Getting things done for your research

Our profession (if you can call it that) is what people call "creative", meaning that the process is as important as the results.  Sometimes it is very easy for the grad student to spend all his time on the process and left none for the results.  A technician told me, in his 25 years working at the university, he has perhaps seen only 1 out of 8 students who completed their research successfully.  Success, as in demonstrating how a new idea works and have the hard data to show, that is.  From my life as a grad student, I would say his estimate is in-line with my observations.

Unfortunately, for the outside world, this is perceived as "not getting things done".  What they think of use, however, has a hint of truth in it, and I would like to suggest ways to turn things around.

Mitigate your risk. Granted your success depends on a lot of things, luck being one of them, I think it is also the risky nature of graduate research.  You are charting unknown territory, and you won't know if something works unless you try it.   I have seen students going "all out", designing the most sophisticated equipment and ordering very expensive parts, only to see their plans fell apart for reasons unknown after investing only 4 years and almost $100k.  Now, if you are that grad student, will you spend that money and time to find out, or would you look for something that would provide feedback faster and cheaper?

Exercise judgement.  Good judgement and experience also help getting results faster.  If a professor suggests to you a possible solution that is not 100% certain, has a 8-week waiting time, and costs 600 bucks, you might want to look elsewhere for a solution.  In my case, I managed to find one that worked every time, worked that very evening, and cost nothing.  Always ask yourself if the end justifies the means, and if not, move on to something that does.

Know your limits.  No matter where you are, at some point you will end up doing things you are not very good at.  They will be the bottleneck of your progress.  Out of bravado (I am still accused of that once in a while) one would try to trek even further into unknown territory and get themselves exposed to further risk.  For things you are not good at, give it a day or two to try.  If the progress don't look too encouraging, look for ways to get around them, quickly.

Accept failures. Sometimes things stagnate out of fear of failure.  I waited about 2 months before I started my very first field test, because I was honestly afraid of seeing something to go horribly wrong.  It did, and it did throughout.  Thankfully things were minor, and could be fixed without causing injury.  The trick is to accept that things will go wrong, but take every steps to minimize your exposure if it does.

More difficult to deal with is, perhaps, when "bad" results arrive.  I had those on my desk, and it was discouraging.  I eventually found out what happened, but only a few months later.  I must admit, a trip to the supervisor's office might not give me a solution right away, but it would give confidence, and would have shaved 2 months off.

Clean up after yourself.  In the chaos of the grad students' labs, it is tempting to make patchwork solution that works right away to get you out of here.  It might sound like a good idea, but it slows everyone down.  A very able colleague got some results, got out, but it didn't take long to find out how he did it.  Good that we did a very thorough check (1 week) of the lab before we started.  Most of that time was spent on weeding out things that he introduced during his stay.

Grad research is no different than other types of work.  True, there is a lot more risk involved, but it is possible to moderate your risk with good judgement.  Once in a while, things will not work the way you wanted, but it is time to look for other interpretations (through other opinions).  Most importantly, knowing that you are not working alone, help others get things done as well.

Presentations for the introverted

Where I went to grad school, there used to be a tradition called the annual seminar.  Students, with their research at various stages of incompletion, were asked to present their work, to be evaluated by the presiding professors. Unfortunately, these presentations almost always devolved into somewhat of a Roman gladiatorial spectacle. The whole system was eventually replaced by a more constructively themed orientation session.

Presentations are unavoidable.  You want your work known to your community, or at least to the right people.  You have 20 minutes to summarize years of hard work.  But you didn't have much time organizing your results, let alone presenting them.  And everyone is on the same boat so they cannot help you in that department.  How do you conquer your butterflies, or at least make them fly in formation?  I have a few notes that I would like to share, and I hope this would make you more confident over your upcoming presentation as well.

You are in control.  When you are presenting, they want to listen to your research.  Stop guessing what your audience want to hear and focus on what you want to say.  You would be surprised how much your audience appreciates it when someone finally takes the driver's seat.

The 10-20-30 rule.  Guy Kawasaki proposed this dead simple rule of thumb: 10 slides, 20 minutes, 30 point font.  Mind you, these rules were meant for start-ups pitching ideas to venture capitalists.  So here is the modified form for grad students:  10 topics, 20 minutes, 30 point font.  Typically these are the topics I cover

1. Introduction (1-2 slides)
2. Your problem (1 slide)
3. Your (proposed) solution (1 slide)
4. Theoretical background (1-3 slides)
5. General solution strategy (1 slide)
6. Each aspect of your solution (1 slide per aspect of your research)
7. Progress/results of your strategy (1-2 slides per aspect)
8. Interpretation of your results (up to 1-2 slides per aspect)
9. Conclusions / recommendations (1-2 slides)
10. Future work (1 slide)

So you have a minimum of 10 slides.  My research consisted of 4 aspects: analytical, experimental, numerical, and practice.  My presentation ended up to be 24 slides, including the title and thank-you slide.  Is it too little? Hell no.  Also, I have a 40-minute version of the same presentation with (just) 36 slides.  But this extended version contains no extra crucial information that the 20-minute version does not provide, other than additional graphs and figures (i.e., eye candy).

