Short notes

My research is going quite well, so I don't have time to write much here. I'll close out the month of July now with some short notes...

I've successfully switched to Google Chrome after getting tired of slowness and "A script on this page may be busy" error messages from Firefox. Especially handy features of Chrome: (1) it has its own PDF reader, so it doesn't arm-wrestle with Adobe Reader when it wants to show you a PDF; (2) you can use your Google Account to synchronize your settings across multiple computers.

In a Roaming User Profiles environment, Chrome loses its settings when you log out. This may not be a problem if you sync through your Google Account. Otherwise, to fix it, get Chrome's Group Policy and install it on your domain controller; set the user data directory; and in the path that you specify, be sure to use Chrome's notation (${roaming_app_data}) rather than Microsoft's notation (%APPDATA%). More about that here.

A choice quote from Pandora.com: "Felix Mendelssohn (1809-1847) had a life opposite the image of the Romantic ideal. He was well-educated, had very stable home and professional lives, and never had to struggle to be accepted."

Like many people, I often look up local house prices on Zillow as an economic indicator (I'm not selling or buying anything myself). Zillow has apparently changed algorithms this week and no longer rounds to the nearest $100. Also, this article makes it clear that Zillow is a seller's index — based on voluntary sales only — while the Case-Shiller Home Price Index is a buyer's index, based on both voluntary and involuntary sales. That's why they often behave differently.

Permanent link to this entry

Spam purportedly from "Michael A. Covington"

Two friends so far have received spam (advertising e-mail) that claimed to be from Michael A. Covington (my name).

Both friends were good enough to send along full copies. I have analyzed the headers and found that, apart from "Michael A. Covington" in the From line, there is nothing to connect the spam with me, my computers, or my networks.

I presume the spammer is simply dropping well-known false names into the messages and chose mine. The spammer does not claim to be me and does not appear to know anything about me.

Permanent link to this entry

30 years

Today Melody and I celebrate 30 years of happy and successful marriage. For the rest of the story, click here.

Permanent link to this entry

Do you know you're in college? Or graduate school?

It's almost time to welcome a new bunch of graduate students, and I'm a bit startled to note that this will begin my 31st year of college teaching (29th at UGA; I started at the University of Southern California). Over the years, I've learned a lot about how to do it.

One thing I've noticed is that some students don't seem to quite realize where they are — they think college or graduate school is something else. At next month's orientation sessions, I'm going to do my best to head off these misconceptions.

We have a substantial number of undergraduates who think they're still in high school. Of course, upon arrival, they can hardly think anything else, but most of them realize very quickly how different college really is. Sadly, a few don't catch on.

Symptoms of thinking you're still in high school include the following:

• Wanting lots of repetition and review. High-school courses are slow and repetitious. In college, we move through the material much faster, and we move through it just once. Review is mostly something you do by yourself, with your books and notes.
• Viewing the teacher as an adversary, or more generally, failing to take ownership of your own education. High schools have a captive audience; colleges don't. You came here voluntarily. If you talk as if you don't want to be here, we'll believe you.
• Being overly concerned with the attendance policy and wanting a quota of days you can miss without penalty; maybe even wanting time off for "school activities" that are purely recreational. In college, we do not have unnecessary class meetings. Whenever you're not here, you're missing something. If you had a good reason to be away, we can sometimes help you make up what you missed. But we can't grant time off. That would be tantamount to saying we didn't need to have the classes in the first place.

Analogously, sometimes graduate students think they're still undergraduates. The biggest symptom of this is expecting the teacher to tell you everything you need to know — relying on the faculty to make you do things.

Graduate school is where you learn how to make new discoveries and learn things even your teachers don't know. You're expected to immerse yourself completely in your studies. In fact, you should be constantly reading and learning, beyond what is assigned, simply because you want to know.

You're expected to arrive prepared, too. We require you to write a thesis, but we don't teach you how to spell, punctuate, or use a word processor. You're supposed to know that already.

In fact, being in denial about the thesis is a serious warning sign for arriving graduate students. The thesis isn't a strange hurdle that we set up at the last stage of your studies. It is what you came here to do. People who don't want to write a thesis shouldn't enroll for a thesis degree. If you don't start thinking about your thesis as soon as you arrive — at least think of thesis-writing as something you want to know how to do, even if you don't know the topic yet — you're going to start it too late or not at all. You're like a soldier who doesn't want to put on a uniform.

