care and feeding of graduate students (life as a professor)

Just back from the UK where my family and I spent a good amount of time exploring the public footpaths in the south of England.  There are two sets of appropriate maps Explorer (red) and Pathfinder (orange).  The red maps are great for driving and show the foot paths, but for the second time in two trips are not quite up to date.  We ended up turning a short 3 mile walk near Avesbury into a 6 mile trek (and climbed the downs several times) as we the path didn’t actually follow the map.  This is very frustrating as we now have several good map readers in the family and we knew where we were and where the path should be - it just wasn’t there.  The orange maps are (usually) more up to date - and show the hedge rows, but when I checked they aren’t always that much better. (wouldn’t have helped with this trip, but would have with the other one).

So database accuracy vs. the real world is still a critical issue.

As a professional scientist, as a computer security person and educator, I worry about cheating.  It’s easy to fool yourself, to think your results are significant when they simply are not and much of scientific ethics and the scientific method exists to catch these issues.  In many ways the apprenticeship that graduate students serve is to teach them about the fundamental honesty that is absolutely needed in science.  You throw the dice, get an experimental result and if it doesn’t agree with your pet hypothesis - you live with it.   I like to say “I don’t gamble for mere money”.

Cheating is a different issue in the real world. Computer security issues are not usually direct attacks on an established protocol but cheats based on human frailty.  No one expects to break a username/password combination by exhaustive trial and error (at least not until they have the /etc/shadow file :->), but people break into machines all the time by trying common combinations of usernames and passwords.  (I must admit the people who use root or administrator with Yahweh deserve the fruit of their blasphemy).

I was reading an interesting book on cheating, “how to cheat at everything” by  Simon Lovell.  (you can find it on amazon).  There are a fair number of combinatorial problems where the odds are different than one would naively expect - these will be homework for my security class.  Some of the hustles are simply magic tricks gone astray, but many of them involve a level of brazenness that is breathtaking.   It really reinforces the inherent sneaky streak that any good computer security person needs.  Between palming a small mirror to see the cards as you deal,  marking only a few cards (for example the face cards for poker or the low cards) to give an edge on your betting, and using slight of hand to swap dice between ones with two sixes and normal ones there isn’t much a dedicated cheat won’t do to fleece his mark.  Even my favorite Nigerian scams make an appearance.  It’s a good read.

And now for something completely different (and related to my day job).

My august university has been going through the throws of reacrreditation and I’ve been reluctantly dragged into the mess. One of the big issues is that we have to formulate learning outcomes. The term ‘learning outcome’ may mean something to members of the college of education but doesn’t mean anything in computer science. So of course, we did our best to define them, which was NOT GOOD ENOUGH. So I actually read some of the (very poor quality) education literature and found what they were talking about.

Learning outcomes must be specific, measurable, and quantitative things that students actually do in the course. Great for a lesson plan, possibly acceptable for elementary school, but insane at a undergraduate and graduate college level.

For example, with “introduction to object-oriented programming”, a sophomore course in computer science, we have general aim of “students will be introduced to the principles of object-oriented programming using the Java language”. Unfortunately, that’s not specific, measurable or quantitative.

So we change it. Let’s follow the instructions.

Students will be introduced to the principles of object-oriented programming using the Java language. They will learn object creation, extension with sub-classing, …, procedural control. Closer, but still not right - we’re specific, and maybe measurable, but not quantitative and it doesn’t say what the students do.

Here’s what they want, and what is simply incorrect.

Students will be introduced to the principles of object-oriented programming using the Java language. They will learn object creation, extension with sub-classing, …, procedural control. They will write the 10 (insert your nonsense number here) principles of object orientation, they will write the 4 (ditto) control structures of Java, they will write the 20 ways to sub-class,… .

Fine - except there aren’t an agreed upon 10 principles of object orientation (which is still an active area of language research), or any of the other things. In science in general, and especially at the professional level expected at university level instruction, there isn’t a simple way to quantify the knowledge. We either end up writing detailed lesson plans for every lecture - plans that are obsolete before class starts - let alone after a few years or we make the aims looser.

I vote for looser aims. Unfortunately I’m not on the accreditation committee.

I realized that I haven’t put one of these in, and could, in principle, be up the proverbial estuary without a means of propulsion.