Know what you want to say.  There is a theme, ain't it?  Even though you don't have concrete results, there is still a point to your presentation.  Make this clear to your audience throughout, and you will become confident with it.  Also, it you can help it, do not present items that you don't feel confident presenting, just for the sake of filling up your time slot.  If you have to bring it up, put it at the end as "future work", instead of in the results section as "work in progress".

Less is more. As a consequence of using at least 30-point fonts on your presentations, you need to discriminate your information.  Figure out the one very important thing that you want to tell your audience with every slide and stick to it.  Everything else is either supporting evidence or embellishment.  Also, mainly out of fear losing track themselves, many presenters load their slides with words and recite them off the screen.  If you are the forgetful/nervous type (I don't blame you), use cue cards.  When lights are low, as it is almost always in presentations, no one will see you reading off cue cards.

Pace your way, your way.  You are the second element of the presentation, and don't forget, they are coming to see you.  Engage your audience with explanations, antedotes, and examples.  These things compliment the main points on your presentation, and makes things interesting for your audience.  You can also control the flow of your presentation much more easily with the amount of verbal communication.

Treat equations like figures.  If you have to present equations, treat it like you would for a figure.  Direct the audience to what are essential (i.e., important terms and their implications, contributing parameters, trends, etc.) instead of reciting the equation term for term.  Like this title suggests, also do the same for figures.

No bells and whistles.  Slide animations and transitions effects are lame.  So are weird slide layouts and inappropriate background and color schemes.  On the other hand, black text on white background is much more hypnotic than most minimalists think.  Find something easy on the eye.

Practice with reason.  I don't usually rehearse rigorously, because I run a risk of "scripting" it, and i will get nervous if I don't remember what I am scripted to say.  Instead, go through the slides quietly, make notes on them.  Present it to your colleagues for casual feedback.  The best way is to go, is knowing what you want to say, and say it.  If there is something the audience does not understand and care about, they will ask for clarification.

Going through these steps will force you to think very hard over your work, and what you should (and should not) present.  By the time you are done, you will have a very solid idea about the presentation.  There will be the first two minutes of awkwardness when you start presenting.  Once you find your momentum, however, you will be at ease, and so will your audience.  Again, there is no way to get rid of the butterflies in your stomach, but at least they will fly in formation.

Postscript: Too much information already?  At least take this one home. Treat your presentation as if it cost $20 per slide five secretary-days to prepare, just like the good old days before Harvard Graphics almost 20-odd years ago (that's even before Mugu's time).  I bet you will change the way you think about presentations very quickly.

Advice for thesis preparation

My thesis seemed like a monumental undertaking at the time of writing.  I started, scrapped and restarted three times, over two years before I gained enough momentum to get that all done.  I was still waiting for a lot of stuff to come together, mind you.  It was not exactly easy

If your grad department lets you compile a publication-based thesis, you're lucky.  For everyone else, I hope the following antedote would help all aspiring writers to get this done and get on with real life (quickly, hopefully)!

Don't panic. Often, you don't know what is missing in your research until after your start writing.  When you have one of those moments (and you will have many), take the time to put the missing pieces together.  You will become more confident with your work and, if not, you will at least know what to do to for improvement.

Use LaTeX.  LaTeX handles all the formatting for you, so you don't have to worry about your figures jumping all over the place.  For references and equations, its BibTeX and math mode will save you a lot of time and mouse clicks than their WYSIWYG alternatives.  Besides, it doesn't cost you a dime to pick up the legal version.  The lack of proofing tools might be a problem, but LaTeX front-ends like Texniccenter usually come with a spell checker.  It's not going to be easy learn, but you will be glad you LaTeX'd your thesis.

Get to the point.  You want your readers to know exactly what your problem is, what you did, and what your findings are.  So your introduction and conclusions/recommendations chapters should be short and direct.  I am talking about less than 10 pages each.  Literature review contains everything under your universe, and no one cares if you can't organize it.  Chapters on your research should contain just enough detail to guide your readers along, but not so much as to detract them.  All other interesting but not directly relevant information can be placed in the appendices.

Think hard about the scope.  When you are very confident about your research direction, invest the time to write down your thesis scope.  These few paragraphs will be a reminder of what you should be doing for the rest of your degree.  If you are on the right track, your scope would only need minor adjustments at most.

Use B/W figures whenever possible.  Unless it's a thesis on Visual Arts, chances are, your thesis will be printed or photocopied on B/W at some point, very likely by your committee.  B/W figures are much easier to read and reproduce.  In addition, it also forces you to look for ways to present the data effectively.  After all, admit it, you used color because you just wanted to copy and paste what your favorite charting program spat out.

Simple language.  If you know already what you want to write, then jot it down right away instead of toiling over a better way to write it.  It might not sound very elegant, but it got the point across, and (most of the time) that was all that matters.  If you still don't like how you wrote it, you will come back to it, which brings me to the next point ...

Proofread your thesis backwards.  A neat little trick I learned from my high school English teacher.  Start from the last page, and work your way back to the first.  This forces you to go through everything, and you will catch a lot of typos and grammatical errors.  You will also have the opportunity to rigorously examine your writing.

Save acknowledgement for last.  Having the acknowledgement page written out first gives you a false sense of accomplishment.  You will slack off, or wasting time embellishing this page.  When the going gets tough, the temptation will be there to write it.  Without this page to distract you, you will find yourself more focused, more motivated (maybe) to devote your energy on other parts of the thesis.