Some graduate students think they're in trade school. That is, they think they're just taking courses to learn techniques — in our case, computer programming languages and algorithms — and don't want to be trained to do research. These are the students who are very conscientious with homework and lab reports but don't know how to go beyond what is fed to them in the classroom.

An even worse misunderstanding is that sometimes, foreign graduate students seem to think they are in a special program for international visitors when they're not. They don't realize that they are going to be expected to write papers and give oral presentations in English, just like the native speakers.

We have plenty of excellent international students, and most of them have no problem with this misconception. But sometimes it crops up, especially when there is a cohesive group of students from the same country. Back in the 1990s, we got bitter complaints from students telling us we were requiring too much English in our courses and that our expectations were too high. What could I say? English is the language of instruction at this university.

Crucially, a graduate student's work is presented to the public, not just the teacher and has to be presentable. Graduate students give talks at conferences. They write papers for publication. Good English is required regardless of where the student came from.

Permanent link to this entry

Congratulations to three new graduates...

Three of my graduate students finished their master's degrees this past week.

Charles Hollingsworth showed that it is possible to use sentence structure to identify the author of a written text. He worked with several examples, among them a set of 1920s detective stories that were quite similar but written by different authors. His method, "syntactic stylometry," worked.

Shayi Zhang unveiled a very user-friendly software package for counting words and other indicators of the nature of a written document. Originally designed to pick up indicators of deceptive language, it also has other uses. The software is called LingCues and she hopes to distribute it to the public after settling some questions about intellectual property rights.

David Robinson used machine-learning techniques to determine whether articles in medical journals contain "patient-oriented evidence" (good-quality research evidence about things that matter to the patient, as opposed to physiological theory or individual case reports).

All three of their theses will very soon be available here (scroll all the way to the bottom; if a PDF copy of the thesis is available, the title will link to it).

Congratulations!

Permanent link to this entry

The Colorado massacre and anti-intellectualism

Our thoughts are prayers are with the victims (and indeed the perpetrator) of the mass shooting in Aurora, Colorado. So far, the media have not reported anything credible about the motives of the shooter. It's still a mysterious tragedy.

What the media are saying is that he was "book-smart," intelligent and scholarly.

I fear that in a time when Americans are already getting frustrated with higher education (due to oversold student loans and a poor job market), this may trigger a wave of anti-intellectualism. People may start saying that intelligent people are dangerous and all on the verge of going mad. We already got a whiff of that in an earlier shooting case.

"Don't trust smart people, they don't have any common sense" is unfortunately a thread that has run through American popular culture for a long time. It may have sprung partly from a well-intentioned desire to honor the vast number of Americans who have worked hard (and intelligently) and succeeded without higher education — combined with some confusion of intelligence with education. But it may also be just envy. "I'm not as smart as him, and I'm better off because of it." (Like saying to a wrestling champion, "You may be strong, but you're muscle-bound.")

Also, the misconception that "great wits are unto madness near allied" dies hard. Partly this is because madmen (deluded people) often claim to be geniuses — and if you're in an environment where real scholars don't congregate, you're proportionally more likely to see crackpots instead, claiming to be geniuses but staying away from anyone who might give them intelligent criticism.

The actual medical evidence is that mental illness tends to go with lower intelligence rather than higher. The tendency is not strong, but it's in the opposite direction than many people believe. Also, mental illness in a more intelligent person may be more noticeable, just as a muscle disorder might be more noticeable in an athlete, simply because more intelligent people are more aware of how well their brains are working.

Permanent link to this entry

Can you stop your car without using the brakes?

I'm not gone, just busy. Today's entry is an outgrowth of two interesting forum conversations. Please share it.

Everybody who drives a car needs to know at least two ways to stop it without using the brakes, and know them so well that they come to mind instantly if the brakes fail or the throttle sticks.

They are: use the parking brake and shift into neutral. The second is likely to be necessary because the parking brake isn't powerful enough to overcome the power of the engine.

"Unintended acceleration" is a new and mysterious automotive failure, but stuck throttles and failed brakes have always been with us. A stuck throttle is what happens when the accelerator linkage sticks so that the pedal doesn't come up when you release it. If you shift into neutral, you can use the brakes to stop the car.

