Toxicity of Grad School

I've written before about how difficult it is to finish a PhD -- specifically, the middle portion is very depressing and it often difficult to summon the motivation to continue.

There's a whole long, boring back-story to this that I won't go into, but one of the regrets that I have about that era is that I declined a lot of wedding invitations. Because I was a single guy in my 20s with not a lot of social skills, I didn't think that it was a big deal. The weddings were in California, or Texas, or wherever. And I figured that the wedding party probably wouldn't even notice that I wasn't there, might even welcome not having to pay for my plate at the reception, or want to have a seat that they could fill with someone else. Now that I'm married and have a completely changed perspective on weddings, I realize how misguided my thinking was.

Looking back on my life when I was in grad school, though, I realize now that the real reason that I declined a lot of the invitations were because I was depressed and embarrassed about being in grad school. My perception of myself was that I was languishing away in grad school, while all of my peers were working in real jobs, getting married, having children, and getting on with their lives. Most people don't understand what the PhD process is like, or why you would even want to get one. Can you imagine being at a wedding and having this conversation?

"Terence, it's been a while! What have you been up to?"
"Uh, I'm working on a PhD right now."
"Wait, you didn't get enough school in college and want more?"
"Uh... something like that."
"You're almost 30 and you're still a student?"
"It's not really like I'm a student..."
"Well, are you almost done with your PhD?"
"Yeah, I'm probably about a year away from graduation."
"Wait, isn't that what you said last year? And the year before that?"
*sigh* "Yeah..."
"And you're not married?"
"Uh, no, but this conversation is making me think about using heroin."

By the way, I don't harbor any resentment towards people who didn't come to our wedding. By the time we got married, everyone had children and in-laws, limited vacation time, and all kinds of other complications. It wasn't like when I was in grad school, and could take off whenever I wanted to.

Perceived Progress in Grad School

I was recently talking with some friends who are still in grad school. One of them joked that if you plotted Perceived Progress on your PhD against Actual Progress, it would look something like this:


You know, there is that long period at the beginning (region A) where you are making progress, but it doesn't actually feel like you're getting anywhere. And then there is that interminably long period at the end (region B), where you think that you are basically finished -- except that you're there for a year or more before you actually are able to finish (because you weren't as close to being finished as you thought you were).

The above figure is really funny (to anyone who has been in grad school), precisely because it is so true.

However, you can't forget those pesky committee meetings, where you discuss your project with the professors who are in the position to approve or disapprove your research plan. Those always seem to go horribly wrong, either because they correct your perception of how far along you are, or because the goalposts get moved ("Hmmmm, that's interesting. You should take more data to support your argument..."). So, I proposed that the plot should actually look more like this:


There are still the characteristic behaviors at the beginning and end. But you can't forget about the inevitable committee meeting right at the end, in which the setback you receive is enough to make you want to give up and become a farmer.

At any rate, the point is that at most points, you perceive yourself as being closer to the end than you actually are.

Reviewer Comments

I have commented before on the peer-review system that most technical journals and conferences use to screen manuscripts that are submitted. Let's say you want to get a paper published. You write up the manuscript and submit it. Then other researchers, who are ostensibly your peers, will review it and make their recommendation about whether it should be published, published after major revisions, or rejected. And almost all reviewers will suggest ways that the manuscript can be improved.

There are many positive and negative qualities about this system. One of the good things is that (when the system is working correctly), your work is being evaluated by people who are qualified to know whether the work is novel or not.

Recently, I received reviewer comments on a manuscript that I had submitted to a conference. And I couldn't help but chuckle because one of the reviewers wrote:

This paper is an excellent example of technical writing, from the organization of the paper, to the diction and grammar, to the clarity of the explanations of difficult subjects. From that point of view, it was a pleasure to read. Unfortunately, the statistics performed are all completely wrong.

I should note that the above is a paraphrase of what I actually received, and not a direct quote. What a slap in the face!

One of the negative things about the peer-review system is that you don't know who your reviewers are. So, you can't tell if they are a graduate student who doesn't know what they are talking about, and there is no mechanism to refute the claims given by the reviewers. So, if the reviewers are wrong, your only recourse is to rewrite the manuscript to be clearer.

PhDs

I apologize in advance for rambling on in this post... A few of our friends are in that last tiny stretch towards finishing their PhDs, and it has reawakened some of the thoughts that I have had about the whole PhD process.

When I meet people for the first time (either in my personal life or professionally), I almost never bring up that I have a PhD. A good chunk of the reason why is because of this underlying feeling that I have that universities will give PhDs to anyone. This is kind of a half truth.

Looking back on the whole grad school thing, it is easy to say that the PhD qualifier was the hardest part of the process. Someone once told me that once you get past the PhD qualifier, the only thing required to get a PhD is perseverance. But now, looking back on things, I'm not so sure that this is true.

If it were really that easy, then we wouldn't know so many people who have gotten divorced during grad school. The truth is that getting your PhD requires a unique set of skills - the ability to motivate yourself, a ridiculously high tolerance to depression, and the arrogance to stick to your guns when everything and everyone is telling you that you are heading down the wrong path with your research.

