students have asked for extra practice problems in calculating parts-per-million concentration units – I learned that in high school, certainly by the time I’d had the first analytical class. Are they ready for this class at all? If they’re not, how can I get them ready?)

We spent an hour in class talking about spectral resolution of different instruments. We saw in class what narrow lines look like seen through too-wide slits. The textbook also covers this in detail. I asked them to sketch what two closely-spaced emission lines would look like through a monochromator without enough resolution (actually two different slits to choose from, one 0.2 nm that was just barely adequate and one 0.7 nm that wasn’t).

One of the better students gave me a very nice graph showing a narrow line at the position of each emission line. One had a height labeled 0.2, the other’s height was labeled 0.7. Another drew two broad peaks and labeled one “excitation” and one “emission”.

Is that a case of not understanding the problem? or not knowing the answer?

I really want to know which. In either case I consider that I’ve done poorly – in one case, for getting too clever on the test so that they wasted their time instead of showing me what they do know; in the other case, I failed to present this essential content in ways that they could understand and retain in useful ways.

First thing to do, I think, is to talk about this pretty frankly with them when we go over the test in class tomorrow. They’ll be worried about their grades; I’m more worried about how we’re collectively doing in picking apart the material in ways that make sense to them.

Second thing to do, I think, is to assign more homework. (They’ll be so thrilled.) Most of them need practice in basic math like converting mass and volume to concentration and back – I can write homework to target that, at the very least. I think they’ll benefit more also from a steadier stream of smaller problems, set explicitly “out of context” so that they can focus on the problem itself.

(I still want them to be able to recognize the problem in context, because that’s a far more useful thing to take out into the world. But they need to walk first before I make them run.)

When I teach this again next fall, I think I’m going to spend the first week, maybe 2, in a review of the first semester of analytical. It should be boring review for the undergrads who’ve already had the quant class here… but most of the grad students taking this class are badly underprepared on the basics. I assumed that they’d all had those basics already, and… well, the question on relative standard deviation should not have been missed by anyone, if that were true.