Behind all this, every driver needs to think of the car as a machine and know what various parts of it actually do. For example, that "N" on your gearshift isn't just for decoration — it provides a way to uncouple the engine from the wheels. And you can shift into N from D or R without stopping. That's why the order is P R N D L.

(Other ways to stop the car might include heading up a steep hill or into shrubbery — not trees! Switching off the engine is not a good idea because it may lock the steering. If you can shift into neutral, you don't need to switch off the engine.)

I'm convinced many, if not all, cases of "unintended acceleration" actually come from having the foot on the wrong pedal. In that situation, do you have enough common sense to realize that if the pedal is making the car do the wrong thing, you shouldn't push harder to make it do more of the wrong thing? More generally, don't you always start off gently, so that if the wrong thing is happening, you'll notice before you get up to speed?

I've heard too many accounts of people who do absolutely tragic things in an emergency (or in response to a small mistake on their own part) because their understanding of the car goes no farther than, "Push this pedal to go, push the other pedal to stop." Please don't be one of them.

Permanent link to this entry

End of the checkbook era

Yesterday Melody and I passed a minor milestone as we move deeper into the 21st Century.

We no longer have a checkbook. That is, we no longer have a pocket-sized book of small (2.5×6-inch) checks.

We can still print checks on the computer, and we can also fill out computer checks by hand when we need to. (We have more than a lifetime supply of them.) That is how we are going to handwrite checks from here on.

And on the business side (Covington Innovations), we still have a big ledger-style checkbook in which I write checks with an elegant italic pen — for income tax and very little else. I write about twelve checks per year, eight of which are quarterly tax payments.

For the most part, payments from our checking account are made by "PC banking" — the bank prints and mails the check for us. Of course, whenever possible, they do an electronic transfer instead, and that's entirely their concern, not ours.

As I mentioned earlier, Britain is abolishing paper checks, replacing them with electronic transactions. America has decided that paper checks should be electronic transactions — that is, the image of the check should substitute for the real thing — and that hasn't worked quite as well as people hoped. It's too easy to fake. Instead of relying on either an authentic piece of paper or an authenticated electronic transaction code, we've created an awkward mix, and it won't last.

Permanent link to this entry

Outsourcing clerical work to the customer?

I've detected another wave of one of this decade's business fads: making customers do clerical work via the Internet.

In some cases this is beneficial. I enjoy picking items from catalogues without a human intermediary. The same goes for picking airfare and hotel bargains.

And if the work really involves entering numbers into a computer, I might as well enter them myself rather than dictate them to a human being sitting at a computer.

But outsourcing clerical work can go too far. I want to offer some "don'ts."

(1) Don't try to make the customer do something that actually requires an employee's expertise.

Remember, customers are not employees, and the most knowledgeable 10% of the customers are not typical. Some interactions really require things only an employee knows. That's especially the case if the product is something commonly bought by people who are unfamiliar with it (such as life insurance or air conditioners) or if there is specialized terminology involved.

What's more, it's important to let customers contact you and ask you free-form questions. The ability to do this by e-mail (or secure web messaging) can be very convenient. All too often, though, a company's web site seems to be designed to prevent customers from ever asking for anything not already foreseen.

(2) Don't expect the customer to remember an elaborate procedure for something he only does rarely.

Procedures that may be fine for an employee who does them 100 times a day will mystify — or needlessly burden — people who only need them infrequently. No, I don't remember exactly how I submitted an insurance claim 18 months ago. You'd better make it obvious how to submit the next one, without relying on any memory of past procedures.

(3) Don't make the customer keep track of a user name and password (or even ID number) unless absolutely necessary.

I already have to keep up with about 200 user-name-and-password combinations. I do not want to be burdened with yet more of them. And no, I'm not going to use the same user name and password that I already used on something important. I don't trust you enough to give you a password that might also get you into my bank account.

Why can't you recognize me by my name and address, and (when appropriate) my credit card number?

(4) Don't do all this unless there is a clear benefit.

Unfortunately, some people think the Internet is such a glamorous novelty that people will jump at the chance to use it, even in order to do drudgework. For many of us, those days are past (and were probably past before you even heard of the Internet).

I don't want to access your business through a web site unless the web site does something that benefits me and — crucially — is a useful addition, not a substitute for good customer service.