When you get to that middle portion of the work -- it's been a few years since the qualifier, but you know that you are a few years away from defending, too -- it's almost impossible to continue. Any normal person would give up or get depressed about the mountain of work that lies ahead. And what's worse is that no one cares (except for you) if that work gets done or not. You're not curing cancer (well, I wasn't, anyway), you're not fighting for a promotion (unless you consider being head lackey in a university lab to be a step above being a regular, run-of-the-mill lackey).

And that's just if you want to get a PhD. It's ten times worse if you actually want to get a good PhD (one that will make you marketable). Just thinking about all of this is making me feel depressed again. It's a wonder that grad school doesn't produce more alcoholics...

Conference Peer Review Systems

I am becoming increasingly frustrated with the peer-review systems that are used by major conferences and journals.

Let's say that you write an article or a paper, and submit it to a technical conference or a journal. Copies of your manuscript will then go out to other researchers in the same field, who evaluate the work, decide whether or not it should be accepted, and suggest (or mandate) changes that should be made to the manuscript before it is published.

The idea is good, because only people who are doing research in a similar field of study are going to be qualified to fully understand what is contained in the manuscript and evaluate the quality of the research.

The problem is that by definition, other people in a similar field of study are your competitors. They are competing with you for funding from government agencies, publications, recognition in the field, etc.

Increasingly, what I have seen is that the comments coming back from reviewers are suggestions to cite certain papers or works. And although the identities of the reviewers aren't known, it appears that most reviewers are suggesting that authors cite works that they (the reviewers) have published in the past. Papers that are cited more often gain recognition in the field (along with their authors), so people are definitely motivated to try to get their works cited as much as possible.

It seems to me that the peer-review system is broken (or is, at least, close to being broken). Instead of receiving constructive comments, authors now seem to be receiving a lot more requirements to cite particular works. I'm not sure if I perceive the increase of this problem because I was much more naive a few years ago, or because the types of conferences and journals I have been submitting manuscripts has changed over the years.

But I'm not sure what the solution is. It is a hard sell to convince journals and conferences to police their reviewers, since 1) these organizations are already overworked as it is, and 2) reviewers are difficult to find.

Akron Pride

Now that I work in Akron, OH, I am constantly hearing all kinds of Akron propaganda -- you know, how every city claims that they invented everything and that they are the linchpin of human civilization... You wouldn't believe the things that the world has to thank Akron for.

The latest one that I heard was the the zipper was invented in Akron -- a claim that seemed really unlikely. So, I did some "research" (which is really just a fancy word for doing a Google search and then believing any crap that I found on the internet).

Here's what I found... A number of sources cite Whitcomb Judson as the first person to try to market a device for clothes fastening which caused two edges to become attached to each other. Judson was from Chicago, and according to the Wikipedia article on zippers [wikipedia.org]:

Judson, together with business partner Harry Earle, founded... the Universal Fastener Company. The design deficiencies [of the fastener], combined with difficulties in getting the machinery needed for mass production to work... led to a series of reorganizations and... relocations, first to Catasauqua, Pennsylvania; then to Elyria, Ohio; Hoboken, New Jersey; and finally Meadville, Pennsylvania.

"Gideon Sundback, a Swedish-born immigrant, joined the company, then called the Automatic Hook and Eye Company, in Hoboken, in 1906. At that time the company's product, still based on hooks and eyes... had a strong tendency to pull apart, and wasn't any more successful than the previous versions. Sundback finally solved the pulling-apart problem in 1913, with his invention of the first version of the zipper based on interlocking teeth, the 'Hookless Fastener No. 1'."

So, it seems clear that the invention of the attachment device had nothing to do whatsoever with Akron. According to this source [about.com]:

The popular 'zipper' name came from the B.F. Goodrich Company [which was headquarted in Akron at the time], when they decided to use Gideon's fastener on a new type of rubber boots or galoshes and renamed the device the zipper, the name that lasted.

This assertion is supported by the Wikipedia article about the University of Akron [wikipedia.org] which states:

The University of Akron's athletic teams are known as the 'Zips,' originally short for 'Zippers,' overshoes that were nationally popular in the 1920s and 1930s... The university's mascot is 'Zippy,' a kangaroo... Zippy is one of only eight female college mascots in the United States.

The above reference claims that the mascot was chosen because the inventor of the zipper (Judson) was of Akron, but I could find no source which supported that claim.

So, it appears that Akron's claim to fame was the marketing of the first products in which the attachment device was popular, and the invention of the name "zipper", but not the invention of the attachment device itself.

Slide Rules

I have been looking for a slide rule for a while now. You know those people who believe that they need to know how to light a fire using nothing but twigs, just in case that nuclear war destroys all civilization, and someone will need to light fires without having lighters?

Well, I'm the same way. In my mind, the definition of a cataclysm threatening the future of mankind is if we can no longer use Matlab [wikipedia.org]. And if there are only 100 people left on the face of the planet, there damn well better be someone that still knows how to calculate logarithms and trig functions.