Opinions, comments, and reviews on this site are my opinions and only my opinions.  They are not warranted to be accurate or unbiased.  They should not be construed to be the opinion or an endorsement by Georgia State University, the State of Georgia or any other legal entity.

so there.

Congratulations to Jeff C. on completing his dissertation defense.  There are still a few things to finish up Jeff, but you’re on your way.

After spending the 4th (you know the English should celebrate being rid of us troublesome “yanks”) shooting off fireworks in the great state of Alabama - where it is legal and trying to have fun messing around in boats - a thought occurred to me (rare event - I know).

It’s much more fun to be pushed around by the wind albeit slowly, than to be shoved around by a noisy infernal combustion engine however fast.

Patrick McManus is one of my favorite outdoor humorists.  While his focus is more on hunting, fishing and “base camping” than I’m wont to, he has a unique way of bringing out the oddball side of the outdoor life.

One of his stories, “cabin fever”, describes the effects of living in enforced isolation, and he describes various varieties of cabin fever - such as “continent fever” (I’ve had this), “villa fever”, and “two-man tent fever”.

I’d like to add one more to this list, “trail fever”.  Trail fever affects adult scout leaders who don’t quite trust each other, themselves, or the scouts to handle themselves in the back country.  It’s symptoms are the “troll-like” behavior seen with the other fevers, including sensitivity to otherwise minor issues that can be resolved by simply talking.

While returning to “civilization” (I wonder about whether we’re really civilized) can cure it, I think there are some good preventive measures.

1. Experience and training.  Train together, build experience together, learn to trust each other and the scouts.  Develop the skills to handle anything (almost) that mother nature can throw at you.
2.  Be Prepared.   Not just the scouts motto.  Think about what might happen and then choose equipment (or figure out alternative uses for equipment) that can help.   For example, if you are concerned about hypothermia - make sure someone has a light thermal blanket and remember that you can use packliners (trash bags) or raingear as a warming suit.
3. Self-awareness.  Monitor your self.  If you find that you’re snapping at someone or that they are consistently annoying you - take some time off and when you’re calm talk to them.  If you don’t then the annoyance will only build.
4. Communicate.   I like to tell my students, after doing a bit of slight of hand, that I dropped out of wizard school before learning mind reading.  No one else sees things the way you do - someone else might have a different viewpoint.  For me at Philmont the challenge was to get everyone through a challenging trail without injuries - and not how fast we went.  Other challenges - like how fast we went or how much we carried are valid challenges - just not my challenge.

On thinking about this, these same issues apply to managing a lab, and probably off the trail in regular life as well.

Just a short post with some philosophical reflections on the art and practice of backpacking.

Backpacking is not running.  It’s not a competitive sport like soccer, tennis or even (gag) golf.  So what is it? It’s what I call a “Zen sport”, where self-control and self-awareness are critical.  It’s a sport of resilience - not endurance or performance - where you get up, break camp, walk until you’re tired, make camp, eat, and go to sleep.  Then you do the same thing again and again and again.

Backpacking is an excuse to be in the country - to walk in beauty and harmony with the natural world that we urbanized people so seldom see.

The only reason to bring a watch on  a trip is to time the water purification pills.

Like a Zen master, one of the young assistant scoutmasters at my son’s troop used to answer any question like: “when will we get there?” or “How far have we gone?” or “How far do we have to go?” with the simple answer “10.2″

Until you understand the beauty of that answer, you won’t really enjoy the sport.

Philmont was (and is) great and I highly recommend it to anyone who can go.

Ten days with one shower is not bad in the dry New Mexico mountains. We smelled better than the homeless do here in Atlanta.

I have some quick gear pointers and then a few crew management pointers.