Permanent link to this entry

Customized clips secure GPS cable

Here is a very small car-improvement project I carried out just now. The problem was that the power cable for the GPS device was flopping all around in front of the dashboard. The solution:

These are All-States FRC-1 flat ribbon cable clips, with peel-and-stick adhesive pads. I got them at Fry's; you can order them here. They are handy for other wiring projects and for taming computer cables on desks. Radio Shack has similar but smaller clips designed for telephone wire.

After using a couple of the clips right out of the package, I decided that although they did their job, they looked a bit conspicuous on the dashboard of my Escape. So I got two more and spray-painted them...

And here they are, in situ, doing their job. When the GPS is packed away, the clips are not very conspicuous, and the paint makes them look as if they were meant to be there.

Permanent link to this entry

Three automotive notes

I'm still occupied with other things — especially taking time off to enjoy Cathy's visit — and am not blogging much. But here are three notes, all having to do with cars.

Matching wits with windshield-wiper blades

After driving through a heavy thunderstorm the other day, I arrived home with determination to put new wiper blades on our cars. (You need them before you think you do. Blades that look more or less OK can be a lot less safe than really good ones.)

On the 2012 Escape, Ford uses a new squeeze-and-pull system that makes the blades very easy to change, but most aftermarket manufacturers don't support it at all, or at best, they use clumsy adapters. After looking through the selection at an auto parts store, I ended up at the Ford parts department and came away with a pair of Motorcraft WW-2046 20-inch blades. Snap, snap, and the Escape is in much better shape.

The 2004 Taurus uses the old J-hook system, making it easy to install a premium aftermarket blade. I chose Rain-X Fusion and was temporarily outwitted in the following manner. I put the blades on quickly and easily, observing, "Hmmm, this synthetic rubber is yellow instead of black." Then I turned the wipers on — and — ouch! — the rubber part seemed to come right off the blades! Further inspection showed that the yellow squeegee was actually a protective cover, and the real squeegee was beneath it. So it was supposed to come off. But it gave me a moment's fright.

Driving with GPS

While in Lexington, Kentucky, the other day, I finally got around to buying a Garmin Nüvi 50LM GPS navigation device. Some thoughts:

Paradigm shift: The GPS is not primarily for finding things whose location you are completely unaware of. It's for recovering from route changes, as when a blocked road or traffic conditions force you to go an unexpected way. You can rely on it to get you back to a major road and headed in the right direction.

Having said that — of course, it's very accurate, and it really does find things. I recommend reviewing the route before heading out, just in case there's been a mistake or it has chosen something that you know is not optimal.

It talks... but it often mispronounces things ("Nicholasville" as "Nitch-o-lass-ville" was my favorite) and calls roads by unfamiliar names when they have more than one designation (Tennessee 456 came out as Tennessee 297, which is apparently officially correct but not on the signs).

Update your maps: Downloading a full set of updates took a couple of hours but led to much better route preferences.

Akins Ford (Athens) is now Athens Ford

You may recall that I mentioned a few years back that Athens, Georgia, temporarily lacked a Ford dealer. To fill the gap, Akins Ford, of Winder (25 miles away), took over the Athens dealership. That lasted until very recently. Now Akins has retreated to its original habitat and the Athens dealership is under new ownership and is known as Athens Ford. I haven't done much business with them yet, but I certainly got good service at their parts counter this morning.

Permanent link to this entry

What competition ought to be

In a fight, competition means beating the opponent.

In a team sport, it means much the same thing, but the beating is symbolic.

In an economy, competition is something radically different. It does not mean vanquishing the opponent — it means offering people a choice.

That is exactly what has happened with two auto repair shops near me. With their employer's approval, good will, and even assistance, Tracy and Cristie Morgan and Tracy Miranda, formerly of Car Craft in Athens, Georgia, have spun off a separate business called Classic City Motorworks. I am happy to recommend both establishments.

Permanent link to this entry

Short notes

This is a busy time, with much good news that is not yet ripe for disclosure. Regarding July 4 and American independence, see my entry for Memorial Day. And regarding the surprising survival of Obamacare, I have nothing to add to what I said here and here and here and especially here. Bottom line: If you say there are technical and economic flaws in the plan, I agree. But if you object to the very idea of spending your own money to help manage risks faced by other people, not just yourself, you don't know what insurance is.

Permanent link to this entry