Anyway, I got some slide rules and some manuals from Heimlich's grandfather. So, in my free time over the holidays, these are the sorts of things I will bring with me when I curl up in front of the fire.

Are We Being Watched by Flying Robot Insects?

I keep seeing links to articles written by alarmists. Many speculate about the latest robotics projects DARPA (the Defense Advanced Research Projects Agency) is currently working on, or whether we're already being watched by flying robot insects [typepad.com].

I suspect that most of us would be horrified if we really knew what was state of the art.

There have been a lot of articles in the press about the latest crop of robotic insects that are being developed (such as the one at the Harvard Microrobotics Lab [harvard.edu], or the one being developed at Berkeley [berkeley.edu]). Here's one of the articles [economist.com] that I was speaking of. Many articles cite the difficulty of generating sufficient lift from flapping wings, or getting the right motion to utilize phenomenons such as wake capture or delayed stall (you can't just flap wing-like things up and down and expect to generate lift).

But most people probably don't know that the problem has already been mostly solved. The CIA's Office of Research and Development created an autonomous flying insect robot that was literally the same size as a dragonfly. It carried its own fuel, and could be directed by a human operator. It was called the "insectothopter", and it was developed in the 1970s. It is now sitting in a display case at CIA headquarters (which I have actually seen). Here is an article [boston.com] that mentions its existence. I wonder what has been done in the 30 years since...

DARPA Urban Challenge

Well, the DARPA Urban Challenge is over [theregister.co.uk] and the winner (Carnegie Mellon) has been announced. The DARPA Urban Challenge was a competition between teams developing autonomous cars that have to run around in an urban environment, obeying traffic laws, interacting with other vehicles, and planning and replanning their routes without human guidance or intervention.

It has been a little difficult getting detailed information. I am interested because my PhD adviser has been involved, along with several of my former co-workers. Team Case apparently did not get selected from the National Qualifying Event to go to the finals. I caught this interesting little blurb [wired.com] on Wired about what happened to them:

"It was a case of too much work, too little sleep, and too much voltage that nearly doomed Team Case's entry in the DARPA Urban Challenge. One team member accidentally connected 24-volt laser range finders belonging to the team's robot dune buggy, named Dexter, into a 48-volt power supply, frying them instantly."

"Dexter is the only car in the race without any seating for humans, and it's the only one to be built from the ground up as a robot car. It made a good showing in the 2005 Grand Challenge, covering 81 miles of the 132-mile course for Team ENSCO before a blown tire forced it out."

The blog on Wired [wired.com] actually has had the most detailed information that I have been able to find on the event.

Headshots

Over the years, I have submitted quite a number of technical papers to be published in conference proceedings or journals, etc. One of the things that has been becoming more common is for those organizations to request a picture of yourself to publish with your article or paper. I can't tell you how annoying this is. I guess this is one of the drawbacks to the ubiquity of digital cameras.

What am I, an actor? What difference does it really make to the reader to know what I look like? Ok, the real problem is that I look like a moron in every single picture that has ever been taken of me. Here are a couple of examples:



So, which one of the above makes me look the most like a responsible researcher?

My Bio

So, back at Case [case.edu] where I used to work, my former co-workers are submitting another journal article. Part of the research that performed there supports a conclusion that they are trying to assert in a journal article, so they added me as an author (5th or 6th in a list) to the journal article so that they could use some of my research. And that's totally fine, because it doesn't really matter to me either way -- I'm actually happy that they're getting more mileage out of that research.

Anyway, then I found out that I was going to have to submit a bio... Ugh. They just wanted to have these short blurbs at the end of each article, with a picture of each author, so that the readers would have some idea of whether the authors were qualified to make the conclusions they are making, etc.

I find it very difficult to write a summary of my life or accomplishments. I mean, I usually go out of my way to make sure that I don't sound like I am bragging about myself. And writing a bio about yourself just seems to close to bragging, I guess. Anyway, it makes me very uncomfortable. Here is what I ended up submitting:

Terence Wei received his Ph.D. in Mechanical Engineering from Case Western Reserve University in 2005. While performing research at the Biologically Inspired Robotics Laboratory at Case, he developed many types of vehicles, including single-legged hopping robots and six-legged walking robots. He has specially designed dynamic simulations for the purposes of optimizing robotic design and robot-ground interaction. He has also created several specialized apparatus for the purposes of obtaining insect locomotion and climbing behavior data, as well as novel feeding mechanisms for robotic manufacturing workcells. After finishing his Ph.D., he received an Intelligence Community (IC) Postdoctoral Fellowship. As a postdoctoral fellow, he worked on the design and integration of insect-inspired attachment mechanisms for climbing robots. His other interests include mechanism design and vehicle dynamics.

Now that I've finished writing it, I'm not sure what is weirder -- trying to figure out what accomplishments in my very short career are really worth writing about, or seeing my entire life summarized in a few sentences.