• My lightweight pack worked very well. On the first day with 5 days of Philmont’s heavy and bulky food and 5 liters of water it weighed 32 lbs which was the lightest in the crew. At the finish with some food, trash and about a liter of water it weighed 25 lbs. (my son’s was 20 lbs - so he is still a better light-weight backpacker than me).
• I used an outdoor research waterproof extension bag to hold the food and some extra’s (belts to be branded, a pack of jolly rancher candy (wrapped in paper as a surprise), and an extra crew shirt for the ranger).
• The shoulder pockets for the miraposa plus were not very useful - in the end I just kept the camera in a short pocket.
• We just used the chemical sterilizer - it was fine.
• The six moon design lunar solo is well adapted to Philmont conditions - I could always find place for the tent pegs (it’s not self-supporting) and the condensation was usually not too bad. The amount of condensation was a pretty accurate predictor of the amount of rain that the afternoon thunderstorm would produce. I’d make sure to bring a microfiber towel (get them cheap in the auto section of Target) along to wipe off when it is bad. I expect that any decent single layer ’silon’ tent will work well too.
• Make sure everyone brings an extra 2.5 L platypus and fills it for dry camps.
• bring a good book that you can stand to re-read.
• bring at least one bandanna to use as a “dew rag” and wash to avoid “monkey butt”
• The Philmont issued food is too bulky and too heavy. It also contains cans and pouches of meat and tuna that are impossible to clean thoroughly in the backcountry and can contaminate packs. Even though we could get rid of trash every few days, this is a serious issue in bear and panther country. It needs to be completely re-thought. When I was a scout (30-35 years ago) we carried ten days of food with less trouble than three days of Philmont food. Food technology has improved since the 1970’s. You will also get sick of meat bars and (bad commercial) jerky for breakfast.

Crew management is a more difficult issue.

• Establish clear communications with the other adults as soon as possible. This should happen during the training. (This was my one big mistake).
• The isolation will amplify personality quirks - be sure you like and (more importantly) respect and trust each other before you enter the backcountry.
• Except for health and safety issues or immediate coaching for technique, be sure to go via the “crew chief” (the crew chief is the elected senior scout). Insist that the other adults do the same.
• Beware of protective fathers who want their “father and son” experience to be extra special. You will, if you have your son in the crew, have a “father and son” experience and get to know your child much better. But it is critical to give them the room to grow. A good scouter will always consider what is good for the crew and do his best to treat his son like any other scout.
• Try to avoid injuries early in the trek by keeping the speeds down so that everyone can keep up. Having an injured scout will demoralize the group, reduce efficiency, and ruin at least one person’s trek. We had a total for the crew of 2-3 minor blisters over this trek. Can you beat that?
• Do Not Worry About Speed Early in the Trek. There really isn’t any need to be anywhere at any time. Everyone will improve in condition and by the end the group will fly along easy trails at speeds that they could not approach in the beginning. (the long long steep uphills will still be slow).
• We used a caterpillar technique on half of the 41 river crossings between miner’s camp and black mountain. It rocks. (basically the first person across the bridge waits for the last - yells clear and everyone moves on. It gives everyone an even rest break and avoids the stop and start that happens otherwise).
• Learn to use the rest-step.  We beat a much older and physically stronger crew up Mt. Phillips by rest-stepping.
• Hydrate, Hydrate, Hydrate. Use the toasting method to “camel up”.
• There will be disagreements about issues, it’s normal.
• There will be disagreements about issues, it’s normal.
• Let the crew chief lead the crew.
• Scouting is a “game with a purpose” - sometimes you can play with the scouts in the game or start it off. Association with adults is one of the methods of scouting and treating the scouts as equals includes occasional playing around.
• The scouts will have a good time.

It looks like I have an interlude to write an entry after all.  One of the things our crew has done correctly is to get to high altitudes several days before starting the trek.  Training at 1000ft just isn’t the same as at 6-12000 feet.

It’s also given the crews (we’re sending two) time to get oriented.  I was quite worried about fitting in our odd scout (we’d moved from 8 to 7 as a result of a needed shift in adults), who I must add was my son - which only makes it harder.  So we finally got them together and viola it was solved.  One of the other pairs liked the idea of dividing a seven pound tent in three vs. a 6 pound tent in two and since they also were buddies with the odd scout everything was fine.  What a relief.

Otherwise we spent the day up Pike’s peak via the cog railway.  They still tell the same joke about having two springs (Manitou and Colorado Springs) to stop the train if the breaks fail that they told 30 years ago when I visited with my family.

We also walked around the garden of the gods which is a neat eroded sandstone feature and well worth the visit.

After dinner at a Mexican restruant,  Arceo’s where the food was good and inexpensive,  (really - the Carnita’s are quite different from how I’ve always done them and they had a Mexican coleslaw that was a cross between coleslaw and salsa) the boys returned and used the motel pool.

I feel a little as if I should end as Samuel Pepys “and so to bed”